VLSI, Communication and Signal Processing: Select Proceedings of the 5th International Conference, VCAS 2022 9819909724, 9789819909728

This book covers a variety of topics in Electronics and Communication Engineering, especially in the area of microelectr

218 33 52MB

English Pages 866 [867] Year 2023

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Contents
About the Editors
Machine Learning-Based FSOC Link Performance Estimation
1 Introduction
2 Results and Discussions
2.1 FSO System
2.2 ML Model Evaluations
3 Conclusion
References
BER Efficiency of Outdoor Optics Links Using Hybrid-SIM with Pointing Errors Operating on Extreme Turbulence Regime
1 Introduction
2 System and Channel Model
3 Bit Error Rate Evaluation
4 Numerical Analysis
5 Conclusion
References
An Exhaustive Review of Various Optical Devices for Biomedical Applications
1 Introduction
2 Background of Various Optical Devices for Bio-medicinal Applications
3 PCF SPR Sensor Modeling for Human Fluid Analysis
4 Conclusion
References
Cell Optimization and Realization of XOR-Based Logic Design in QCA
1 Introduction
1.1 Basic QCA Cell
1.2 Clocking Scheme in QCA
1.3 3-Input Majority Gate (MG)
2 Literature Survey
2.1 2-Input XOR and XNOR Gate
2.2 3-Input XOR and XNOR Gate
2.3 1-Bit Full Adder
2.4 3-Bit Parity Generators
3 Proposed Structures Using QCA
3.1 2-Input QCA XOR and XNOR Gates
3.2 3-Input QCA XOR and XNOR Gate
3.3 One-Bit Full Adder Design
3.4 Parity Generator Circuit Using QCA
4 Simulation Result and Discussion
5 Conclusion
References
Effect of Scandium Doping on Crystallization Kinetics and Glass Transition of Te(1−x) (GeSe0.5) Scx (X = 0.1) Glassy Alloy for PCM Applications
1 Introduction
2 Experimental Methods
3 Theoretical Background
4 Results and Discussion
4.1 DSC Curves
4.2 Glass Transition
4.3 Iso-conversional Methods
4.4 Crystallization Kinetics Analysis
5 Conclusion
References
Design of Ultralow-Power and High-Speed Comparator Using Charge Sharing Technique
1 Introduction
2 Some Predefine Architectures of Double-Tail Comparator
2.1 Double-Tail Dynamic Comparator (DoTDC)
2.2 Hybrid Double-Tail Dynamic Comparator (HDoTDC)
3 Proposed Circuit
4 Results and Performance Analysis
5 Conclusion
References
Underwater Image Enhancement Using Color Correction and Fusion
1 Introduction
2 Methodology
3 UW Image Quality Measure
4 Results and Discussion
5 Conclusion
References
Study and Analysis of Three-Stage Single-Miller CMOS OTA
1 Introduction
2 Three-Stage CMOS OTA Circuit
2.1 HV Amplifier [2]
2.2 Three-Stage ASMIHF Amplifier with Slew Rate Enhancer [3]
2.3 Three-Stage Single-Miller CMOS OTA [1]
3 Simulation Results and Analysis
3.1 Steady State Analysis
3.2 Transient Analysis
3.3 Parametric Analysis of CMOS OTA
4 Comparative Analysis
5 Conclusion
References
Stacked Bi-LSTM Network and Dual Signal Transformation for Heart Sound Denoising
1 Introduction
2 Proposed Technique
2.1 Signal Transformations
2.2 Proposed Network Architecture
2.3 Dataset
2.4 Model Configuration and Training Setup
2.5 Experimental Setup
2.6 Testing
3 Results
4 Conclusion
References
Abnormality Detection in Heart Using Combination of CNN, RNN and U-Net
1 Introduction
1.1 Motivation
1.2 Contribution
2 Proposed Work
2.1 Architecture of Proposed Model
3 Result and Analysis
3.1 Result Produced by LSTM
3.2 Result Produced by Xception Network
3.3 Result Produced by Proposed Work
4 Conclusion and Future Scope
References
High Speed LVDS Driver Design with Fast Settling Common Mode Feedback Circuit
1 Introduction
2 Current Mode LVDS Circuit Design
2.1 Circuit Bandwidth Calculation
3 Proposed Circuit to Improve upper V Subscript CMVCM (Common Mode Voltage) Settling Time
3.1 Proposed Common Mode Feedback Circuit
3.2 Stability of Proposed Circuit
4 Results and Discussion
References
Design and Analysis of Low Power Frequency Divider Circuit
1 Introduction
2 Frequency Divider Circuit Design
3 Analysis
4 Conclusion
References
Static Single Phase Contention Pulsed Latch for Low Voltage Operation
1 Introduction
1.1 Introduction
1.2 Literature Review
1.3 Theoretical Background
1.4 Proposed Work
1.5 Result and Discussion
2 Conclusion
References
Design of Y-shaped Multiband Antenna Using the Parametric Approach for Wireless Networks
1 Introduction
2 Design Analysis and Parametric Approach
3 Results and Discussion
3.1 Reflection Coefficient (S11)/ Return Loss
3.2 VSWR
3.3 Radiation Pattern
4 Conclusion
References
Design and Implementation of 1 × 4 & 4 × 4 RAM Using Quantum-Dot Cellular Automata
1 Introduction
2 Background of QCA
2.1 QCA Binary Wire
2.2 QCA Clock
2.3 QCA Basic Gates
3 Proposed System
3.1 1 × 4 RAM
3.2 4 × 4 RAM
4 Results and Discussion
5 Conclusion
References
FPGA and ASIC-Based Design of Fast and Low Power SEC-DAEC and SEC-DAEC-TAEC Codecs
1 Introduction
2 Construction of H-Matrix for Proposed SEC-DAEC and SEC-DAEC-TAEC Codes
3 Design of Proposed SEC-DAEC and SEC-DAEC-TAEC Code
4 Analysis of Different Design Aspects
5 Implementation Results
5.1 FPGA-Based Implementation Result
5.2 ASIC Implementation Result
6 Conclusion
References
Reliable Implementation of Finite State Machine for On-Board VLSI Designs
1 Introduction
1.1 Problems with an FSM
2 Designing of the FSM
3 Verification of FSM
3.1 Static Design Rule Analysis
3.2 Functional Simulations of the Design
4 Case Study
4.1 Synthesis Options
4.2 Result Analysis
5 Conclusion
References
Performance Improvement of SAC-OCDMA FSO System Under Rain and Snow Conditions Using Different Zero Cross-Correlation Codes
1 Introduction
2 Code Construction
3 FSO Attenuation
3.1 Rain Attenuation
3.2 Snow Attenuation
4 SAC-OCDMA FSO System Scheme
5 Simulation Results and Discussion
5.1 Comparative Evaluation of the SAC-OCDMA FSO System’s Performance Using Two Different Codes (ZCC and DW-ZCC)
5.2 Simulation of 7 Users SAC-OCDMA FSO System Performance Using DW-ZCC Code and FBG Filters
6 Conclusion
References
Robust and Imperceptible Color Image Watermarking Using LWT, Schur Decomposition, and SVD in YCbCr Color Space
1 Introduction
2 Preliminaries
2.1 YCbCr Color Space
2.2 QR Code
2.3 Lifting Wavelet Transform (LWT)
2.4 Schur Decomposition
2.5 Singular Value Decomposition (SVD)
3 The Proposed Watermarking Technique
3.1 Embedding Process
3.2 Extraction Process
4 Results and Discussion
5 Comparison of Results
6 Conclusion
References
Design of Three Slot-Based Multiband MIMO Circular Patch Antenna
1 Introduction
2 Antenna Design
2.1 Antenna Design Procedure
2.2 Single Element Design
2.3 MIMO Configuration
3 Simulation Results
3.1 Return Loss
3.2 Isolation
3.3 VSWR (Voltage Standing Wave Ratio)
3.4 ECC (Envelope Correlation Coefficient)
3.5 Radiation Pattern
4 Conclusion
References
An Imperceptible Semi-blind Color Image Watermarking Using RDWT and SVD
1 Introduction
2 Watermark Embedding and Extraction Process
3 Simulation Results and Discussion
4 Conclusions
References
Construction of Deterministic Measurement Matrices Using Polyphase Sequences for Compressed Sensing Radar
1 Introduction
2 CS Radar
2.1 LFM Signal
2.2 Dictionary Matrix (Ψ)
2.3 Measurement Matrix (Φ)
2.4 Sparse Signal Recovery
3 Results and Discussion
4 Conclusion
References
A Non-invasive Planar Resonant Microwave Sensor for Unknown Liquid Permittivity Estimation
1 Introduction
2 Principal of Microwave Resonant Sensor
3 Design of Proposed Sensor
4 Analysis of Proposed Sensor
5 Sensitivity Analysis of the Sensor
6 Mathematical Modelling for Characterizing Unknown Liquid Sample
7 Validation of the Derived Numerical Model
8 Conclusion
References
Design of SRAM Cell for Improved Performance
1 Introduction
2 Previous Work
3 Proposed Work
3.1 Circuit Description
3.2 Operation of Circuit
3.3 Device Parameters for Proposed Circuit
3.4 Results and Discussions
4 Conclusion
References
Performance Analysis of Various Charge Pump Topologies for PLL Application
1 Introduction
2 Charge Pump Non-idealities
3 Different Topologies of Charge Pump
3.1 Classic Current Steering Charge Pump
3.2 Two-Stage Charge Pump
3.3 NMOS Switch Cascode Charge Pump
3.4 Low Mismatch High Speed Charge Pump
3.5 Self-cascode-Based Charge Pump
3.6 OPAMP-Based Charge-Pump
3.7 Double-Ended Charge Pump
4 Results and Analysis
5 Conclusion
References
Low Power Radix-4 Booth Multiplier Design Using Pass Transistor Logic
1 Introduction
2 Study of Booth Multipliers
2.1 Chang et al. Model
2.2 Venkatachalam et al. Model
2.3 Ranasinghe et al. Model
3 Proposed Low-Power Radix 4 Booth Multiplier
4 Results and Discussions
5 Conclusion
References
Impact of Different Buffer Layers on Performance of Cu2O Based Solar Cell: SCAPS 1D Analysis
1 Introduction
2 SCAPS 1D Parameters
3 Results and Discussion
3.1 Basic Solar Cell Simulation
3.2 Effect of Absorber Layer Thickness Variations
3.3 Influence of Carrier Concentration and Defect Concentration of Absorber Layer
3.4 Effect of Absorber Layer Defect Concentration
3.5 Impact of Interface Defect Concentration
3.6 Influence of Buffer Layer Thickness and Electron Affinity
3.7 Impact of Temperature
4 The Optimized Device
5 Conclusion
References
Sum Rate Maximization Using Unlicensed and Licensed Bands in UAV-Assisted Cellular Communication
1 Introduction
2 System Model
2.1 Licensed Subchannels
2.2 Unlicensed Subchannels
2.3 Channel Model
3 Proposed Optimization
4 Simulation Results
5 Conclusion
References
Performance Analysis of mmWave V2V Communication Using Relay Vehicle for Advanced Safety Applications
1 Introduction
2 Relay Vehicle Transmission Model
3 Problem Formulation and Solution
4 Simulation Parameters
5 Conclusion
References
Low-Voltage Low-Power Fully Differential Double Recycling OTA with Positive Feedback
1 Introduction
2 Proposed Class AB OTA Employing Positive Feedback
2.1 Proposed OTA Small Signal Circuit Analysis
3 Simulated Results
4 Conclusion
References
Multilingual Indian Musical Type Classification
1 Introduction
2 Survey of Recent Deep Learning Approaches
3 Finding from the Survey
4 Conclusion
References
Design and Implementation of Matched Filtering and Timing Recovery Algorithm for IEEE 802.15.4 Digital Baseband Front End
1 Introduction
2 MATLAB/Simulink Simulation of Matched Filter
2.1 MATLAB Implementation of Matched Filter
2.2 Simulink Implementation of Matched Filter as Direct Form-1 FIR Filter
3 FPGA Implementation of Matched Filter as Direct Form-1FIR Filter
4 Symbol Timing Synchronization
4.1 MATLAB Implementation of Symbol Timing Synchronization
4.2 FPGA Implementation of Symbol Timing Synchronization
5 Results
5.1 Matched Filter Simulation Results after FPGA Implementation
6 Conclusion
References
Minimal Group Delay Multi-objective Finite Impulse Response Filter Design Using Salp Swarm Algorithm and Its Improved Version
1 Introduction
2 Problem Formulation of Multi-objective Small Group Delay Sparse FIR Filter
3 Multi-objective Salp Swarm-Based Optimization Algorithms
3.1 Modified Multi-objective Salp Swarm Algorithms
4 Result and Discussion
4.1 Case 1: Multi-objective FIR Filter Design
4.2 Case 2: Implementation of Multi-objective Optimization with Three Objectives
5 Conclusion
References
Performance Analysis of Downlink Cooperative NOMA System
1 Introduction
2 System Model
3 Performance Analysis of C-NOMA
4 Simulation Results
5 Conclusion
References
Design and Analysis of Noise Immune, Energy Efficient 1-bit 8T SRAM Cell
1 Introduction
2 Proposed 8T SRAM Cell
2.1 Operation of the Proposed 8T SRAM Cell
2.2 Power Analysis
2.3 Delay Analysis
2.4 Read and Hold Stability
3 Simulation Results and Discussion
4 Conclusion
References
Automated Design Rule Checker for VLSI Circuits Using Machine Learning
1 Introduction
2 Literature Survey
3 Proposed Architecture for DRC Violation Classification
4 Results and Analysis
5 Conclusion and Future Work
References
Design of Low Power High Speed 2nd Order Discrete Time Sigma-Delta Modulator Using Charge Shared Double Tail Dynamic Comparator
1 Introduction
2 Working Principle of Sigma-Delta Modulator
2.1 Oversampling
2.2 Noise Shaping
3 Implementation and Analysis of Modulator
3.1 Passive Integrator
3.2 Hybrid Passive-Active Integrator
3.3 Preamplifier
3.4 1-Bit Quantizer
4 Complete Model for 2nd Order Discrete Time Sigma-Delta Modulator
5 Simulation Results
6 Conclusion
References
Solar Powered Low Cost Air Quality Monitoring System
1 Introduction
2 System Design
2.1 System Design Schematic Diagram
2.2 Microcontroller and Its Specifications
2.3 Programming Flowchart
3 Sensors
3.1 Gas Sensor (MQ135)
3.2 Digital Sensor (BME280)
3.3 PM Sensor (PMS5003)
4 Power System
4.1 Solar Panels
4.2 Solar Charge Controller (TP4056)
4.3 Power Saving System
4.4 Power Calculations
5 AQMS Data Center
6 Results
7 Conclusion
References
High Gain Hexaport Millimetre Wave MIMO Antenna for 5G Service LMDS and N257 Band Applications
1 Introduction
2 Proposed Design Configuration
2.1 Design Implementations of the Single Element Proposed Antenna and Its Performance
3 Proposed Hexaport MIMO Antenna Simulation Results and Discussion
3.1 Electrical Simulation Results
3.2 Far-Field Results
3.3 MIMO Antenna Simulation Parameters
4 Comparison
5 Conclusion
References
Aspect Ratio Estimation of a Two-Stage Operational Amplifier
1 Introduction
2 Dataset Preparation and Pre-processing
2.1 Dataset Preparation Methodology
2.2 Dataset Pre-processing
3 Proposed Approach
3.1 Proposed Neural Network Architecture
4 Results and Discussions
4.1 Performance Comparison of Machine Learning Models
4.2 Performance Improvement of the Neural Network Model
4.3 Testing Data Versus Predicted Data Plot
4.4 Proposed Model Predictions
5 Conclusion
References
EM Power Absorption and RCS Analysis of Novel FSS-Based Broadband Radar Absorbing Structure
1 Introduction
2 EM Performance Analysis of Novel FSS-Based Unit Cell
3 Equivalent Circuit Model (ECM) of the Unit Cell
4 RCS Analysis
5 Conclusion
References
Performance Analysis of Pulsed Latches for Low-Voltage Operation
1 Introduction
1.1 Pulsed Latches
1.2 Energy Metrics
1.3 Other Design Issue
2 Review of Pulsed-Latch Design Using Different Techniques
2.1 Pulsed-Latch Design Using Low-Power Synchronous Circuit Design
2.2 Pulsed-Latch Design Using Low-Power Area Efficient Shift Register
2.3 Edge-Triggered Pulse Latch Design with Delayed Latching Edge Design
2.4 Conditional Push–Pull Latches With Energy Delay Product
2.5 Low-Power Pulse-Triggered Flip-Flop Design with Conditional Pulse-Enhancement Scheme
2.6 Semi-Dynamic Pulsed-Latch Flip-Flop
2.7 Self-Time Pulsed Latch For Low-Voltage Operation With Reducing Hold Time
3 Performance Analysis
3.1 Comparison of Parameter PLULPSC PLUAESR ETPL CPPL LPPTFF STPL
4 Conclusion
References
A Comparative Performance Analysis of Different Methodology Operational Transconductance Amplifier Using Cadence
1 Introduction
1.1 CMOS Based on the Operational Transconductance Amplifier (OTA)
1.2 The Single-Stage Based on Operational Transconductance Amplifier (OTA)
1.3 Two-Stage Operational Transconductance Amplifier
1.4 The Ultra-Low-Power and Reconfigurable, Two-Stage OTA
1.5 Two-Stage, Based on Fully Differential Miller OTA
2 Simulation Results and Discussion
2.1 AC Analysis
2.2 Transient Analysis
2.3 Comparative Analysis of Different Topologies of OTA
3 Conclusion
References
Checkpoint Snapshot Placement in the Cloud Data Center Using Fuzzy Inference System
1 Introduction
2 Related Work
3 Proposed Method
3.1 An Explanation of the Input and Output Variable's Membership Function (MF)
3.2 Projected Set of Fuzzy Rules for PSFIS1 and PSFIS2
4 Performance Evaluation
5 Conclusion
References
4-Disjoint Path Multistage Interconnection Network
1 Introduction
2 Background and Motivation
2.1 Background
2.2 Motivation and Contributions
3 Layout of the Proposed Design (4 DPMIN)
3.1 Topology
3.2 Connection Pattern
4 Performance Analysis
4.1 Reliability Analysis
4.2 Cost Analysis
5 Conclusion and Future Scope
References
Design of Low-Power High Performance SAR Analog-To-Digital Converter Based on MOS Capacitor DAC Structure
1 Introduction
2 Circuit Design
2.1 Sample and Hold
2.2 Comparator
2.3 Dac
2.4 SAR Logic
3 Simulation Result and Discussion
3.1 Sample-And-Hold Output (Transient Response)
3.2 Capacitive DAC
3.3 Comparator
3.4 SAR Logic
3.5 SAR ADC Output
4 Conclusion
References
FGMOS and QFGMOS-Based Super-Wilson Current Mirror and Its Application in Full-Wave Rectifier
1 Introduction
2 LVLP Design Techniques
2.1 Floating Gate (FG)
2.2 Quasi-Floating Gate (QFG)
3 Proposed Circuit
3.1 Super-Wilson Current Mirror
3.2 FGMOS-Based Super-Wilson Current Mirror
3.3 Proposed QFGMOS-Based Super-Wilson Current Mirror
3.4 Proposed Full-Wave Rectifier (in Current Mode) Using QFGMOS-Based Super-Wilson Current Mirror
4 Simulation Results
5 Conclusion
References
Review of the SAR ADC’s Energy Efficient Switching Scheme
1 Introduction
2 Conventional Switching
3 Monotonic Switching
4 Merged Capacitor Switching or VCM-Based Capacitor Switching
5 Tri-Level Capacitor Switching
6 Charge Average Switching (CAS) Technique
7 Detect-And-Skip Switching Scheme
8 Higher Side-Reset-And-Set (HSRS) Switching Scheme
9 Comparison of Reviewed Switching Scheme
10 Latest Innovative Switching Scheme
11 Conclusion
References
Design of an Elliptical-Shaped Meta-material for SAR Reduction Application
1 Introduction
2 An Intelligent Meta-material Structure
3 The Effect on Antenna
4 Conclusions
References
Improved Fuzzy Logic-Based Localization in Wireless Sensor Networks
1 Introduction
2 Literature Review
3 System Model
3.1 Mamdani FIS Model
3.2 Sugeno FIS Model
4 FIS-Based Localization Algorithm
5 Simulation Results and Analysis
6 Conclusions
References
Residential Level Short-Term Demand Forecasting Using ANFIS Model
1 Introduction
2 ANFIS
2.1 Adaptive Neuro-fuzzy Inference System (ANFIS)
2.2 Flowchart Diagram of ANFIS Model
3 Artificial Neural Network (ANN)
4 Results and Discussion
4.1 Description of Dataset
4.2 Performance Evaluation Criteria
4.3 Learning Environment Setup
4.4 Experimental Result
5 Conclusion
References
Signal Integrity Analysis in Coupled GNR Nano Interconnects Integrated with Buffer Repeaters
1 Introduction
2 Geometric Modelling of Interconnect Parasitic
2.1 Geometry of Cu Interconnect
2.2 Geometry of GNR Interconnect
3 Signal Integrity Analysis Using DIL Test Bench
4 Results and Discussion
5 Conclusion
References
Effect of Electrical Vehicle Charging on Power Quality
1 Introduction
2 Power Quality with EV System
2.1 Block Diagram Representation
3 Results and Discussions
3.1 Case-1:—When R-Load is Connected
3.2 Case-2:—When One EV-Battery is Connected
3.3 Case-3:—When Three EV-Batteries Are Connected
3.4 Case-4:—When Six EV-Batteries Are Connected
3.5 Case-5:—Mitigation Circuit
3.6 Case 6:—Grid to Vehicle and Vehicle to Grid Analysis
4 Conclusion
References
Design and Analysis of Ion Selective Field Effect Transistor for Biomedical Application
1 Introduction
2 Device Structure
3 Results and Analysis
4 Conclusion
References
Narrowband Data Waveform Development and Simulation for Achieving Low BER in Wireless Communication
1 Introduction
2 Baseband Processing Blocks
2.1 Baseband Processing Chain
3 Selection of Baseband Processing Blocks Based on Simulation
3.1 Modulations
3.2 Channel Coder
3.3 Pulse Shaping Filters
4 Simulation Result of the Finalized Chain
5 Conclusion and Future Works
References
Design and Analysis of CIC Decimation Filter Using Redundant Number System
1 Introduction
2 CIC Filter and Signed Digit-Based FIR Filter
2.1 Comb-Based Filter
2.2 FIR Filter
2.3 SD-Based Filter Operation
3 Computationally Efficient Compensation Filter
4 Proposed Design for Compensated Cascaded-Integrated-Comb Structure
5 Results and Discussion
5.1 Comparison of Passband Ripple in Magnitude Response
5.2 Comparison of Stopband Attenuation in Magnitude Response
5.3 Number of LUTs Utilized Filter Operation
6 Conclusion
References
Enhanced Energy-Aware Fault Tolerance Technique for Real-Time Task on Heterogeneous Multicore System
1 Introduction
2 Literature Survey
3 Motivation
4 Problem Statement
5 Models and Assumption
5.1 Platform Model
5.2 Task Model
5.3 Energy Model
5.4 Fault Model
6 Proposed Methodology
6.1 Task Partitioning
6.2 Task Priority Assignment
6.3 Frequency Assignment
6.4 Computation of Promotion Time
6.5 Enhanced MPB-PS Algorithm
6.6 Speed Fine-Tuning Algorithm
7 Experiment Result and Analysis
7.1 Impact of Utilization with Fault Rate
7.2 Impact of Utilization with Speed Level
8 Conclusion and Future Work
References
Design and DC Electrical Performance Analysis of SOI-Based SiO2/HfO2 Dual Dielectric Gate-All-Around Vertically Stacked Nanosheet at 5 nm Node
1 Introduction
2 Device Structure and Simulation Methodology
3 Result and Discussion
3.1 Effect of Nanosheet Width on Performance Matrices
4 Conclusion
References
Designing of Ternary to Binary Half Adder Using CMOS
1 Introduction
2 Methodology
2.1 Proposed Design of the Ternary Inverter
2.2 Design and Implementation of Ternary Logic Gates
2.3 Ternary to Binary Half Adder
3 Results
4 Conclusion
References
Performance Analysis of Differential Dual Stage Delay Cells of VCO
1 Introduction
2 Architecture Dual-Delay-Path Differential Ring VCO
3 Different Topologies Used to Tune Frequency in VCO
3.1 Conventional Dual-Delay Cell [20]
3.2 PMOS Varactor Delay Cell [20]
3.3 Dual-Delay Cell with Pre-charge and Pre-discharge Circuit [21]
3.4 Differential Dual-Delay Cell with Active Load and IMOS Varactor [22]
3.5 Dual-Delay Differential VCO with IMOS Varactor [23]
3.6 Dual-Delay Differential VCO with A-MOS Varactor [24]
4 Results and Discussion
5 Conclusion
References
Design and Analysis of Negative Capacitance Graded Channel Junctionless Nanowire for Analog/RF Applications
1 Introduction
2 Device Structure and Simulation Parameters
3 Analog RF Analysis of NCJLNW
3.1 Band Energy, Electric Field, and Potential Distribution
3.2 Analog/RF Analysis
4 Conclusion
References
Prospects of Black Phosphorous in Transit Time Devices
1 Introduction
2 Simulation Method
3 Results and Discussions
4 Conclusions
References
Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA
1 Introduction
2 Circuit Description
2.1 Voltage Differential Transconductance Amplifier (VDTA)
2.2 Circuit Description
2.3 VDTA as Three-Phase Oscillator
3 Results and Discussion
4 Conclusion
References
Double-Core Photonic Crystal Fiber for Liquid Sensing Detection
1 Introduction
2 Design of PCF
3 Conclusion
References
Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes
1 Introduction
1.1 Research Contribution and Motivation
2 System Model
2.1 Channel Model
2.2 CR Mobility
3 Performance Analysis of Cooperative Spectrum Sensing
3.1 Area Under the ROC Curve (AUC)
4 Results and Discussion
5 Conclusion
References
Sensitivity Enhancement of Kretschmann Configured Surface Plasmon Resonance Sensor with 2D Nanomaterial: MXene
1 Introduction
2 Theoretical Model, Sensor Design, and performance parameters
3 Results and analysis
4 Conclusion
References
Metasurface-Based Tunable Radar Absorbing Structure for Broadband Applications
1 Introduction
2 EM Performance Analysis of Active Unit Cell Based on PIN Diodes
2.1 Power Absorption Characteristics
2.2 Surface Current Distribution Over Unit Cells
2.3 Equivalent Circuit Model (ECM) for Unit Cells
3 Conclusion
References
Recommend Papers

VLSI, Communication and Signal Processing: Select Proceedings of the 5th International Conference, VCAS 2022
 9819909724, 9789819909728

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Lecture Notes in Electrical Engineering 1024

R. K. Nagaria V. S. Tripathi Carlos Ruiz Zamarreno Yogendra Kumar Prajapati   Editors

VLSI, Communication and Signal Processing Select Proceedings of the 5th International Conference, VCAS 2022

Lecture Notes in Electrical Engineering Volume 1024

Series Editors Leopoldo Angrisani, Department of Electrical and Information Technologies Engineering, University of Napoli Federico II, Napoli, Italy Marco Arteaga, Departamento de Control y Robótica, Universidad Nacional Autónoma de México, Coyoacán, Mexico Samarjit Chakraborty, Fakultät für Elektrotechnik und Informationstechnik, TU München, München, Germany Jiming Chen, Zhejiang University, Hangzhou, Zhejiang, China Shanben Chen, School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai, China Tan Kay Chen, Department of Electrical and Computer Engineering, National University of Singapore, Singapore, Singapore Rüdiger Dillmann, Gebäude 07.21, University of Karlsruhe (TH) IAIM, Karlsruhe, Baden-Württemberg, Germany Haibin Duan, Beijing University of Aeronautics and Astronautics, Beijing, China Gianluigi Ferrari, Dipartimento di Ingegneria dell’Informazione, Sede Scientifica Università degli Studi di Parma, Parma, Italy Manuel Ferre, Centre for Automation and Robotics CAR (UPM-CSIC), Universidad Politécnica de Madrid, Madrid, Spain Faryar Jabbari, Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA, USA Limin Jia, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China Janusz Kacprzyk, Intelligent Systems Laboratory, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Alaa Khamis, Department of Mechatronics Engineering, German University in Egypt El Tagamoa El Khames, New Cairo City, Egypt Torsten Kroeger, Intrinsic Innovation, Mountain View, CA, USA Yong Li, College of Electrical and Information Engineering, Hunan University, Changsha, Hunan, China Qilian Liang, Department of Electrical Engineering, University of Texas at Arlington, Arlington, TX, USA Ferran Martín, Departament d’Enginyeria Electrònica, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain Tan Cher Ming, College of Engineering, Nanyang Technological University, Singapore, Singapore Wolfgang Minker, Institute of Information Technology, University of Ulm, Ulm, Germany Pradeep Misra, Department of Electrical Engineering, Wright State University, Dayton, OH, USA Subhas Mukhopadhyay, School of Engineering, Macquarie University, NSW, Australia Cun-Zheng Ning, Department of Electrical Engineering, Arizona State University, Tempe, AZ, USA Toyoaki Nishida, Department of Intelligence Science and Technology, Kyoto University, Kyoto, Japan Luca Oneto, Department of Informatics, Bioengineering, Robotics and Systems Engineering, University of Genova, Genova, Genova, Italy Bijaya Ketan Panigrahi, Department of Electrical Engineering, Indian Institute of Technology Delhi, New Delhi, Delhi, India Federica Pascucci, Department di Ingegneria, Università degli Studi Roma Tre, Roma, Italy Yong Qin, State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing, China Gan Woon Seng, School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, Singapore Joachim Speidel, Institute of Telecommunications, University of Stuttgart, Stuttgart, Germany Germano Veiga, FEUP Campus, INESC Porto, Porto, Portugal Haitao Wu, Academy of Opto-electronics, Chinese Academy of Sciences, Haidian District Beijing, China Walter Zamboni, Department of Computer Engineering , Electrical Engineering and Applied Mathematics, DIEM—Università degli studi di Salerno, Fisciano, Salerno, Italy Junjie James Zhang, Charlotte, NC, USA

The book series Lecture Notes in Electrical Engineering (LNEE) publishes the latest developments in Electrical Engineering—quickly, informally and in high quality. While original research reported in proceedings and monographs has traditionally formed the core of LNEE, we also encourage authors to submit books devoted to supporting student education and professional training in the various fields and applications areas of electrical engineering. The series cover classical and emerging topics concerning: • • • • • • • • • • • •

Communication Engineering, Information Theory and Networks Electronics Engineering and Microelectronics Signal, Image and Speech Processing Wireless and Mobile Communication Circuits and Systems Energy Systems, Power Electronics and Electrical Machines Electro-optical Engineering Instrumentation Engineering Avionics Engineering Control Systems Internet-of-Things and Cybersecurity Biomedical Devices, MEMS and NEMS

For general information about this book series, comments or suggestions, please contact [email protected]. To submit a proposal or request further information, please contact the Publishing Editor in your country: China Jasmine Dou, Editor ([email protected]) India, Japan, Rest of Asia Swati Meherishi, Editorial Director ([email protected]) Southeast Asia, Australia, New Zealand Ramesh Nath Premnath, Editor ([email protected]) USA, Canada Michael Luby, Senior Editor ([email protected]) All other Countries Leontina Di Cecco, Senior Editor ([email protected]) ** This series is indexed by EI Compendex and Scopus databases. **

R. K. Nagaria · V. S. Tripathi · Carlos Ruiz Zamarreno · Yogendra Kumar Prajapati Editors

VLSI, Communication and Signal Processing Select Proceedings of the 5th International Conference, VCAS 2022

Editors R. K. Nagaria Department of Electronics and Communication Engineering Motilal Nehru National Institute of Technology Allahabad Prayagraj, Uttar Pradesh, India Carlos Ruiz Zamarreno Department of Electrical, Electronic and Communications Engineering Public University of Navarra (UPNA) Pamplona, Spain

V. S. Tripathi Department of Electronics and Communication Engineering Motilal Nehru National Institute of Technology Allahabad Prayagraj, Uttar Pradesh, India Yogendra Kumar Prajapati Department of Electronics and Communication Engineering Motilal Nehru National Institute of Technology Allahabad Prayagraj, Uttar Pradesh, India

ISSN 1876-1100 ISSN 1876-1119 (electronic) Lecture Notes in Electrical Engineering ISBN 978-981-99-0972-8 ISBN 978-981-99-0973-5 (eBook) https://doi.org/10.1007/978-981-99-0973-5 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Contents

Machine Learning-Based FSOC Link Performance Estimation . . . . . . . . Rohith Mankala and Yogendra Kumar Prajapati BER Efficiency of Outdoor Optics Links Using Hybrid-SIM with Pointing Errors Operating on Extreme Turbulence Regime . . . . . . . Dheeraj Dubey, Jahnvi Tiwari, Ajay Kumar Yadav, Yogendra Kumar Prajapati, and Rajeev Tripathi

1

13

An Exhaustive Review of Various Optical Devices for Biomedical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ayushman Ramola, Surinder Singh, and Anupma Marwaha

23

Cell Optimization and Realization of XOR-Based Logic Design in QCA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ayushi Kirti Singh, Subodh Wairya, and Divya Tripathi

39

Effect of Scandium Doping on Crystallization Kinetics and Glass Transition of Te(1−x) (GeSe0.5 ) Scx (X = 0.1) Glassy Alloy for PCM Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Surbhi Agarwal, Pooja Lohia, and D. K. Dwivedi

71

Design of Ultralow-Power and High-Speed Comparator Using Charge Sharing Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Kunal, Anurag Yadav, and Subodh Wairya

81

Underwater Image Enhancement Using Color Correction and Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T. M. Sachin and G. P. Prerana

97

Study and Analysis of Three-Stage Single-Miller CMOS OTA . . . . . . . . . 107 Dinesh kumar, Vikas Kumar, and P. Karuppanan Stacked Bi-LSTM Network and Dual Signal Transformation for Heart Sound Denoising . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Abhinav Chandel, Vinit Kumar, and Priya Ranjan Muduli v

vi

Contents

Abnormality Detection in Heart Using Combination of CNN, RNN and U-Net . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Utkarsh Sharma, Nimish Nigam, Ujjawal Kumar, Vinay Kumar, Sadanand Yadav, Ashish Pandey, and Rakesh Kumar Singh High Speed LVDS Driver Design with Fast Settling Common Mode Feedback Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 Munish Malik, Neelam R. Prakash, Ajay Kumar, and Jasbir Kaur Design and Analysis of Low Power Frequency Divider Circuit . . . . . . . . . 161 Priyanka Pandey, Amit Kumar, and Rajiv Kumar Singh Static Single Phase Contention Pulsed Latch for Low Voltage Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Pragati Mishra and Neelam Srivastava Design of Y-shaped Multiband Antenna Using the Parametric Approach for Wireless Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Leeladhar Bodu and Vinodh Kumar Minchula Design and Implementation of 1 × 4 & 4 × 4 RAM Using Quantum-Dot Cellular Automata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195 Shivani Kamaganikuntla, Sri Charan Cherepally, Sudeep Sharma, and S. V. S. Prasad FPGA and ASIC-Based Design of Fast and Low Power SEC-DAEC and SEC-DAEC-TAEC Codecs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Sayan Tripathi, Jhilam Jana, and Jaydeb Bhaumik Reliable Implementation of Finite State Machine for On-Board VLSI Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Sourabh Kumar Jain, Usha Mehta, Aarushi Bhandari, Kamal Poddar, S. M. Trivedi, and A. K. Lal Performance Improvement of SAC-OCDMA FSO System Under Rain and Snow Conditions Using Different Zero Cross-Correlation Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Sokaina Boukricha, Kamal Ghoumid, Abdelfattah Mazari, El miloud Ar-Reyouchi, Réda Yahiaoui, and Omar Elmazria Robust and Imperceptible Color Image Watermarking Using LWT, Schur Decomposition, and SVD in YCbCr Color Space . . . . . . . . . 259 Divyanshu Awasthi and Vinay Kumar Srivastava Design of Three Slot-Based Multiband MIMO Circular Patch Antenna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273 Anshika Shrivastav, Pradeep Kamal, Yash Deshmukh, Patri Upender, B. Yakub, and Amarjit Kumar

Contents

vii

An Imperceptible Semi-blind Color Image Watermarking Using RDWT and SVD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Ranjana Dwivedi and Vinay Kumar Srivastava Construction of Deterministic Measurement Matrices Using Polyphase Sequences for Compressed Sensing Radar . . . . . . . . . . . . . . . . . 295 Sudha Hanumanthu and P. Rajesh Kumar A Non-invasive Planar Resonant Microwave Sensor for Unknown Liquid Permittivity Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Ojaswita Mankar and Smriti Agarwal Design of SRAM Cell for Improved Performance . . . . . . . . . . . . . . . . . . . . . 317 Priyanshi Bhatia and Santosh Kumar Gupta Performance Analysis of Various Charge Pump Topologies for PLL Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Nashra Khalid and Ram Chandra Singh Chauhan Low Power Radix-4 Booth Multiplier Design Using Pass Transistor Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 Sandhya Kanoujia, Rishav Kumar, and P. Karuppanan Impact of Different Buffer Layers on Performance of Cu2 O Based Solar Cell: SCAPS 1D Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365 Chandni Tiwari and Varun Mishra Sum Rate Maximization Using Unlicensed and Licensed Bands in UAV-Assisted Cellular Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Nageswara Rao Kota and Kalpana Naidu Performance Analysis of mmWave V2V Communication Using Relay Vehicle for Advanced Safety Applications . . . . . . . . . . . . . . . . . . . . . . 389 Aakash Jasper, Arun Prakash, Sara Paiva, and Raghavendra Pal Low-Voltage Low-Power Fully Differential Double Recycling OTA with Positive Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Nikhil Deo and Tripurari Sharan Multilingual Indian Musical Type Classification . . . . . . . . . . . . . . . . . . . . . . 419 Swati P. Aswale, Prabhat Chandra Shrivastava, Roshani Bhagat, Vikrant B. Joshi, and Seema M. Shende Design and Implementation of Matched Filtering and Timing Recovery Algorithm for IEEE 802.15.4 Digital Baseband Front End . . . . 431 Preeti Shukla, Manish Tiwari, Mohanasundaram, and P. Haribabu Minimal Group Delay Multi-objective Finite Impulse Response Filter Design Using Salp Swarm Algorithm and Its Improved Version . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Sonelal Prajapati and Sanjeev Rai

viii

Contents

Performance Analysis of Downlink Cooperative NOMA System . . . . . . . 455 M. Ramadevi, S. Anuradha, and L. Padma Sree Design and Analysis of Noise Immune, Energy Efficient 1-bit 8T SRAM Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465 Hena Shivhare, Avaneesh Kumar Dubey, and Sumit Kumar Jha Automated Design Rule Checker for VLSI Circuits Using Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475 Mihir Rana, Nimit Malani, Ruchi Gajjar, Manish I. Patel, and Dipesh Panchal Design of Low Power High Speed 2nd Order Discrete Time Sigma-Delta Modulator Using Charge Shared Double Tail Dynamic Comparator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Virat Mathur, Vikas Tiwari, and R. K. Nagaria Solar Powered Low Cost Air Quality Monitoring System . . . . . . . . . . . . . . 501 Amit Kumar Singh Chauhan, Anant Singh Suryavanshi, Anoushka Awasthi, Shivani Rai, Shruti Mishra, and Arun Kumar Singh High Gain Hexaport Millimetre Wave MIMO Antenna for 5G Service LMDS and N257 Band Applications . . . . . . . . . . . . . . . . . . . . . . . . . 515 Ch Murali Krishna, N. Suguna, N. V. K. Maha Lakshmi, R. Saravanakumar, Shiv Charan Puri, and M. K. V. Subbareddy Aspect Ratio Estimation of a Two-Stage Operational Amplifier . . . . . . . . 531 Aravind Kannan, Aftaab Siddiqui, Ruchi Gajjar, Manish I. Patel, and Dipesh Panchal EM Power Absorption and RCS Analysis of Novel FSS-Based Broadband Radar Absorbing Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 Trideeb Bhattacharya, Shrikrishan Baghel, Syed Tabassum Nazeer, Vineetha Joy, and Hema Singh Performance Analysis of Pulsed Latches for Low-Voltage Operation . . . 557 Pragati Mishra and Neelam Shrivasrva A Comparative Performance Analysis of Different Methodology Operational Transconductance Amplifier Using Cadence . . . . . . . . . . . . . . 571 Km Jyotsana and Amit Kumar Checkpoint Snapshot Placement in the Cloud Data Center Using Fuzzy Inference System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 Priti Kumari, Vandana Dubey, Adarsh Kumar, and G. R. Mishra 4-Disjoint Path Multistage Interconnection Network . . . . . . . . . . . . . . . . . . 597 Vipin Sharma, Abdul q. Ansari, and Rajesh Mishra

Contents

ix

Design of Low-Power High Performance SAR Analog-To-Digital Converter Based on MOS Capacitor DAC Structure . . . . . . . . . . . . . . . . . . 609 Kumar Shubham, Vikas Tiwari, and R. K. Nagaria FGMOS and QFGMOS-Based Super-Wilson Current Mirror and Its Application in Full-Wave Rectifier . . . . . . . . . . . . . . . . . . . . . . . . . . . 623 Renuka Bhardwaj, Ashwni Kumar, and Richa Srivastava Review of the SAR ADC’s Energy Efficient Switching Scheme . . . . . . . . . 637 Vikas Tiwari and R. K. Nagaria Design of an Elliptical-Shaped Meta-material for SAR Reduction Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Abhay Krishna Yadav, Sudhanshu Verma, Abhimanyu Yadav, and Anand Sharma Improved Fuzzy Logic-Based Localization in Wireless Sensor Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659 Rishabh Tarachand Tarte, Shilpi, and Arvind Kumar Residential Level Short-Term Demand Forecasting Using ANFIS Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 671 Chandrashekhar Azad, Parvez Ahmad, Priyanshul Niranjan, Deepak Kumar Singh, Niraj Kumar Choudhary, and Nitin Singh Signal Integrity Analysis in Coupled GNR Nano Interconnects Integrated with Buffer Repeaters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683 Afreen Khursheed, Kavita Khare, and Zainab Aizaz Effect of Electrical Vehicle Charging on Power Quality . . . . . . . . . . . . . . . 695 Parvez Ahmad, Kanahaiya Kumar, Tarun Varshney, Sandeep Gupta, Ruchi Varshney, Niraj Kumar Choudhary, and Nitin Singh Design and Analysis of Ion Selective Field Effect Transistor for Biomedical Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709 Arathi V. Suresh, Merin Thomas, and Lucky Agarwal Narrowband Data Waveform Development and Simulation for Achieving Low BER in Wireless Communication . . . . . . . . . . . . . . . . . . 719 Gopal Agarwal, Sushil Kumar Bahuguna, Nawaj Shikalgar, S. K. Lahiri, and Anuradha Mayya Design and Analysis of CIC Decimation Filter Using Redundant Number System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 Ajeet Kumar Srivastava, Krishna Raj, and Alok Kumar Enhanced Energy-Aware Fault Tolerance Technique for Real-Time Task on Heterogeneous Multicore System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 745 Priyanka Gupta and Ranvijay

x

Contents

Design and DC Electrical Performance Analysis of SOI-Based SiO2 /HfO2 Dual Dielectric Gate-All-Around Vertically Stacked Nanosheet at 5 nm Node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763 Ram Krishna Dewangan, Vinay Kumar Singh, and Mohammad Rafique Khan Designing of Ternary to Binary Half Adder Using CMOS . . . . . . . . . . . . . 773 Rajan Singh, Bittu Kumar, Kiran Dasari, S. V. S. Prasad, Kota Maneela, and Bhagavathi Gadi Performance Analysis of Differential Dual Stage Delay Cells of VCO . . . 785 Mohd Saqib, Subodh Wairya, and Anurag Yadav Design and Analysis of Negative Capacitance Graded Channel Junctionless Nanowire for Analog/RF Applications . . . . . . . . . . . . . . . . . . . 801 Manish Kumar Rai, Shubham Verma, and Sanjeev Rai Prospects of Black Phosphorous in Transit Time Devices . . . . . . . . . . . . . . 813 Girish Chandra Ghivela and Joydeep Sengupta Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 819 Abhishek Kumar, Santosh Kumar Gupta, and Vijaya Bhadauria Double-Core Photonic Crystal Fiber for Liquid Sensing Detection . . . . . 837 Shreya Gupta, Dharmendra Kumar, Vijay Shanker Chaudhary, and Sneha Sharma Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 847 Ashish K. Rao, Siddharth Srivastava, Arun Kumar Singh, and Neelam Srivastava Sensitivity Enhancement of Kretschmann Configured Surface Plasmon Resonance Sensor with 2D Nanomaterial: MXene . . . . . . . . . . . . 861 Rajeev Kumar, Maneesh Kumar Singh, Sarika Pal, and Alka Verma Metasurface-Based Tunable Radar Absorbing Structure for Broadband Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873 Hrishit Mohan Das, Syed Tabassum Nazeer, Shrikrishan Baghel, Vineetha Joy, and Hema Singh

About the Editors

R. K. Nagaria is working as Professor and Head in Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology (MNNIT) Allahabad, Prayagraj (India). He received B. Tech. and M.Tech. degrees in Electronics Engineering from Kamla Nehru Institute of Technology (KNIT) Sultanpur, India, and Ph.D. (Engg.) from Jadavpur University, Kolkata, India. He has over 33 years of teaching UG, PG, and research experience. He has published over a hundred research papers in national and international conferences/journals. His name is enlisted in an exclusive directory Marquis Who’s Who in the world. He is also nominated for International Educator of the year award 2005, by International Biographical Centre, Cambridge, England. He has guided the thesis of more than 30 PG students and supervised 12 Ph.D. theses; presently, four research scholars are working under his supervision. His area of interest is Mixed-Mode/Analog Signal Processing, High-Speed Networks, and VLSI Design and Applications. V. S. Tripathi received his B.Tech. from the J. K. Institute of Applied Physics and Technology, the University of Allahabad, in 1988. He did his M.Tech. and Ph.D. from Motilal Nehru National Institute of Technology Allahabad in 1999 and 2007, respectively. Presently, he is a Professor in the Department of Electronics and Communication Engineering at MNNIT Allahabad. He has published more than 70 research papers in reputed refereed international journals and conferences. He has guided seven Ph.D. scholars and more than 15 M.Tech. scholars. His area of specialization includes Antenna, SDR, and Non-invasive RF Sensors. Carlos Ruiz Zamarreno is a Professor in the Electrical, Electronic, and Communications Engineering Department at the Public University of Navarra (UPNA), Pamplona-Iruña, Spain. Professor Zamarreño received his M.S. degree in Electrical and Electronic Engineering and his Ph.D. in Communications from the Public University of Navarra (UPNA) in 2005 and 2009, respectively. He obtained a permanent position as Associate Professor at the Electrical, Electronic, and Communications Engineering Department at UPNA. In 2008, he was Visiting Scientist at the Massachusetts Institute of Technology (Boston, MA, USA) under the guidance of xi

xii

About the Editors

Prof. Henry I. Smith. He also performed a post-doctoral research stay at Siemens AG (Munich, Germany) in the group of Prof. Maximilian Fleischer. He has also obtained research grants from the Bank of Santander and the Spanish Ministry of Economy and Competitiveness (José Castillejo) in 2013 and 2015, respectively, to perform a research stay at the Photonics Laboratories at Universidade Technologica do Parana (Parana, Brazil) in the group of Prof. Cicero Martelli. In 2013, he received the IEEE GOLD Award for his contributions to developing novel optical sensing waveguides based on micro- and nanostructured films. He is the author of more than 200 scientific journals and conference publications. He has participated in 20 different research projects with public and private entities. Yogendra Kumar Prajapati received his Ph.D. degree from the Electronics Engineering Department, Uttar Pradesh Technical University, Lucknow, Uttar Pradesh, India, in 2010. He is currently an Associate Professor at Motilal Nehru National Institute of Technology Allahabad, Prayagraj, Uttar Pradesh, India. Dr. Yogendra has made a significant contribution to experimental and theoretical research in the areas of Photonics. He has completed several R&D projects sponsored by government funding agencies such as DST, BRNS, etc. He has published more than 120 SCI research papers in peer-reviewed journals and 40 International Conference proceedings. He is a recipient of the Sir Visvesvaraya Young Faculty Research Fellowship Award 2016, Ministry of Electronics and Information Technology, New Delhi, India, and a teaching fellowship from Dr. A. P. J. Abdul Kalam Technical University, Lucknow in 2007. Also, Dr. Yogendra is a recipient of the Young Scientist Award by DST, SERB New Delhi in 2014. His research interests include Optical Communication, Optical Fiber Sensors, Photonic Spin Hall Effect, and Fabrication and characterization of SPR sensors. He is an active reviewer of many scientific sensors-related journals, and Associate Editor of the IEEE Access and Frontiers in Physics—Optics and Photonics.

Machine Learning-Based FSOC Link Performance Estimation Rohith Mankala and Yogendra Kumar Prajapati

Abstract Free space optical communication (FSOC) is the optical wireless technology that provides enormous bandwidth, unlicensed spectrum, and excellent performance in comparison with radio frequency (RF) technology, which has its own impairments that degrade signal quality in the environment. Numerous innovative techniques, including as channel coding, diversity approaches, adaptive optics, higher-order spatial modes, and machine learning, have been included into the FSO. In the proposed study, a performance analysis of the FSOC system is conducted, and five machine learning (ML) techniques are used to determine the entire system’s quality factor. Coefficient of determination and mean absolute error (MAE) are used to evaluate the performance of these models. The collected findings demonstrate that the random forest ML model predicts the quality factor with the lowest MAE and the highest coefficient of determination compared to the other considered ML models. Keywords Free space optical communication · Machine learning · Random forest

1 Introduction Globally, radio frequencies are used extensively in wireless technologies. As technology evolves, so does the need for high capacity and performance. Due to spectrum band restrictions and spectrum use regulations, the RF spectrum gets crowded as a consequence of the rising number of users. Optical wireless communication (OWC) is a complementing technology to radio frequency (RF) technology that provides fast data transfer speeds and a broad variety of unregulated spectrum bands. OWCs are wireless connections that use the optical spectrum. Free space optics (FSO) is an optical wireless communication technology where the medium of transmission is free space with data, voice, and video communication that can be fulfilled by tens of gigabits per second data rate over a vast range of communication that ranges from meters to kilometers with the high-speed line of sight technology [1, 2]. It delivers huge R. Mankala · Y. K. Prajapati (B) ECED, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, Uttar Pradesh, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_1

1

2

R. Mankala and Y. K. Prajapati

capacity and is simpler to implement than fiber-based solutions. This technology’s primary benefits include unlicensed spectrum, extremely secure communications, and minimal implementation costs. Despite its numerous benefits, the FSO connection suffers from signal deterioration owing to channel imperfections that hinder system performance. In contrast to fiber-based and wireless systems, FSO systems have channel impairments of their own. Beam divergence, atmospheric attenuation, air turbulence, and pointing errors rely on several characteristics such as weather conditions, temperature, atmospheric pressure, and transmitter and receiver alignment. A few strategies such as machine learning and neural network may be applied in FSO to improve system performance, but they are still experimental, in the research stage, and well defined; nonetheless, they are situation dependent due to their complexity and expense [2]. The current trend in the fiber optic system is machine learning, which may also be utilized in FSO. ML algorithms have been implemented in numerous domains of fiber-based systems, including performance monitoring, modulation format detection, and signal quality reporting. ML algorithms have been successfully implemented in various optical wireless communication applications, such as the detection of users transmitting simultaneously with different wavelengths by using their histogram of the amplitude [3], the improvement of the performance of the turbulent system by the relevant neural network [4], and the estimation of the channel under the influence of different turbulence [5]. Literature depicts the prediction of the quality factor of the FSO system for numerous transceivers [1, 6]. Depending on the problem statement, supervised (classification or regression) or unsupervised (clustering) machine learning techniques could be implemented. To the best of knowledge, FSO with the array of amplifiers for different power range and in the presence of adverse weather conditions has been simulated and estimated using ML algorithms with the huge set of data for the better prediction in this work. This work analyzes the performance of the FSO system with an array of amplifiers and describes the quality factor (Q-factor) prediction using ML algorithms. Here, the accuracy of the regression algorithms’ predictions is assessed by the R2 coefficient (the coefficient of determination) and the MAE. In this study, many supervised ML techniques are used. For predicting the Q-factor of the proposed FSO system, support vector machine (SVR), decision tree (DT), linear regression (LR), random forest (RF) regression, and K closest neighbors (KNN) regression models have been trained and evaluated. This paper is organized into the following sections: In Sect. 2, the description of the FSO model, the generations of the datasets for the ML model training, along with results and discussion are given. In Sect. 3, the conclusion of the work is given.

Machine Learning-Based FSOC Link Performance Estimation

3

2 Results and Discussions 2.1 FSO System 2.1.1

System Description

In this section, the FSO simulation setup and the performance analysis of the system with an array of amplifiers are presented for the different transmitting wavelengths and the different ranges of the link coverage. The block diagram of the FSO system is shown in Fig. 1, and the simulation parameters used are described in Table 1. In this work, the Opti system (optical communication system design software) is used to simulate the proposed FSO system. The whole system is organized around the basic communication modules, which are the transmitter, channel, and receiver. The description of each module is as follows: The transmitter system block includes the pseudorandom bit sequence generator, NRZ pulse generator, CW laser, Mach– Zehnder modulator, and optical amplifiers. Random bit sequences, which are periodic and deterministic series of digital zeros and ones, are generated by the PRBS generator. The NRZ pulse generator turns the produced binary sequence into the NRZ line code electrical pulse. Continuous wave (CW) lasers generate a continuous and unbroken beam across time. For performance evaluation, the operating wavelength was set to 850, 1310, and 1550 nm. The FSO channel is the traveling medium for optical waves between the transmitter and receiver in free space. Due to varying weather conditions and turbulence effects, the broadcast signal degrades in the channel. As per the variations of weather effects, the attenuation of the signal being received is obtained from Beer–Lambert’s law by evaluating the aerosol and the molecular

PRBS Generator

NRZ Generator

CW Laser

MZM

BER Analyzer

Filter

FSO Channel

PIN Photodiode

Array of Amplifiers

Fig. 1 Block diagram of FSO simulation setup

Table 1 Design parameters specification of FSO system

Parameter

Value

Attenuation

Clear air (0.245 dB/Km)

Link range

1000 m to 5000 m

Wavelengths

850 nm, 1310 nm, 1550 nm

4

R. Mankala and Y. K. Prajapati 010101. PRBS Generator

NRZ Pulse Generator

Array of Amplifiers=N

MZM

FSO Channel CW Laser BER Analyzer

3R Generator LP Bessel Filter

PIN Photo detector

Fig. 2 Simulation setup of the FSO system

particles in the atmosphere. The law depends on the absorption and scattering coefficients of the attenuation parameter. For fog and haze conditions, attenuation is given by Kim’s model depending upon the visibility and transmitting wavelength that define the signal degradation occurring in dB/km. The fluctuations that are due to turbulence are classified as weak, moderate, and strong and are measured in terms of S.I (scintillation index). As per these variations in turbulence, there are channel models that describe the intensity fluctuations. The receiver block consists of the photo detector, low-pass Bessel filter, 3R generator, and BER analyzer. A photo detector (PIN diode) converts the optical signal to an electrical signal. The low-pass Bessel filter has a smooth roll-off factor with a flat gain that removes the unwanted highfrequency components. The 3R generator generates the received input bit sequence and an electrical NRZ pulse. The BER analyzer evaluates the BER, eye pattern, and Q-factor by using the received signal from the 3R generator. The simulation setup of the model with the array of the amplifiers is shown in Fig. 2. The Q-factor can be calculated from BER as follows [7, 8]. Pe =

) ( Q 1 erfc √ , 2 2

(1)

where Q-factor can be calculated from eye diagram s follows Q=

|μ1 − μ0 | , σ1 − σ0

(2)

where μ1,0 are the mean values of the voltages and σ1,0 are the standard deviation.

Machine Learning-Based FSOC Link Performance Estimation

2.1.2

5

Discussions

After the simulation of the FSO system, the results have been observed at the wavelengths of 850 nm, 1310 nm, and 1550 nm, with the link distance ranging from 100 to 5000 m. The obtained output parameters are the Q-factor and received optical power of the FSO system, which are described in this section. The Q-factor is an important parameter that determines the performance or quality of the signal being received. The system is simulated using Opti system software. The design parameters of the system are mentioned in Table 1. Figure 3 illustrates the received signal power over the link range of 100–5000 m. Observed results indicate that as the number of amplifiers (N) increases, the signal power received also increases. Figures 4, 5, and 6 show the variation of the Q-factor against a link length of 100–5000 m for the different wavelengths of 850 nm, 1310 nm, and 1550 nm under clear weather conditions. It has been seen that as the number of amplifiers goes up, the quality of the signal that is received goes up. Figures 7, 8, and 9 illustrate the eye pattern of the FSO design at 1550 nm wavelength with a link range of 5000 m. It is observed that the eye-opening of the signal increases as the number of amplifiers increases, and it is also observed that the eyeopening of the system with four amplifiers is greater compared to the FSO system with one or two amplifiers, as shown in Figs. 7, 8, and 9, respectively. The obtained parameters such as Q-factor, maximum eye-opening, and minimum BER achieved and received signal power for the N number of amplifiers at the link range of 5000 m under clear weather conditions are given in Table 2. 20

Recieved Signal power

Fig. 3 Received power versus against link range at 1550 nm

N=1 N=2 N=4

0 -20 -40 -60 -80

0

1000

2000

3000

Range (m)

4000

5000

6

R. Mankala and Y. K. Prajapati

Fig. 4 Q-factor against link range at 850 nm wavelength for different no. of amplifiers

3500

N=4 N=2 N=1

3000

Q- Factor

2500 2000 1500 1000 500 0 0

1000

2000

3000

4000

5000

Range (m) Fig. 5 Q-factor against link range at 1310 nm wavelength for different no. of amplifiers

3500

N=4 N=2 N=1

3000

Q Factor

2500 2000 1500 1000 500 0 0

1000

2000

3000

4000

5000

Range (m) Fig. 6 Q-factor versus link range at the 1550 nm wavelength for different no. of amplifiers

1800 1600

N=4 N=2 N=1

Q Factor

1400 1200 1000 800 600 400 200 0 0

1000

2000

3000

Range (m)

4000

5000

Machine Learning-Based FSOC Link Performance Estimation

7

1

0

0.5

1

10μ 0

Q

0.5

2.9 3.1 3.3 3.5 3.7 3.9 4.1

0

Amplitude (a.u.)

Time (bit Period)

Fig. 7 BER analyzer output of FSO system at 1550 nm wavelength for single amplifiers

Time (bit Period) Time (bit Period) 0

0.5

1

0

0.5

1

0

3 2

Amplitude (a.u.)

10μ

5 4

Q

6

7

20μ

8

Fig. 8 BER analyzer output of FSO system at 1550 nm wavelength for two amplifiers

Time (bit Period)

1

0.5

1

20μ 30μ 40μ

0.5

0

10μ

10

Q

0

0

Time (bit Period)

Amplitude (a.u.)

Time (bit Period)

Fig. 9 BER analyzer output of FSO system at 1550 nm wavelength for four amplifiers

8

R. Mankala and Y. K. Prajapati

Table 2 Simulated output parameters of FSO system at 1550 nm wavelength for a link range of 5000 m Parameter

Optical amplifiers N =1

N =2

Q-factor

4.18298

8.76246

Eye height

2.93947e−006

1.37836e−005

BER

1.43447e−005 −73.4132

Received signal power (dBm)

9.53044e−019 −67.3487

N =4 15.9058 3.39446e−005 2.88208e−057 −61.3227

2.2 ML Model Evaluations 2.2.1

Dataset Generation

ML regression algorithms have been used in this section for estimating the Q-factor of the proposed FSO system, which is implemented in Sect. 2.1. Different ML model performance evaluations are carried out by comparing their coefficient of determination and the mean absolute error. The evaluation process of the ML models is shown in Fig. 10. The ML algorithms used are support vector regression (SVR), linear regression (LR), decision tree (DT), random forest (RF algorithm), and K-nearest neighbors regression models. The dataset generation is done by the Opti system software by the parameter sweep. Around 10,725 datasets are generated for the input parameters. The input parameters considered are link range, power transmitted, wavelength, beam divergence, and the atmospheric attenuations for different weather conditions and the output parameter of Q-factor. These parameters are given in Table 3. After nested parameter sweeping, the simulations are performed. Then the datasets of 10,725 × 6 are exported and saved in a file. This generated data of the Q-factor is used for training the ML model. The whole dataset is divided into two sets, which are training and testing datasets. The training datasets are used for the model fitting and training. After the model fitting, the test datasets have been tested for their performance by

Model Comparison

Data Generation

Training and Test Data Separation

Training Data (80%)

Test Data (20%)

Fig. 10 ML models flow of evaluation

Model Training and Building

Model Testing

Machine Learning-Based FSOC Link Performance Estimation

9

Table 3 Simulation parameters for the dataset generation Parameter

Value

Data rate

10 Gbps

Optical transmitted power

5–25 mW

Transmission length

100–5000 m

Transmitter and receiver aperture size

5 cm and 8 cm

Beam divergence

2–5 mrad

Wavelength

850 nm, 1310 nm, and 1550 nm

Weather conditions

Based on Kim’s model for haze, fog attenuations, and approximate rain expression [7]

the R2 coefficient and the mean absolute error. Then overall comparisons of the ML models are done with the evaluated performance values. The attenuation coefficients of different weather conditions are as follows [7]. The weather attenuations for the environmental conditions are given by the Eqs. (3)–(5). For fog attenuation, ( ) 3.91 λ −q βfog (λ) = , V 550

(3)

where V is visibility range in kilometers; λ is the wavelength in nm; q is the size distribution coefficient of scattering. ⎧ ⎪ 1.6 ⎪ ⎪ ⎪ ⎪ ⎨ 1.3 q = 0.16V + 0.34 ⎪ ⎪ ⎪ V − 0.5 ⎪ ⎪ ⎩0

V > 50 6 < V < 50 1Drain

C Gate>Source

1.2x10-15

-16

6.0x10-16 4.0x10-16

6.0x10-16

4.0x10-16

2.0x10-16

2.0x10-16

0.0

0.0 0.0

0.2

0.4

0.6

Gate Voltage (V)

0.8

1.0

0.0

0.2

0.4

0.6

Gate Voltage (V)

Fig. 6 C gs versus V gs and C gd versus V gs in JLNW, NCJLNW, and GCNCJLNW

808

M. K. Rai et al. 0.05

Transconductance (S)

8.0x10-3

Output Conductance (S)

JLNW NCJLNW GCNCJLNW

1.0x10-2

6.0x10-3 4.0x10-3 2.0x10-3

0.04

0.03

0.02

0.01

0.00

0.0 0.0

JLNW NCJLNW GCNCJLNW

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0

0.2

Gate Voltage (V)

0.4

0.6

0.8

1.0

Gate Voltage (V)

Fig. 7 gm versus V gs and gd versus V gs in JLNW, NCJLNW, and GCNCJLNW

At a fixed gate to source voltage, the ratio of drain current fluctuation to the equivalent variation in the drain to source voltage is known as output conductance. Mathematically gd =

∂ Id . ∂ Vds

(1)

Figure 7 depicts the output conductance variation as a function of Vgs . In the linear region, increasing the drain to source voltage decreases output conductance. gd = µCox

 W Vgs − VT − Vds , L

(2)

gd,sat = 0.

(3)

and in the saturation region:

From Fig. 7, it is evident that output conductance (gd ) for NCJLNW is slightly higher than JLNW for smaller values of V ds (V ds < 0.3 V) and for V ds > 0.3 V, gd value for GCNCJLNW becomes equal to gd of JLNW. Figure 8 shows the early voltage (V EA = I d /gd ) variation, which shows that it is higher by 26.8% for NCJLNW, showing better analog performance. The intrinsic gain of a device is the ratio of its transconductance (gm ) to its output conductance (gd ). It reflects the device’s voltage gain, and a greater intrinsic gain value suggests better performance [29, 30]. The fluctuation in intrinsic gain with regard to V gs for JLNW and NCJLNW is shown in Fig. 8. The intrinsic gain of NCJLNW is higher than that of JLNW, as seen in Fig. 8. Cutoff frequency (f T ) is the frequency at which the device functions as an amplifier, i.e., the drain current equals the gate current at this frequency, and small-signal current gain equals unity

Design and Analysis of Negative Capacitance Graded Channel …

809

8 3.5x1012

JLNW NCJLNW GCNCJLNW

3.0x1012

7

5 Early Voltage (V)

Cutoff Freq (Hz)

2.5x1012 2.0x1012 1.5x1012 1.0x1012 5.0x1011

4 3 2 1

0.0 -5.0x1011 0.0

JLNW NCJLNW GCNCJLNW

6

0 0.2

0.4

0.6

0.8

-1 0.0

1.0

Gate Voltage (V)

0.2

0.4

0.6

0.8

1.0

Gate Voltage (V)

Fig. 8 f T versus V gs and V EA versus V gs in JLNW, NCJLNW, and GCNCJLNW

fT =

gm . 2π(Cgs + Cgd )

(4)

The maximum operating frequency of the circuit is limited by f T . The variation in cutoff frequency (f T ) with regard to V gs is depicted in Fig. 8. NCJLNW has a higher cutoff frequency than JLNW at lower V gs levels. NCJLNW, on the other hand, has a lower cutoff frequency than JLNW at larger V gs levels. For high-frequency applications, the transconductance frequency product (TFP) is an essential metric. It is described as (TFP) =

gm ∗ f T . Id

(5)

Here, transconductance is gm , the cutoff frequency is fT , and the drain current is I d . Figure 9 shows TFP as a function of V gs . NCJLNW has a greater TFP than JLNW for a lower gate voltage value; however, when V gs increases, TFP value declines. Transconductance generation factor (TGF) is the ratio of transconductance (gm ) to output current (I d ) (TGF) =

gm . Id

(6)

The circuit consumes less power and may work efficiently at low supply voltage if the TGF is high. gm /Id is highest around the device’s subthreshold region, as seen in Fig. 9. NCJLNW has a larger TGF than JLNW at lower gate voltages, indicating a high gain per unit of power dissipation. Both devices have remarkably similar TGF when Vgs is increased. Table 3 shows the peak values of different analog parameters. Transconductance is increased by 30.9%, intrinsic gain is increased by 1.6%, cutoff frequency is increased by 26.6%, TGF is increased by 15.3%, TFP is increased by 21.8%, and early voltage is

810

M. K. Rai et al. 7x1013

40 6x1013

35 30

4x1013

TFP (Hz/V)

25 TGF (V-1)

JLNW NCJLNW GCNCJLNW

5x1013

JLNW NCJLNW GCNCJLNW

20 15 10

3x1013 2x1013 1x1013

5

0

0

13

-1x10

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

Gate Voltage (V)

0.4 0.6 Gate Voltage (V)

0.8

1.0

Fig. 9 TGF versus V gs and TFP versus V gs in JLNW, NCJLNW, and GCNCJLNW

Table 3 Analog/RF parameter values S. No. Parameter

JLNW

NCJLNW

GCNCJLNW

1

4.2 × 10−4

9.5 × 10–3

9.8 × 10−3

26.16

44.02

Transconductance (S)

2

Output conductance (mS)

4.02

3

Cutoff frequency (Hz)

3.03 × 1011 3.51 × 1012 3.01 × 1012

4

Transconductance generation factor (V−1 )

36.2

5

Transconductance frequency product (Hz/V) 4.2 ×

6

Early voltage (V)

6.9

38.5 1012

5.13 × 6.1

39.2 1013

6.32 × 1013 1.1

increased by 49.2% using the negative capacitance technique. The improved performance metrics of the NCJLNW device show that it is a good choice for analog applications.

4 Conclusion The thorough study of NCJLNW for low power high density ICs is presented in this paper. First, the impact of the negative capacitance approach on the performance of the JLNW was investigated. In comparison with previous devices at this node, it has been discovered that the proposed device operates at low power and successfully accommodates short-channel effects. To further improve the device performance, graded channel approach is incorporated in the NCJLNW device. The GCNCJLNW structure has a step-like potential profile and minima near to the channel’s source side, resulting in a lower DIBL. Furthermore, below the insulator gap, the channel electron concentration is found to be uniform, resulting in a higher conduction current density. Furthermore, the suggested device has an off-state leakage current of 1.34 × 10–16 A, which is sufficient to off-flow the leakage during a circuit’s switch-off

Design and Analysis of Negative Capacitance Graded Channel …

811

condition, and a drive current of 0.62 mA at a silicon thickness of 5 nm for 20 nm channel length.

References 1. McFardland, Flynn M (1995) Limits of scaling MOSFETs. Technical report CSL-TR-95-662 2. D’Agostino F, Quercia D (2000) Short channel effects in MOSFETs 3. Romli NB et al (2015) An overview of power dissipation and control techniques in CMOS technology. J Eng Sci Technol Malaysia 10(3):365–382 4. Rai MK, Gupta A, Rai S (2021) Comparative analysis & study of various leakage reduction techniques for short channel devices in junctionless transistors: a review and perspective. Silicon 1–23 5. Samal A, Tripathi SL, Mohapatra SK (2020) A journey from bulk MOSFET to 3 nm and beyond. Trans Electr Electron Mater 1–13 6. Mendiratta N, Tripathi SL (2021) 18 nm n-channel and p-channel Doping less asymmetrical Junctionless DG-MOSFET: low power CMOS based digital and memory applications. Silicon 1–12. 7. Colinge J-P, Lee C-W, Afzalian A, Akhavan ND, Yan R, Ferain I, Razavi P et al (2010) Nanowire transistors without junctions. Nat Nanotechnol 5(3):225–229 8. Colinge JP, Lee CW, Dehdashti Akhavan N, Yan R, Ferain I, Razavi P, Kranti A, Yu R (2011) Junctionless transistors: physics and properties. In: Semiconductor-on-insulator materials for nanoelectronics applications. Springer, Berlin, Heidelberg, pp 187–200 9. Salahuddin S, Datta S (2008) Use of negative capacitance to provide voltage amplification for low power nanoscale devices. Nano Lett 8(2):405–410 10. Böscke TS, Müller J, Bräuhaus D, Schröder U, Böttger U (2011) Ferroelectricity in hafnium oxide thin films. Appl Phys Lett 99(10):102903 11. Kittl JA, Locquet, J-P, Houssa M, Afanasiev VV (2020) A critical analysis of models and experimental evidence of negative capacitance stabilization in a ferroelectric by capacitance matching to an adjacent dielectric layer. arXiv:2003.00424 12. Hoffmann M, Peši´c M, Slesazeck S, Schroeder U, Mikolajick T (2018) On the stabilization of ferroelectric negative capacitance in nanoscale devices. Nanoscale 10(23):10891–10899 13. Mehta H, Kaur H (2018) Impact of Gaussian doping profile and negative capacitance effect on double-gate junctionless transistors (DGJLTs). IEEE Trans Electron Devices 65(7):2699–2706 14. Lin CI, Khan AI, Salahuddin S, Hu C (2016) Effects of the variation of ferroelectric properties on negative capacitance FET characteristics. IEEE Trans Electron Devices 63(5):2197–2199 15. Choi SJ, Moon DI, Kim S, Ahn JH, Lee JS, Kim JY, Choi YK (2011) Nonvolatile memory by all-around-gate junctionless transistor composed of silicon nanowire on a bulk substrate. IEEE Electron Device Lett 32(5):602–604 16. Liu TY, Pan FM, Sheu JT (2015) Characteristics of gate-all-around junctionless polysilicon nanowire transistors with twin 20-nm gates. IEEE J Electron Devices Soc 3(5):405–409 17. Cheng CH, Fan CC, Tu CY, Hsu HH, Chang CY (2018) Implementation of dopant-free hafnium oxide negative capacitance field-effect transistor. IEEE Trans Electron Devices 66(1):825–828 18. Gupta V, Awasthi H, Kumar N et al (2021) A novel approach to model threshold voltage and subthreshold current of graded-doped junctionless-gate-all-around (GD-JL-GAA) MOSFETs. Silicon 19. Choi Y, Hong Y, Shin C (2019) Device design guideline for junctionless gate-all-around Nanowire negative-capacitance FET with HfO2 -based ferroelectric gate stack. Semicond Sci Technol 35(1):015011 20. Jimenez D, Miranda E, Godoy A (2010) An analytic model for the surface potential and drain current in negative capacitance field-effect transistors. IEEE Trans Electron Devices 57(10):2405–2409

812

M. K. Rai et al.

21. Gaidhane AD, Pahwa G, Verma A, Chauhan YS (2018) Compact modeling of drain current, charges, and capacitances in long-channel gate-all-around negative capacitance MFIS transistor. IEEE Trans Electron Devices 65(5):2024–2032 22. Srivastava NA, Priya A, Mishra RA (2020) Analog and radio-frequency performance of nanoscale SOI MOSFET for RFIC based communication systems. Microelectron J 98:104731 23. Si P, Zhang K, Yu T, Zhao Z, Lyu WF (2020) Analog/RF performance analysis of nanometer negative capacitance FDSOI transistors. Informacije MIDEM 50(1):47 24. Roy NCh, Gupta A, Rai S (2015) Analytical surface potential modeling and simulation of the junction-less double gate (JLDG) MOSFET for ultra-low-power analog/RF circuits. Microelectron J 46(10):916–922 25. Srivastava NA, Mishra RA (2019) Linearity distortion assessment and small-signal behavior of nano-scaled SOI MOSFET for terahertz applications. ECS J Solid State Sci Technol 8(12):N234 26. Srivastava NA, Priya A, Mishra RA (2022) Performance analysis of gate-stack nanoscaled recessed-S/D SOI-MOSFET for analog applications. In: Recent trends in electronics and communication, Springer, Singapore, pp 613–624 27. Rai S (2017) Reliability analysis of Junction-less Double Gate (JLDG) MOSFET for analog/RF circuits for high linearity applications. Microelectron J 64:60–68 28. Jo J, Shin C (2014) Impact of temperature on negative capacitance field-effect transistor. Electron Lett 51(1):106–108 29. Xiao YG, Tang MH, Li JC, Cheng CP, Jiang B, Cai HQ, Tang ZH, Lv XS, Gu XC (2012) Temperature effect on electrical characteristics of negative capacitance ferroelectric field-effect transistors. Appl Phys Lett 100(8):083508 30. Saha R, Bhowmick B, Baishya S (2018) Temperature effect on RF/analog and linearity parameters in DMG FinFET. Appl Phys A 124(9):642

Prospects of Black Phosphorous in Transit Time Devices Girish Chandra Ghivela and Joydeep Sengupta

Abstract Due to adjustable bandgap of black phosphorous, for the first time an attempt has been made here for the design and development IMPATT diode. Using MATLAB software, the DC performance of the black phosphorous-based double drift region (DDR) IMPATT has been analyzed here at Ka band. The performance is governed through basic transport equations and validated through boundary conditions. Black phosphorous-based IMPATT is giving higher efficiency of 14.45% at 35.5 GHz, which is higher than its counterpart Si and Ge. Keywords Black phosphorous · IMPATT · Drift · Avalanche

1 Introduction Impact ionization and avalanche transit time (IMPATT) diode is one of the premier solid-state devices due to its superior performance over other devices. From long time, silicon (Si) and germanium (Ge) are tested and used for the design of IMPATT. However, these materials are not meeting the required power performances. IMPATT can generate required power performances by using the more advanced and suitable materials. So, from the material perspective and recent advancement in the different fabrication technologies, authors have modeled and designed the IMPATT based on black phosphorous. Black phosphorous is now one of the emerging competitors and became the hot topic of research from last few years. The adjustable band gap of black phosphorous (0.2–2 eV) leads to possibility of black phosphorous as a semiconducting material and for the design and development of IMPATT [1–5]. Also, it has higher mobility. G. C. Ghivela (B) Electronics and Communication Engineering Department, Indian Institute of Information Technology, Nagpur, India e-mail: [email protected] J. Sengupta Electronics and Communication Engineering Department, Visvesvaraya National Institute of Technology, Nagpur, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_62

813

814

G. C. Ghivela and J. Sengupta

Higher mobility contributes more current, as a result the performance of the device improves. This advantage of black phosphorous inspired the authors for considering in design and development of double drift region (DDR) structure IMPATT diode.

2 Simulation Method To compute the DC parameters of black phosphorous-based IMPATT, the 1D-model used is shown in Fig. 1 [6, 7]. The 1D-model is consisting of high-moderatemoderate-high profiles as shown in the Fig. 2. This profile is basically termed as DDR structure since two drift regions are there, one for electron and other for hole charge carriers. In Fig. 1, the red dotted line shows the current density profile. The operating frequency is 26.5–40 GHz. Depletion width of the diode with frequency ‘f ’ is calculated using W n,p = 0.37 Vns,ps /f , where Vns,ps are electron and hole saturated velocity [8, 9]. Any transport phenomena related devices can be studied by using transport equations given by Eqs. (1) and (2) [10–12]. q dE(x) = [ND − NA + p(x) − n(x)], dx ε

(1)

where q, N D and N A are charge of electron, donor and acceptor concentrations; p(x) and n(x) are hole and electron concentrations along the diode length ‘x’ and 1 is the permittivity material.

Fig. 1 Model of IMPATT

Prospects of Black Phosphorous in Transit Time Devices

815

Fig. 2 Concentration profile

1 dJn dp 1 dJ p dn = + g and = + g, dt q dx dt q dx

(2)

where g and J are generation rate and current density of charge carriers, respectively. Solving the Eqs. (1) and (2) gives the field profile. The breakdown voltage (V B ) is the total drops across drift and avalanche (V A ) layers. The DC to RF conversion efficiency is obtained by using Eq. (3). / η(%) = 2m(VB − VA ) π VB ,

(3)

where m = 0.5 [33].

3 Results and Discussions The obtained field profile from the black phosphorous-based diode is shown in Fig. 3. Its DC to RF conversion efficiency is shown in Fig. 4, which is compared with its counterpart Si and Ge. Efficiency from black phosphorous is dominating over Si and Ge over whole frequency range. Black phosphorous IMPATT has higher efficiency value of 14.45% at 35.5 GHz, whereas Si and Ge have 11.279% and 13.22%, respectively, at 36 GHz. The higher efficiency from former one is due to higher mobility and velocity of charge carriers, which results in contribution of more current.

4 Conclusions For the first time, the black phosphorous is tested for the DDR IMPATT and its conversion efficiency is reported here. To the best of author’s survey, black

816

G. C. Ghivela and J. Sengupta

Fig. 3 Electric field profile

Fig. 4 DC to RF conversion efficiency from black phosphorous-based IMPATT, compared with Si and Ge

phosphorous-based DDR IMPATT is not reported yet. Black phosphorous-based IMPATT is giving the better conversion efficiency than Si and Ge-based IMPATTs. This simulation-based data can be taken as a measure for the successful realization of black phosphorous-based DDR IMPATT.

References 1. Island JO, Castelno-Gomez A (2016) Black phosphorus-based nanodevices, In: Lacopi John F, Jagadish BC (eds) Semiconductor and semimetals series, vol 95: 2D material. Elsevier, Amsterdam, pp 279–303 2. Keyes RW (1953) Department of Physics and Institute for the Study of Metals. In: The electrical properties of black phosphorous, vol 92, no 3. University of Chicago 3. Chen X et al (2018) Large-velocity saturation in thin-film black phosphorous. ACS Nano 12(5):5003–5010 4. Madelung O (1996) Semiconductors—Basic data, 2nd edn. Springer, Berlin, p 161 5. Mukhopadhyay J (2009) Studies on the effects of some physical phenomena on the high frequency properties of impatts based on different materials, chapter 5. Ph.D. thesis

Prospects of Black Phosphorous in Transit Time Devices

817

6. Ghivela GC, Sengupta J (2019) Prospects of impact avalanche transit-time diode based on chemical-vapor-deposited diamond substrate. J Electron Mater 48(2):1044–1053 7. Ghivela GC, Sengupta J, Mitra M (2019) Quantum corrected drift diffusion based noise model for impact avalanche and transit time diode. Superlattices Microstruct 128:402–407 8. Sze SM, Ng KK (2007) Physics of semiconductor devices, 3rd edn. Wiley 9. Sze SM, Ryder RM (1971) Microwave avalanche diodes. Proc IEEE 59:1140 10. Ghivela GC, Sengupta J (2019) Noise performance of avalanche transit-time devices in the presence of acoustic phonons. J Comput Electron 18(1):222–230 11. Ghivela GC, Sengupta J (2019) Modeling and computation of double drift region transit time diode performance based on graphene-SiC. Int J Numer Model e2601. https://doi.org/10.1002/ jnm.2601 12. Ghivela GC, Sengupta J (2018) Effect of acoustic phonon scattering on impact ionization rate of electrons in monolayer graphene nanoribbons. Appl Phys A 124(Article no 762):1–8

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA Abhishek Kumar, Santosh Kumar Gupta, and Vijaya Bhadauria

Abstract The paper presents quadrature-phase, in-phase, and out-of-phase inductor-less voltage-controlled oscillator (VCO) using single Voltage Differential Transconductance Amplifier (VDTA). The oscillating frequency is externally controlled by using control voltage (Vctrl). Sensitivity of oscillating frequency with control voltage variation is analyzed and found to be negligibly sensitive to control voltage after 1.5 V. The effect of temperature and process variation on oscillating frequency is analyzed. The circuit is found to be tolerant to variation in control voltage, process, and temperature. The variation in frequency is less than 1% at all corners except at point “SS”. The circuit can be tuned in frequency range 1.76– 2.32 GHz resulting in tuning range of 24%. The power dissipation and phase noise are calculated to be 1.01mW and −95.28 dBc/Hz at 1 MHz offset frequency resulting FOM to be −136.05 dBc/Hz. The maximum total harmonic distortion (THD) is found to be 2.8%. The circuit has been implemented using 180 nm standard library in Cadence Virtuoso tool. Keywords VDTA · Arbel-Goldminz · THD · VCO · FOM

1 Introduction The current-mode approach showed a new path for circuit designer due to its higher bandwidth and higher linearity [1]. Voltage differential transconductance amplifier (VDTA) is one of the active block, which can used as a voltage and current mode, and A. Kumar (B) · S. K. Gupta · V. Bhadauria Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, India e-mail: [email protected] S. K. Gupta e-mail: [email protected] V. Bhadauria e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_63

819

820

A. Kumar et al.

thus gained interest of designer since its introduction by Biolek et al. [2]. Further, the simpler circuit and external tunability make the circuit more ideal for its application in external electronically controllable applications like tunable bi-quad filter, sinusoidal oscillator, frequency dependent negative resistor (FDNR), and active inductor [3–24]. Various reported literatures give more emphasis to design high-performance VDTA structures like low-power CMOS [4], DTMOS, and FD-FVF [5, 6]-based VDTA. Improving performance characteristic plays significant role in deciding overall critical parameters such as bandwidth, quality factor, cutoff frequency, and operating frequency which largely depends on transconductance of Arbel-Goldminz transconductance pair (g M1 or g M2 ) of a VDTA. So, changes in transconductance either due to temperature, process, or any reason adversely affects the performance parameter of the circuit. One of the applications of VDTA is Q-VCO [7–22]. The reported Q-VCO in [12– 15] is using two VDTA and two grounded capacitors, but they generate it by using four independent current sources which results in variation of current in Arbel-Goldminz transconductance pair (g M1 or g M2 ). This independent tunability of transconductance leads to variation in transconductance of g M1 and g M2 . Floating active inductor [16]based digitally tuned VCO generates sinusoidal VCO, which is digitally controlled, but they have used two VDTA for generation of sinusoidal signal and one an active inductor. These reported circuit [4–19] uses independent current source of each Arbel-Goldminz transconductance pair, so there is large effect of process variation on the operating frequency. In this work, three-phase VCO is designed using single VDTA. The VDTA designed for generating VCO do not have any independent current source. The tunability of the VCO is obtained by tuning Vctrl of M16 (shown in Fig. 2) transistor. The four/two independent current source is replaced by single current source. The reference current thus generated is copied through current mirror which act as biasing current for VDTA. The all current source required for basing VDTA is controlled by single voltage source so all variations in current are equally varied. The circuit presented is robust to process, temperature and control voltage variations. Gate voltage of M16 transistor is varied linearly for frequency tunability. The paper also indicates that for minimum variation of oscillating frequency from control voltage (Vctrl) noise, the Vctrl should be biased in deep saturation region. The effect of temperature variation on oscillating frequency is also studied. The maximum sensitivity of oscillating frequency with temperature is obtained is only −0.29%. The circuit faithfully oscillates a from temperature range from −25 to 125 °C. This also indicates that the circuit is tolerable to variation in temperature. Effect of process variation on oscillating frequency is also studied. The effect of process variation affects all NMO equally and similarly in all PMOS, so the effect of process variation will have less impact on oscillating frequency. The circuit is designed using standard 0.18 um CMOS library. The simulation is done using Cadence Virtuoso tool.

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA

821

Fig. 1 Block diagram of VDTA

2 Circuit Description 2.1 Voltage Differential Transconductance Amplifier (VDTA) VDTA is one of the alternative active blocks that provides differential output at one of its output terminal (Z). It also provides positive and negative current at its other output terminals (Ix + at X + and Ix− at X−). Block diagram of VDTA is shown in Fig. 1. The complimented output terminal (ZC) is connected to ground, so VDTA can be described by following matrix: ⎡

⎤⎡ ⎤ ⎤ ⎡ Iz Vp gm 1 −gm 1 0 ⎣ I x− ⎦ = ⎣ 0 ⎦ ⎣ Vn ⎦ 0 −gm 2 0 0 gm 2 Vz I x+

(1)

where V p , V n and V z are the voltage at terminals P, N, and Z, respectively. From above matrix, we get following sets of equations:   Iz = V p − Vn gm 1

(2)

Ix− = −gm 2 Vz

(3)

Ix+ = gm 2 V

(4)

2.2 Circuit Description The circuit consists of two Arbel-Goldminz transconductance [17] (transistor M1-M8) as shown by Fig. 2. The gm 1 of the circuit is calculated as:

822

A. Kumar et al.

Fig. 2 Circuit Diagram of VDTA

 gm 1 =  gm 2 =

g M1 g M2 g M1 + g M2 g M5 g M6 g M5 + g M6



 +



 +

g M3 g M4 g M3 + g M4 g M7 g M8 g M7 + g M8

(5) (6)

In our design, we have taken all NMOS in Arbel-Goldminz transconductance with same transconductance, i.e. gm1 = gm2 = gm5 = gm6 ; and all PMOS with same transconductance i.e. gm3 = gm4 = gm7 = gm8 ; to have minimum process variation effect on Arbel-Goldminz transconductance pair. The above equation reduced to:  gm 1 =  gm 2 =

g M1 + g M3 2 g M5 + g M7 2

(7) (8)

where g M1−8 is transconductance of MOSFETs M1-M8. Circuit diagram of VDTA including biasing circuit is shown in Fig. 2. M9-M16 transistor is used to generate bias current for VDTA. M9-M10, M13-M14 are current

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA Table 1 Transistor sizing

823

Transistors

W(width)(µm)

L(length)(µm)

M1, M2, M5, M6

4.14

0.18

M3, M4, M7, M8

10.08

0.18

M9, M10, M11, M12

1.8

0.18

M13, M14, M15

8.46

0.18

M16

10

0.18

mirrors which are used to copy reference current from transistor M16. Vctrl is the control voltage used to control oscillating frequency. M10-M15 is used to copy current from mirrors. The current from M16 transistor is circulated in circuit for biasing both transconductance pairs (transistor M1-M8) to remove possibility of transconductance variation in M1-M8, and thus, the effect of process and temperature variation is having negligible effect on overall circuit (Table 1).

2.3 VDTA as Three-Phase Oscillator The proposed VCO circuit configuration is shown in Fig. 3. The circuit consists of one resistance R4 and two capacitors C1 and C4. Resistance R4 act as load resistance whose value is 50 Ω (standard load resistance) (Table 2). Applying Kirchhoff law in Fig. 3, the characteristic equation of the circuit is obtained as [15]: Fig. 3 VDTA as Q-VCO

824

A. Kumar et al.

Table 2 Parameters used for simulation

Parameters

Value

R1(load resistance)

50 Ω

R4

1.5 KΩ

C1

18fF

C4

18fF

VDD, VSS

± 0.9 V

1 S +s C4 2



1 gm gm − gm 1 + 1 2 = 0 R4 C1 C4

(9)

So, the condition of oscillation is obtained as: 1 − gm 1 ≤ 0 R4

(10)

Oscillating frequency obtained as: / ω0 =

gm 1 gm 2 C1 C4

(11)

From above equation, oscillating frequency ω0 can be tuned externally by varying gm 1 and gm 2 . Transconductance of a MOSFET is given by: / gm =

I μn Cox

W L

(12)

where μn Cox WL for a MOSFET in IC is fixed so, gm can be varied by varying current through MOSFET. The current can be varied by varying Vctrl (control voltage). For VCO working as linear chirp signal (time varying frequency) generator, M16 transistor should be working in linear region (0.5−0.75 V) for linear variation of drain current. The current at output terminals is obtained from Fig. 3 as: I x+ I x− = gm 2 R4 gm 2 R4

(13)

sC4I x− sC4I x+ = gm 2 gm 2

(14)

I R4 = − IC4 = −

I R4 = −

I x+ I x− = gm 2 R4 gm 2 R4

From Eqs. (1) and (2), it is concluded that:

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA

825

I R4 1 = I x+ gm 2 R4

(15)

IC4 sC4 = I x+ gm 2

(16)

It can be concluded from Eqs. (9) and (10) that the phase difference between I R4 , IC4 and I x− with respect to I x+ are 0◦ , 90◦ , and 180° respectively. The amplitude of the current through IC4 and I R4 with respect to I x+ are not the same, and the amplitude of current I x+ and I x− are same but with phase shift of 180°. The current from IC4 and IC1 (same as I x+ current) form quadrature phase, I R4 and I R1 (same as I x− current) form out-of-phase and IC4 and I R4 form in-phase oscillation.

3 Results and Discussion The VCO is designed using standard 0.18 µm CMOS technology. The generated output is in-phase, out-of-phase, and quadrature phase with each other. Transient waveform is shown in Fig. 4. The circuit takes 23 ns to give steady-state response, and its output is shown in Fig. 5. The current through capacitor C4 and C1 is in quadrature phase with each other and having same amplitude. The current through R1 and R4 is out of phase with each other and having same magnitude. It is clear from Eqs. (9) and (10) that magnitude of current through R1 and R4 is of same magnitude and amplitude of current through I R4 and IC4 dependent on passive component value (R4 and C4) and gm 2 as seen in Fig. 5. Furthermore, location of poles for system at 2.2 GHz frequency is shown in Fig. 6. Majority of poles lies on left of imaginary axis makes system stable. No poles

Fig. 4 Transient response VCO

826

A. Kumar et al.

Fig. 5 Response after transient died (after 23 ns)

lie on right, so system cannot be unstable. Two conjugate poles lie on imaginary axis makes system marginally stable. This pole makes system to oscillate. Since no poles at right hand side of imaginary axis, the oscillating frequency do not increase but oscillate at constant frequency. The Pole-Zero plot indicates that system oscillate at constant frequency keeping all parameters (V ctrl , resistance, capacitance, and Transconductance) as constant.

Fig. 6 Pole location at 2.2 GHz frequency

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA

827

Fig. 7 Variation of frequency with control voltage (V ctrl )

Fig. 8 Sensitivity of frequency with respect to control voltage (V ctrl )

The oscillating frequency can be tuned by using control voltage (V ctrl ) which is seen in Fig. 7, and sensitivity of frequency with control voltage is shown in Fig. 8. When Gate voltage of M16 transistor is above threshold voltage (0.42 V), the transistor enters in saturation region which results in exponential rise drain current. Increasing bias current to VDTA results in increase of transconductance, which can be verified from Eqs. (1–6). When gate voltage is more than 1.2 V, change in drain current is less thus variation in transconductance is least, which is observed in Figs. 7 f and 8. The sensitivity of oscillating frequency with control voltage (sv ctrl ) is given by Eq. (11). The sensitivity is as high as 150% at V ctrl = 0.4 V which decreases to 25% at 0.85 V. The sensitivity decreases to 0.1% beyond 1.5 V so must be biased

828

A. Kumar et al.

Fig. 9 Effect of temperature variation on oscillating frequency

above 1.5 V to have minimum variation of frequency with V ctrl variation. The circuit becomes tolerant to V ctrl noise. f

sv ctrl =

∂f ∂vCtrl

(17)

The differential structure of VDTA results in negligible effect of temperature on oscillator. The proposed biasing circuit affect all transistor equally, thus results in constant incoming (current source) and outgoing current (current sink). Both pair of differential pair is biased by same current source, resulting in equal variation of current in both halves of differential pair. Therefore, effect of temperature is least in the proposed circuit. The effect of temperature on oscillating frequency is shown in Fig. 9. The proposed oscillator is working perfectly for temperature range from −25 to 125 °C. The oscillating frequency is linearly decreasing when decreasing temperature from −25 to 125 °C. The maximum variation of oscillating frequency with respect to room temperature (27 °C) is calculated and presented in Table 3. The maximum variation in frequency is only 13.9% as compared to room temperature. Table 3 indicates less variation in oscillating frequency with significant increase in temperature. The increase in oscillating frequency is due to variation in gm of transistor with respect to temperature.

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA Table 3 Variation of oscillating frequency with Temperature

829

Temperature (°C)

Oscillating frequency (GHz)

Variation (with respect to 27 °C) (%)

−25

2.32

4.4

0

2.27

2.22

25

2.22

0.2

50

2.20

0.9

75

2.12

4.6

100

2.03

8.9

125

1.93

13.9

The basic building block of VDTA, i.e., differential amplifier has inherent tolerant to change in process and mismatch variation. The effect of process variation and device mismatch is performed by Monte-Carlo analysis by taking 1000 sample points. The effect of process variation on oscillating frequency is analyzed. The change in oscillating frequency all corner point is shown in Figs. 10, 11, 12, 13 and 14 and the results obtained are summarized in Table 4. The maximum deviation in frequency is 0.16 GHz at corner point SS (both NMOS and PMOS are slow). The percentage change in standard deviation is less than 1% in all corner point except at corner SS. Total harmonic distortion (THD) with frequency is shown in Fig. 15. The maximum THD is 2.8% at output R1, and minimum is 0.6% at output C1.

Fig. 10 Monte-Carlo analysis at nominal value of transistor

830

A. Kumar et al.

Fig. 11 Monte-Carlo analysis at corner point fast–fast

Fig. 12 Monte-Carlo analysis at corner point fast-slow

Result from similar work has been compared in Table 5. The propose circuit draw 1.01mW power at 2.2 GHz oscillating frequency. The phase noise is found to be −95.28dBc/Hz @1 MHz offset frequency. The circuit can be tuned from 1.76 to 2.32 GHz by varying control voltage from 0.5 to 1.8 V (Fig. 16).

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA

Fig. 13 Monte-Carlo analysis at corner point slow-fast

Fig. 14 Monte-Carlo analysis at corner point slow-slow Table 4 Oscillating frequency variation at different corners Parameter

Mean (GHz)

Standard deviation (GHz)

% Change in variation

Nominal (TT)

2.30

0.017

0.739

Fast–Fast (FF)

2.47

0.009

0.364

Slow-Fast (SF)

2.15

0.018

0.837

Fast-Slow (FS)

2.5

0.017

0.68

Slow-Slow (SS)

1.93

0.16

8.27

831

832

A. Kumar et al.

Fig. 15 Total harmonic distortion at all output

4 Conclusion A voltage tunable Q-VCO is designed using single VDTA, which generates in-phase, quadrature-phase, and out-phase sinusoidal wave. The tunable range obtained by varying control voltage is 0. 56 GHz. The circuit also consumes lesser power when compared to similar work compared in Table 5. The result from Monte-Carlo analysis and from temperature variation indicates that the circuit is tolerable to various environmental, voltage, and process variations. Due to low-power consumption, the proposed circuit is best suited for low-power QAM transmitters.

45-nm CMOS

Yes

1.1–1.8

Technology

Inductor less

Tuning Range (GHz)

1.1

Power consumption (mW)

Figure of merit −176.69 (FOM)

0.9

±1

Supply voltage (V)

−179.03

4.5

−111.56

−98.37

41.55

1.22–1.86

Yes

90-nm CMOS

[23]

Phase noise @ 1 MHz offset (dBc/Hz)

Tuning 48.27 percentage (%)

[17]

References

Table 5 Comparison with similar work

−175.73

21.6

1.8

−111 @3 MHz offset

65.69

1.85–3.66

Yes

180-nm CMOS

[24]

−137.03

8.26

1

−75.2

81.69

2.1–5

Yes

90-nm CMOS

[25]

−182.2

3.6

1.2

−108.84

13.8

5.64–6.48

No

65-nm CMOS

[26]

−167.67

13

1.8

−92.2

81.05

1.13–2.67

Yes

180-nm CMOS

[27]

−175.98

34.8

1.2

−100.5

42.4

6.5–10

No

180-nm CMOS

[28]

−136.05

1.01

± 0.9

−95.28

24.14

1.76–2.32

Yes

180-nm CMOS

This work

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA 833

834

A. Kumar et al.

Fig. 16 Layout diagram of the VDTA including I/O pad (630 × 695 µm2 )

References 1. Toumazou C (1990) Analogue IC design: the current-mode approach. IEEE Circuits and Syst Ser 2 2. Biolek D, Senani R, Biolkova V, Kolka Z (2008) Active elements for analog signal processing: classification, review, and new proposals. Radioengineering 17(4):15–32 3. Leila S, Pandey N, Herencsar N, Khateb F (2017) Special issue on current-mode circuits and systems: recent advances, design and applications 4. Shankar C, Singh SV (2016) Single VDTA based multifunction transadmittance mode biquad filter. Int J Eng Technol (IJET) 7(6):2180–2188 5. Uygur A, Kuntman H (2013) DTMOS-based 0.4 V ultra low-voltage low-power VDTA design and its application to EEG data processing. Radioengineering 22(2):458–466 6. Narang N, Aggarwal B, Gupta M (2017) DTMOS and FD-FVF based low voltage high performance voltage differencing transconductance amplifier (VDTA) and its application in MISO filter. Microelectron J 63:66–74 7. Moonmuang P, Pukkalanun T, Tangsrirat W (2022) Floating/grounded series/parallel RL, RC and LC immittance simulators employing VDTAs and only two grounded passive elements. AEU-Int J Electron Commun 145:154095 8. Borah SS, Singh A, Ghosh M, Ranjan A (2021) Electronically tunable higher-order quadrature oscillator employing CDBA. Microelectron J 108:104985 9. Aggarwal B, Mittal A (2019) Design of temperature and input voltage insensitive VDTA and impedance multiplier. Microelectron J 85:34–51 10. Bhardwaj K, Srivastava M (2021) Wide-band compact floating memristor emulator configuration with electronic/resistive adjustability. Microelectron J 117:105284 11. Roy S, Pal RR (2021) Electronically tunable third-order dual-mode quadrature sinusoidal oscillators employing VDCCs and all grounded components. Integration 76:99–112 12. Suresh LN, Manickam B (2022) Design and application of CMOS active inductor in bandpass filter and VCO for reconfigurable RF front-end. Integration 82:115–126

Tunable Three-Phase Voltage Controlled Oscillator Using Single VDTA

835

13. Yesil, A., Babacan, Y., Kacar, F. (2020) An electronically controllable, fully floating memristor based on active elements: DO-OTA and DVCC. AEU-Int J Electron Commun 123:153315 14. Jantakun A (2016) Voltage differencing transconductance amplifiers based mixed-mode quadrature oscillator. Revue Roumaine des Sciences Techniques-series Electrotechnique et Energetique 61(1):68–72 15. Punnavich P, Lamun P, Kumngern M, Torteanchai U (2014) Current-mode third-order quadrature oscillator using VDTA and grounded capacitors. In: Information and communication technology, electronic and electrical engineering (JICTEE), 2014 4th joint international conference on, IEEE, pp 1–4 16. Mehra R, Kumar V, Islam A (2018) Floating active inductor based class-C VCO with 8 digitally tuned sub-bands. AEU-Int J Electron Commun 83:1–10 17. Dinesh P, Bhaskar DR (2012) Electronically controllable explicit current output sinusoidal oscillator employing single VDTA. ISrN Electron 18. Mayank S, Prasad D, Bhaskar DR (2014) Voltage mode quadrature oscillator employing single VDTA and grounded passive elements. Contemporary Eng Sci 7:1501–1507 19. Arbel AF, Goldminz L (1992) Output stage for current-mode feedback amplifiers, theory and applications. Analog Integr Circ Sig Process 2(3):243–255 20. Kumar A, Gupta SK, Bhadauria V (2022) Low-power and low glitch area current steering DAC. Eng Sci Technol an Int J 29:101035 21. Kumar A, Gupta SK, Bhadauria V (2022) A 12-bit SC3 partially segmented current steering DAC with improved SFDR and bandwidth. Int J Circuit Theory and Appl 22. Kumar A, Gupta SK, Bhadauria V (2022) Design of IF-RF-based heterodyne transmitter using current steering DAC with 5.4 GHz spur-free bandwidth. IETE J Res 1–16 23. Jang S-L, Chen M-H, Chang C-W, Juang M-H (2012) A complementary cross-coupled voltagecontrolled oscillator using differential active inductor. Microw Opt Technol Lett 54(9):2039– 2042 24. Huang G, Kim B-S (2008) Programmable active inductor-based wideband VCO/QVCO design. In: IET Microwaves, Antennas & Propagation, vol 2(8). pp 830–838 25. Tsitouras A, Plessas F, Kalivas G (2011) A linear, ultra wideband, low-power, 2.1–5 GHz, VCO. Int J Circuit Theory Appl 39(8):823–833 26. Amin T, Mak P-I, Martins RP (2015) A 3.6-mW 6-GHz current-reuse VCO-buffer with improved load drivability in 65-nm CMOS. Int J Circuit Theory Appl 43(1):133–138 27. Jeong Y-J, Kim Y-M, Chang H-J, Yun T-Y (2012) Low-power CMOS VCO with a low-current, high-Q active inductor. IET Microwaves, Antennas and Propagation 6(7):788–792 28. Yuan F, DiClemente D (2015) Hybrid voltage-controlled oscillator with low phase noise and large frequency tuning range. Analog Integr Circ Sig Process 82(2):471–478

Double-Core Photonic Crystal Fiber for Liquid Sensing Detection Shreya Gupta, Dharmendra Kumar, Vijay Shanker Chaudhary, and Sneha Sharma

Abstract In this paper, photonic crystal fiber with two cores is designed for sensing application and numerically studied using the finite element method-based COMSOL Multiphysics software. One central air hole forms the two cores can be filled with different liquids for sensing applications. The significant mode of the fiber structure is examined, and the designed sensor has a maximum coupling length of is 90.1, 104.4, 110.3, and 119.1 mm, respectively, for sensing pentanol, benzene, alcohol, and water at the wavelength of 1800 nm. According to theoretical analysis, the suggested photonic crystal fiber-based sensor with a length of only 0.03 cm has a sensitivity of 3750 nm/RIU, 4166.666 nm/RIU, 5000 nm/RIU, and 7500 nm/RIU, respectively, for pentanol, benzene, alcohol, and water. Keywords Single-mode PCF · Liquid detector · Sensitivity · V parameter

1 Introduction A significant turning point in the history of optical fibers was the development of photonic crystal fibers (PCF) [1] with outstanding optical characteristics in birefringence [2], dispersion [3], and nonlinearity [4]. Fiber optics is a field that has expanded beyond its traditional applications in telecommunication and medicine and now has S. Gupta (B) · D. Kumar (B) · V. S. Chaudhary Department of Electronics and Communication Engineering, Madan Mohan Malaviya University of Technology Gorakhpur, Gorakhpur 273010, India e-mail: [email protected] D. Kumar e-mail: [email protected] S. Sharma Department of Electronics and Communication Engineering, Dayananda Sagar University, Bangalore, India V. S. Chaudhary Department of Computer Science Engineering, Galgotias College of Engineering and Technology, Greater Noida, Uttar Pradesh 201310, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_64

837

838

S. Gupta et al.

a broad range of applications. In the field of fiber optics, a new type of optical fiber called PCF is nowadays having many applications. PCF can be employed as both a transmission medium and an optical functional device. PCFs have additional design parameters compared to standard optical fiber, such as air hole diameter, pitch size, and ring count, which offer to get around many of the drawbacks of conventional fiber. PCFs have drawn a lot of attention in the development of opto devices and sensors because of their well-known benefits, including increased design freedom, low cost, quick detection, compact size, reliability, high sensitivity, and flexibility. Due to their extraordinary performance and wide range of potential applications, photonic crystal fibers (PCFs) have received a lot of interest. PCF can be utilized as filters, switches, electro-optical modulators, polarization converters, sensors, and more [5–21]. PCFbased sensors are intelligent fiber-optic applications that have been researched and developed during the previous decade. There are numerous sensing applications for PCF, including temperature sensors [11], R.I sensors [12], chemical sensors [13], mechanical sensors [14], pressure sensors [15], gas sensors [16, 17], stress sensors [18], pH sensors [19], liquid sensors [20], biosensors [21], and among others. To solve safety concerns, highly sensitive chemical (e.g., liquid and gas) sensors are crucial in industrial operations [22], particularly for detecting poisonous and highly explosive compounds (such as harmful gases or liquids). Enhancing the performance of liquid sensors has thus emerged as one of the major issues. Researchers have been particularly interested in the development of PCF-based sensors for safety and environmental monitoring [23, 24]. Since the core of the fiber directly interacts with the material being studied, liquid and gas sensors based on PCF exhibit exceptional performance in the terms of sensitivity. The double-core fiber sensor is extremely simple and structured because of its high response and lack of a metal wire. In this study, we suggest a simple triangular lattice-based double-core PCF (DCPCF) for liquid sensing application. Based on the electric field distribution in the cores of the fiber, the sensor’s sensitivity can be determined. When the hole exactly at the middle of the fiber is infiltrated with various types of liquids pentanol, benzene, alcohal, and water the suggested structure gives the sensitivity of up to 3750 nm/RIU, 4166.666 nm/RIU, 5000 nm/RIU, and 7500 nm/RIU with length of 0.03 cm.

2 Design of PCF A cross-sectional view of a proposed liquid detector DC-PCF is shown in Fig. 1. The guiding characteristics of sensor is examined using COMSOL Multiphysics software which is based on FEM. The suggested DC-PCF is made up of a triangular lattice of circular air holes with two missing holes as two fiber cores distinguished by one air hole. The center air hole that divides the cores of the fiber is packed with various liquid agents like pentanol, benzene, alcohol, and water. The pitch “A” and diameter “d” of air holes are 2.5 µm and 1.85 µm, respectively. Based on the Sellmeier equation [25], the RI of the silica material is 1.444 at 1.55 µm wavelength, and the RI of the air is considered as 1.0. Based on mode-coupling theory [26], the double-core fiber

Double-Core Photonic Crystal Fiber for Liquid Sensing Detection

839

Fig. 1 Proposed structure of dual core PCF for chemical sensing

contains four different modes in the orthogonal polarization axis, such as (x-even and odd, y-even and odd), respectively. Using COMSOL software, these four basic modes of the designed fiber sensor are examined, and electric-field propagation is displayed in Fig. 2. In Fig. 3a–b shows the refractive index of the two modes x-even and y-even of different chemicals which is continuesly decreases with wavelength. The value of Δneo = |ne –no | in the wavelength range of 1.8−3.1 µm is shown in Fig. 4, which reveals that the coupling length changes slightly for various operation wavelengths. Equation specifies the coupling length (L i ) [26] Li =

λ 2(n ie − n io )

(1)

where i = x, y and L i indicate the coupling length, and nio and nie stand the effective RI of the i-polarized even and odd modes, respectively. Figure 5 shows how the variation of effective refractive index differences with wavelength for several liquids. As light spreads out of the core, the RI of basic mode of the proposed fiber is decreases with wavelength. By increasing the wavelength, light confinement in the core region decreases. As it spreads out from the core, the RI drops for all four modes. The V parameter of the fiber can be find using V PC F =

2π A / 2 n λ − n 2e f f λ

(2)

where nλ is the RI of the fiber background material and neff is the real part of the effective refractive index for either of the modes. V parameter is the important optical characteristic that determines the single-mode operation of the fiber. According to a literature survey, photonic crystal fiber can be called a single-mode operation fiber if its V parameter value is less than 3.142. Figure 6 shows the V-parameter of the

840

S. Gupta et al.

(a)

(b)

(c)

(d)

Fig. 2 Electric field distribution of the proposed DC-PCF a x-odd b x-even c y-odd d y-even polarized modeat 1.55 µm wavelength

proposed fiber structure with respect to wavelength. It can be observed that fiber gives a single mode of operation in the complete wavelength range of 1.8–3 µm. Optical power exchanged between one core to another core in a proposed DC-PCF after length L is determined by coupling mode theory [26]. P(λ) = sin2

|n e − n o |πL λ

(3)

The analyzed transmission curve for mode x-polarized of the proposed double core PCF liquid detector with a fixed length of 0.5 cm for different liquids which is filled in the central air hole is shown in Fig. 7. The DC-PCF transmission curve is sinusoidal. It is a continuous wave as well as a uniform periodic function. Focusing on the peak wavelength reveals a linear relationship between it and changing liquid refractive index (RI), as illustrated in Fig. 6. The structure of PCF characteristics influences how the liquid sensor responds to a wavelength shift. According to Fig. 7, it can be possible to further optimize the structure of PCF for liquid sensing by comparing the calculated wavelength-shift of the transmission spectra of PCFs with various structure parameters versus the RI of liquids.

Double-Core Photonic Crystal Fiber for Liquid Sensing Detection

841

1.39

Fig. 3 Graph between refractive index of polarized mode a x-even and b y-even with wavelength

Benzene Water Alcohol Pentanol

(a)

1.38

Refractive Index (n ) x

1.37 1.36 1.35 1.34 1.33 1.32 1.31

3.8

3.6 3.4 Wavelength (μ m)

3.2

3

4

1.38 1.37

Benzene Water Alcohol Pentanol

(b)

y

Refractive Index (n )

1.36 1.35 1.34 1.33 1.32 1.31 1.3

3

3.6 3.4 Wavelength (μ m)

3.2

3.8

4

0.03

Fig. 4 Graph b/w the wavelength and refractive index for different liquids

Benzene Water Alcohol Pentanol

0.025

Δ neo

0.02

0.015

0.01

0.005

2

2.2

2.4 2.6 Wavelength (μ m)

2.8

3

842 120

Benzene Water Alcohol Pentanol

110 Coupling length (mm)

Fig. 5 Coupling length of proposed dauble core-PCF with wavelength for different liquids

S. Gupta et al.

100 90 80 70 60 50

2

2.6 2.4 Wavelength (μ m)

2.2

2.8

3

2.46

Fig. 6 V-parameter of the proposed structure with respect to wavelength V PCF

2.45

2.44

2.43

2.42 1.8

2.2 2.4 2.6 Wavelength (μ m)

2.8

3

1

0.8

Transmission

Fig. 7 Transmission spectra for x-polarized light of a proposed sensor with a length of 0.03 cm for different liquid such as pentanol, benzene, alcohol, and water

2

0.6

0.4

Benzene Water Alcohol Pentanol

0.2

0 1.8

2

2.2

2.4 2.6 Wavelength (μ m)

2.8

3

Double-Core Photonic Crystal Fiber for Liquid Sensing Detection

843

Table 1 Comparison of the proposed fiber with other published works Reference

Proposed structure

Sensitivity

[33]

Titanium nitride coated PCF-SPR sensor

wavelength sensitivity 10,000 nm/RIU in the refractive index (RI) range of 1.385–1.40

[34]

Octagonal cladding with a rotated-Hexa-core in PCF

Relative sensitivity at 77.14%, 78.06%, and 76.11% at 1 THz for Ethanol (n = 1.354), Benzene (n = 1.366), and Water (n = 1.330), respectively

[35]

Plasmon resonance sensor based on D-shape PCF

wavelength sensitivity is 3940 nm/RIU with analyte RI b/w 1.35 and 1.40

Proposed structure

Triangular lattice-based double core PCF liquid sensor

Sensitivity of 3750 nm/RIU, 4166.66 nm/RIU, 5000 nm/RIU, 7500 nm/RIU, for pentanol, benzene, alcohol, and water using 0.03 cm length fiber

To calculate the sensitivity of the sensor [27–32] Sλ =

Δλpeak Δn a

(4)

where λpeak denotes a change in the transmission curve and Δna represents a change in the liquid RI. The proposed DC-PCF produces the sensitivity at a length 0.03 cm for pentanol, benzene, ethanol, and water such as 3750 nm/RIU, 4166.66 nm/RIU, 5000 nm/RIU, 7500 nm/RIU, respectively. The proposed design of double core PCF is very simple to fabricate and easy than other related research works. Table 1 shows the comparison between the proposed fiber with other published works.

3 Conclusion A DC-PCF sensor for liquid sensing with a triangular lattice has been proposed. FEMbased software is used for the theoretical analysis of the designed fiber liquid detector. Comparing it to other reported efforts, the suggested DC-PCF chemical sensor design is quite simple and is easily built using existing techniques. The designed liquid sensor DC-PCF provides high sensitivity at the length 0.03 cm for pentanol, benzene, ethanol, and water is 3750 nm/RIU, 4166.66 nm/RIU, 5000 nm/RIU, 7500 nm/RIU, respectively.

844

S. Gupta et al.

References 1. Knight JC (2003) Photonic crystal fibers. Nature 424:847–851 2. Ortigosa-Blanch A, Knight JC, Wadsworth WJ, Arriaga J, Mangan BJ, Birks TA, Russell PSJ (2000) Highly birefringent photoniccrystal fibers. Opt Lett 25:1325–1327 3. Saitoh K, Koshiba M, Hasegawa T, Sasaoka E (2003) Chromatic dispersioncontrol in photonic crystal fibers: application to ultra-flatteneddispersion. Opt Exp 11:843–852 4. Knight JC, Skryabin DV (2007) Nonlinear waveguide optics andphotonic crystal fibers. Opt Expr 15:15365–15376 5. Arjmand M, Talebzadeh R (2015) Optical filter based on photonic crystal resonant cavity. Optoelectron Adv Mater-Rapid Commun 9(1–2):32–35 6. Fasihi K (2014) High-contrast all-optical controllable switching and routing in nonlinear photonic crystals. J Lightwave Technol 32(18):3126–3131 7. Cui K, Zhao Q, Feng X, Huang Y, Li Y, Wang D et al. (2012) Thermo-optic switch based on transmission-dip shifting in a double-slot photonic crystal waveguide. Appl Phys Lett 100(20):201102–1–201102–4 8. Brosi JM, Koos C, Andreani LC, Waldow M, Leuthold J, Freude W (2008) High-speed lowvoltage electro-optic modulator with a polymer-infiltrated silicon photonic crystal waveguide. Opt Expr 16(6):4177–4191 9. Gao Y, Shiue RJ, Gan X, Li L, Peng C, Meric I et al (2015) High-speed electro-optic modulator integrated with graphene-boron nitride heterostructure and photonic crystal nanocavity. Nano Lett 15(3):2001–2005 10. Xuan H, Ma J, Jin W, Jin W (2014) Polarization converters in highly birefringent microfibers. Opt Expr 22(3):3648–3660 11. Chang YH, Jhu YY, Wu CJ (2012) Temperature dependence of defect mode in a defective photonic crystal. Opt Commun 285(6):1501–1504 12. Liu Y, Salemink HWM (2014) All-optical on-chip sensor for high refractive index sensing in photonic crystals. Europhys Lett 107(3):1160–1170 13. Zheng S, Zhu Y, Krishnaswamy S (2012) Nanofilm-coated photonic crystal fiber longperiodgratings with modal transition for high chemicalsensitivity and selectivity. Proc SPIE 8346(14):1844–1864 14. Lee C, Thillaigovindan J (2009) Opticalnanomechanical sensor using a silicon photonic crystal cantilever embedded with a nanocavityresonator. Appl Opt 48(10):1797–1803 15. Olyaee S, Dehghani AA (2013) Ultrasensitivepressure sensor based on point defect resonant cavity in photonic crystal. Sens Lett 11(10):1854–1859 16. Zhang YN, Zhao Y, Wang Q (2013) Multi-component gas sensing based on slotted photonic crystal waveguide with liquid infiltration. Sens Actuators, B Chem 184(8):179–188 17. Morshed M, Arif MFH, Asaduzzaman S, Ahmed K (2015) Design and characterization of photonic crystal fiber for sensing applications. Europ Scient J 11(12):228–235 18. Lu TW, Lee PT (2009) Ultra-high sensitivityoptical stress sensor based on doublelayeredphotonic crystal microcavity. Opt Expr 17(3):1518–1526 19. Hu P, Dong X, Wong WC, Chen LH, Ni K, Chan CC (2015) Photonic crystal fiber interferometricpH sensor based on polyvinyl alcohol/polyacrylicacid hydrogel coating. Appl Opt 54(10):2647–2652 20. Lai WC, Chakravarty S, Zou Y, Chen RT (2013) Multiplexed detection of xylene andtrichloroethylene in water by photonic crystalabsorption spectroscopy. Opt Lett 38(19):3799–3802 21. Akowuah EK, Gorman T, Ademgil H, Haxha S, Robinson GK, Oliver JV (2012) Numericalanalysis of a photonic crystal fiber for biosensingapplications. IEEE J Quantum Electron 48(11):1403–1410 22. Pushkarsky MB, Webber ME, Baghdassarian O, Narasimhan LR, Patel CKN (2002) Laserbasedphotoacoustic ammonia sensors for industrialapplications. Appl Phys B 75(2–3):391–396 23. Whitenett G, Stewart G, Atherton K, Culshaw B, Johnstone W (2003) Optical fibre instrumentation for environmental monitoring applications. J Opt A: Pure Appl Opt 5(5):S140–S145

Double-Core Photonic Crystal Fiber for Liquid Sensing Detection

845

24. Carvalho JP, Lehmann H, Bartelt H, Magalhaes F, Amezcua-Correa R, Santos JL et al (2009) Remote system for detection of low-levels of methane based on photonic crystal fibres and wavelength modulation spectroscopy. J Sens 2009(2):1–10 25. Chen D, Gufeng H, Chen L (2011) IEEE Photon Technol Lett 23:1851–1853 26. Huang WP (1994) Coupled-mode theory for optical waveguides: an overview. J Opt Soc Am A 11:963–983 27. Paul BK, Ahmed K, Ahmed F, Roy S, Abbott D (2018) IEEE Sens J 18:9948–9954 28. Olyaee S, Naraghi A, Ahmadi V (2014) Optik-Int J Light Electron Opt 125:596–600 29. Tian M, Lu P, Chen L, Lv C, Liu D (2012) Opt Commun 285:1550–1554 30. Sun B, Chen M-Y, Zhang Y-K, Yang J-C, Yao J-Q, Cui H-X (2011) Opt Exp 19(5):4091–4100 31. Ayyanar N, Vigneswaran D, Sharma M, Sumathi M, Mani Rajan MS, Konar S (2017) IEEE Sensors J 17(3):650–656 32. Singh A, Chaudhary V, Chaudhary VS, Kumar D (2018) In: 5th IEEE Uttar Pradesh section international conference on electrical, electronics and computer engineering (UPCON) during 2–4 Nov. 2018, M.M.M. University of Technology Gorakhpur, India. https://doi.org/10.1109/ UPCON.2018.8597161 33. Kaur V, Singh S (2019) Design of titanium nitride coated PCF-SPR sensor for liquid sensing applications. Opt Fiber Technol 48:159–164 34. Abdullah-Al-Shafia M, Sen S (2020) Design and analysis of a chemical sensing octagonal photonic crystal fiber (O-PCF) based optical sensor with high relative sensitivity for terahertz (THz) regime. Sens Bio-Sens Res 29:100372 35. Chen A, Yu Z, Dai B, Li Y (2021) Highly sensitive detection of refractive index and temperature based on liquid-filled D-shape PCF. IEEE Photon Technol Lett 33(11):529–532. https://doi. org/10.1109/LPT.2021.3073425

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes Ashish K. Rao, Siddharth Srivastava, Arun Kumar Singh, and Neelam Srivastava

Abstract In this paper, we have studied the combined impact of cognitive radio (CR) node mobility and imperfect channel state information (CSI) on full-duplex cognitive radio (FD-CR) spectrum sensing performance for the cooperative spectrum sensing (CSS) scenario. Jake’s model is used to model CR node mobility. The Nakagamim fading channel environment is considered to obtain the expression for detection probability and false alarm probability. Total error rate (TER), area under the ROC curve (AUC), and receiver operating characteristic (ROC) are all calculated. Monte Carlo simulations are used to verify all the analytical analysis. Keywords CR node mobility · Nakagami-m fading channel · Full-duplex cognitive radio (FD-CR) · Channel state information (CSI)

1 Introduction Today, we cannot imagine a world without wireless communications and sophisticated mobile devices. Mobile services have progressed from simple phone communication to mobile broadband multimedia services and complex wireless apps during the last decade. Any wireless communication depends on the availability of a radio frequency spectrum, a finite resource that cannot be created. Academics and companies have been searching for creative ways to use the wireless spectrum more wisely and effectively as a result of the tremendous increase of wireless devices and services. Full-duplex (FD) communication is a promising technique that has recently been created to improve spectrum use and network efficiency. Its combination with cognitive radio (CR) boosts performance even more. CR technology has been intensively explored for the past 20 years and made available for few commercial applications. A. K. Rao (B) · A. K. Singh Deptartment of ECE, REC Kannauj, Aher, Uttar Pradesh, India e-mail: [email protected] S. Srivastava · N. Srivastava Department of ECE, IET, Lucknow, Uttar Pradesh, India © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_65

847

848

A. K. Rao et al.

In full-duplex wireless communication, the data transmission and reception are done simultaneously [1–3]. The advancement of the self-interference suppression (SIS) technique in FD systems motivates the FD communications technology to use in cellular and cognitive radio technology. In this paper, cooperative spectrum sensing is performed considering the mobile CR node under the generalized Nakagami-m fading channel. The exact channel state information (CSI) is practically not perfect in wireless communication. Thus, this paper analyzes another crucial factor: the FD-CR spectrum sensing performance under imperfect CSI conditions. The previous studies [4–6] covered the spectrum sensing investigation under Nakagami–m fading channel. In [7–9], the mobility of the primary transmitter (PT) and CR is discussed. The author addresses the HD-CR performance study considering channel estimate error in [10–13]. To our knowledge, no study has evaluated how time-selective Nakagami-m fading, CR node mobility, and imperfect CSI interact to affect FD-CRN spectrum sensing performance.

1.1 Research Contribution and Motivation This study analyzes how CR node mobility and CSI error affect spectrum sensing. The following are the crucial contributions. • Under CR node mobility and imperfect CSI conditions, the expression for detection and false alarm probability has been derived. • The ROC, AUC, and TER metrics are used to analyze the FD-CR system performance, and it is found that poor CSI and increased FD-CR mobility led to degrades system performance. • The spectrum sensing performance under CR node mobility and imperfect CSI conditions is evaluated and compared with stationary CR and perfect CSI conditions. The CSI error and CR node mobility are found to be crucial parameters to be considered while developing the FD-CR system. The other sections of the study include Sect. 2 (offers the system and channel models), Sect. 3 (presents the CSS system performance), Sect. 4 (discusses the numerical and simulation results), and Sect. 5 (which concludes the paper).

2 System Model 2.1 Channel Model Nakagami-m distributed random variables are used to describe the sensing and RSI channels. According to [14, (Eq. 2.20)], the channel probability density function

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

849

(PDF) is defined. 

mi j f (h i j ) = 2 Δi j

m i j

2m −1   hi j i j mi j 2   exp − h , hi j ≥ 0 Δi j i j [ mi j

(1)

where h i j is the channel coefficient, which is also the random variable, m is the Nakagami-m fading parameter, and ij transmitter–receiver pair index. As stated in [14, (Eq. 2.21)], the PDF of the square of the random variable, gi j = h i2j can be given as  f (gi j ) =

mi j Δi j

m i j

m −1   gi j i j mi j   exp − gi j , gi j ≥ 0 Δi j [ mi j

(2)

2.2 CR Mobility The CR node is assumed to be movable in the given system model, whereas the PT node is considered stationary. Due to node mobility, the channels between stationary PT and mobile CR, also known as the sensing channel, and between transmitting mobile CR and remaining mobile CR, also known as the interference channel, are represented as time selective  flat  fading channels. Jake’s model is used to determine between adjacent samples for time-varying channels ρ the correlation coefficient i j   2π f v T

c ij s [15]. Here, the carrier frequency is f c , the speed of light is c, the as ρi j = J0 c sampling period is T s , and the J0 (·) function is a zero-order first-kind Bessel function. The CR node mobility is modeled using the first-order autoregressive process. The channel response is thus provided as [16].

h i j (τ1 ) = ρi j h i j (τ2 ) +

/

1 − ρi2j ei j (τ1 )

(3)

where h i j (τ1 ) and h i j (τ2 ), respectively are used to represent the channel gain over complex Gaussian time instants τ1 and τ2 . . This channel gain is circularly symmetric   2 distributed (ZM-CSCG) having zero mean as h i j (τ ) = CN 0, σh ij . ρi j is the correlation coefficient. It is to note that ei j (τ1 ), is the time-varying component   important with ZM-CG as CN 0, σi2j . The kth signaling index is taken into account in the entire paper. If x(k) is the PT transmitted signal, received signal by CR can be as yi j (k) = h i j (k).x(k) + n i j (k)   Here, n i j (k) denotes the noise with variance σn2i j i.e., n i j ∼ CN 0, σn2i j .

(4)

850

A. K. Rao et al.

For every kth transmitted symbol, CR must have an estimated channel coefficient h i j (k) according to the system model that is being used. The time-varying nature of the channel makes it impossible for CR to follow the gain of the fading channel. Therefore, only the first signaling period is estimated by CR. Thus, the received signal can be represented in terms of CSI error, CR node mobility is given as h i j (k) =

ˆ ρik−1 j h i j (1)

+

/

1−

ρi2j





hˆ i j (k)

k−1

ρik−l−1 ei j (l) + ρik−1 j h i j (1) j

l=1

hˆ i j (k) ϕi j (k)

(5)

where hˆ i j (k) is the channel estimation error resulting from  ϕi j (k) is the noise   and 2(k−1) 2 σi j . In order to obtain node mobility. ϕi j (k) is distributed as CN 0, 1 − ρi j the received signal yi j (k) Eq. (5) substituted into Eq. (4) as follows: yi j (k) = hˆ i j (k)x(k) + hˆ i j (k)x(k) + n i j (k)

(6)

The desired signal, noise resulting from node mobility, noise resulting from channel estimate error, and white noise are all components of the received signal in Eq. (6). The effective noise power can be expressed as   2(k−1) 2 σi2j E s + σn2i j σi2j E s + ρi2(k−1) σeff i j = 1 − ρi j j

(7)

where E s is the PT signal’s energy. The stationary CR condition is given by ρi j = 1 in the equation above, and the perfect CSI condition, which is used to compare system performance, is given by σi2j = 0.

3 Performance Analysis of Cooperative Spectrum Sensing Figure 1 depicts the model of the CSS system. The hidden node issue with non-CSS can be avoided using this system model. It comprises an M number of FD-CR nodes, a primary transmitter PT, a primary receiver PR, a CR receiver, and an FC. All of the CRs send the locally detected data to FC. Based on a few fusion rules, the FC provides the global decision about the status of the primary user. In this case, it is presumable that only CR1 will transmit data during each time slot if the PU signal is inactive. Each M CRs in the system model detects the spectrum, and the (M − 1) CRs aid CR1 in sensing. As a result, CR1 can interfere with the other (M − 1) CRs. As a result, the kth received signal sample at the pth CR, where = 2, 3, . . . , M, can be expressed as follows:

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

851

Fig. 1 CSS system model

⎧ (k−1) ⎪ ρ1 p h 1 p (1)s(k) + ϕ1 p (k)s(k) + n p (k) H0 ⎪ ⎪ ⎨ ρt(k−1) h t p (1)xt (k) + ρtk−1 p p h ∊t p (1)x t (k)+ y p (k) = (k−1) ⎪ ρ1 p h 1 p (1)s(k) + ϕt p (k)xt (k) ⎪ ⎪ ⎩ H1 +ϕ1 p (k)s(k) + n p (k) ∆





(8)

The received signal is denoted by y p (k), and the ZM-CG noise denoted by n p (k) has variance of σn2p . The SIS factor ranges (0 < χ < 1). The SI channel gain of CR is represented as h 11 (k). The ϕ1 p (k) and ϕt p (k) represents the noise resulting from node mobility from CR1 to pth CR and from PT to pth CR respectively. The signals transmitted by PT and CR with powers of σt2 and σs2 respectively are denoted as xt (k) and s(k). When the reporting channel is imperfect, the gains of the sensing hˆ t p (1), channel, interference channel, and reporting channel are respectively ρt(k−1) p (k−1) ˆ ρ1 p h 1 p (1), and ρtk−1 p h t p (1). For an energy detection (ED)-based sensing technique, the decision statistic (E p ) | ∑N | | y p (k)|2 , where N denotes the number of the can be expressed as E p = N1 k=1 received signal y p (k) samples. This equation is used to determine the energy of the received signal. For each of the two hypotheses, the E p for pth CR follows the chisquare distribution. Thus, [4], Eq. 4] can be used to express the conditional CDF of the decision statistic E p .

FE (x|H) = 1 −

  Nx [ N , 2σ 2 p

[(N )

(9)

Here [(·, ·) and [(·) respectively, denote the incomplete and complete gamma functions [17, (Eq. 8.310, 8.350)]. For either of the hypotheses, the variance of the received signal is denoted as σ 2 . The simplified form of Eq. (9) can be written as given in [17, (Eq. 8.352)]. The variance for each of the hypotheses at the pth CR is σ p2 . The

852

A. K. Rao et al.

equation above can be expressed as 

−N x FE p (x|H) = 1 − exp 2σ p2

 N −1  l 1 Nx , l! 2σ p2 l=0

(10)

The variance under the null hypothesis H0 can be expressed as | |2   | 2(k−1 2| ˆ h σ p2H = ρ12(k−1) σ + 1 − ρ σ12p σs2 + σn2p (1) | | 1p s 1p p 0



σu2p

(11)

0

| |2 | | where |hˆ 1 p (1)| is a gamma-distributed random variable, and σu2p is the effective 0

variance. Consequently, the PDF for σ p2mathcal H is given as 0

  m 1 p  −m 1 p y m1 p 1 exp 2(k−1) f σ p2 H (y) = 0 [(m1p ) ρ12(k−1) σs2 Δ1 p ρ1 p σs2 Δ1 p p   m 1 p m 1 p −1 m 1 p −r −1 m1 p − 1   m 1 p σu2p × exp 2(k−1) 0 , y r −σu2p 0 r ρ1 p σs2 Δ1 p r=0 y ≥ σu2p

0

(12)

where m 1 p is the shape parameter and ρ12(k−1) Δ1 p E is the interference channel p variance. Therefore, the CDF of the p is expressed under the hypothesis H0 as ⎡

  ∞ N −1 m 1 p −1 (−1)a  N x l+a m 1 p σu2p 1 0 ⎣ FE p0 (x) = 1 − ex p 2(k−1) [(m1p ) 2 ρ1 p σs2 Δ1 p a=0 l=0 r =0 l!a!  m 1 p −r −1 × −σu2p 0 m 1 p −r+l+a−1   m1 p m11 − 1 × r ρ12(k−1) σs2 Δ1 p p   m 1 p σu2p 0 ×[ r − l − a + 1, 2(k−1) (13) ρ1 p σs2 Δ1 p Under hypothesis H1 , the received signal variance is given as | | |2 |2 | | 2(k−1) 2 | ˆ 2| ˆ σ p2H = ρt2(k−1) σ + ρ σ h h (1) (1) | | | | tp 1p p t s 1p 1

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

853

   | |2  + ρt2(k−1) σt2p |h t p (1)| + 1 − ρt2(k−1 σt2p σt2 + 1 − ρ12(k−1 σ12p σs2 + σn2p p p p

σu2p

1

(14) | |2 | | Hence, under H1 , the effective noise variance is σu2p . Gamma variates (|hˆ t p (1)| 1 | |2 | | and |hˆ 1 p | ) are added to create the variance σ p21 . σ p2H ’s PDF can be provided as 1

f σ p2

H1

  r  2 b−r −1 m q b−1 2 Ξ (q, b) b − 1 y −σu p1 2 ηqb (b−1)! (y) = r q=1 b=1 r =0

 2    σu p y 1 × exp − exp , y ≥ σu2p 1 ηq ηq ρ 2(k−1) σ 2 Δ

(15)

ρ 2(k−1) σs2 Δ1 p . m1 p

where m 1 = m t p , m 2 = m 1 p , η1 = t p m t pt t p and η2 = 1 p The CDF of E p is provided under the hypothesis H1 as

    N x l+a (−1)a b−1 Ξ2 (q, b) FE p1 (x) = 1 − r l!a!(b − 1)! 2 a=0 l=0 q=1 b=1 r=0  2     b−r−1 σu p σu2p 2 r−l−b−a+1 1 1 × −σu p ηq exp [ r − l − a + 1, (16) 1 ηq ηq m q r−1 N −1 ∞ 2

For the pth FD-CR, the Pd and P f can be expressed as [5, (Eq. 17)].   P f p () = Pr E p > ε|H0 = 1 − FE p0 (ε)

(17)

  Pd p (ε) = Pr E p > ε|H1 = 1 − FE p1 ()

(18)

To determine the performance of the pth FD-CR spectrum sensing, utilize the equations above. It is assumed that each FD-CR and FC channel has an error probability of δ. All the sensing choices made at FC are combined using the logical OR rule. Accordingly, the FC’s overall probability of detection and the false alarm are given as [5, (Eq. 38)]. P fFC = 1 − PdFC = 1 −

 

 M

1 − P f p (ε)(1 − δ) + δ P f p (ε)

1 − Pd p (ε))(1 − δ) + δ Pd p (ε)

 M

(19) (20)

854

A. K. Rao et al.

The probability of miss detection and false alarm can be added to determine the global TER.   FC Pter ≜ P fFC + 1 − PdFC

(21)

Equations (17) through (20) can be used to calculate the probability of detection and false alarm, and Eqs. (19) and (20) can be used to calculate the TER.

3.1 Area Under the ROC Curve (AUC) AUC gives a single merit score between 0.5 and 1, with 1 considered excellent and 0.5 considered poor. The various thresholds given can be averaged to produce the AUC. 1 A=

Pd p (x)d P f p (x),

(22)

0

In [18], the average threshold. ∞ A=−

Pd (x) 0

∂ P (x)

Here, ∂f px given as

is the partial derivatives of P f p (x). Using (13) and (19), ⎡

⎢ a ∂ P f p (x) = C1 ⎢ ⎣x ∂x

∞  σu2p

×

exp σu2p

−N 2



(23)

∂ P f p (x) ∂x

will be

  m1 p y Nx y (b−a−1) dy + ax (a−1) − 2 2 exp − 2y σs χ Δ1 p

0



∞

where

∂ P f p (x) dx. ∂x





⎥ m1 p y −N x y (b−a) dy ⎥ − 2 2 ⎦ 2y σs χ Δ1 p

0

  N −1 m1 p−1    m1 p 1 N a m 1 p σu2p m1 p 1 0  C1 =  exp σs2 χ 2 Δ1 p a=0 b=0 a! 2 [ m1p σs2 χ 2 Δ1 p .   (m1 p−b−1) m − 1  1p × −σu2p . 1 b

(24)

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

855

Using Eqs. (19), (20), (22), (23) along with (24), and after some simplification,  A=

C1 C2 ηqp−l+1 

σs2 χ 2 Δ1 p m1 p

b−a

⎡ ⎣N 2



∞ x 0

l+a

σu2p

Nx [ p − l + 1, ; ηq 2ηq



  2 2 m1 p N x σs χ Δ1 p d x − a σs2 χ 2 Δ1 p 2σ 2s χ 2 Δ1 p m1 p   ∞ σu2p N x l+a−1 1 × x [ p − l + 1, ; ηq 2ηq 0    m 1 p σu2p m1 p N x 0 ; ×[ b − a + 1, 2 2 dx σs χ Δ1 p 2σ 2s χ 2 Δ1 p ×[ b − a,

m 1 p σu2p



1

0

;

(25)

where  C2 =

m q r −1 N −1 2 l=0 q=1 r =1 p=0



1 Nx l! 2

l Ξ2 (q, r )

 r − 1  2 r − p−1  2  −σu p 1 σu p p 1 exp ηqr (r − 1)! ηq

4 Results and Discussion The simulation results are discussed in this section, considering the impact of CR node mobility and CSI error. Analysis and verification of the impact of estimate error, CR node mobility, and the number of CR users on FD-CR spectrum sensing are performed. For the simulation, the number of samples N = 10 [6], carrier frequency 2 = σtk2 = σn2 = 1 [6], error variance = 1.9 GHz [16], noise variance σt12 = σ1k 2 σ = 0.1, and SIS Factor χ = 0 < χ < 0.1 is taken. To evaluate the FD-CR sensing performance, this work considers the CSI error, CR node mobility, and primary channel as Nakagami-m distributed. Here, the assumption is CR continuously senses the PT channel and always has data to transmit. The ROC plot for different CSI conditions is shown in Fig. 2 at different CR transmit power levels σs2 /σn2 . When considering channel estimation error from σ 2 = 0 to σ 2 = 0.1, the Pd dramatically changes for different CR power (σs2 /σn2 = 10 dB, 15 dB and 20 dB) respectively. It is clear from Fig. 2 that the sensing performance degrades as σs2 /σn2 increases. From Fig. 2, Pd decreases from 0.5425 to 0.4922, at the (P f = 0.01, χ = 0.05, PT power σt2 /σn2 = 5 dB, σs2 /σn2 = 10dB, and CR speed v = 20 mph), with a change in CSI error (σ 2 = 0(perfect ) to σ 2 = 0.1(imperfect)). As a result, the FD’s sensing performance is greatly impacted by the imperfect reporting channel.

856

A. K. Rao et al.

Fig. 2 Probability of detection variation with σ 2 = 0 (perfect) to σ 2 = 0.1 (imperfect) CSI at different σs2 /σn2

Figure 2 compares the probability of detection for perfect and imperfect CSI conditions. It is also clear from the plot that when a channel is imperfect, the probability of detection suffers. The FD-CR performance is shown in Fig. 3 under various CR node mobilities conditions for imperfect CSI. The ROC curve shows that the detection probability decreases as CR node mobility increases. From Fig. 3, at (P f = 0.01, χ = 0.05, CR σs2 /σn2 = 15 dB σ 2 = 0.1 σs2 /σn2 = 4 dB) the Pd falls from 0.3909 to 0.1888 with increases in CR node velocity from v = 0 mph to v = 50 mph. As a result, the FD-spectrum CRN’s sensing performance is considerably worsened by CR node mobility. The AUC is depicted in Fig. 4. The AUC declines as CR node mobility increases, indicating that the FD-CR system’s performance is deteriorating. From Fig. 4 at (σt2 /σn2 = 10 dB, σs2 /σn2 = 10 dB), and CR speed (v = 0 mph to v = 40 mph) AUC goes from 0.9375 to 0.8875. The CR node speed and the CSI error considerably degrade the performance of the FD-CRN system. Figures 4 and 5 show the comparison between CR’s sensing performance for stationary and mobile CR. The performance metrics probability of detection and AUC are significantly affected by static CR conditions v = 0 mph to mobile CR v = 50 mph conditions. Therefore, we can see that CR node mobility significantly deteriorates the spectrum sensing performance. The TER for the number of CR users and their speeds, PT power, and CR powers are shown in Figs. 5 and 6. Due to CR’s strong mobility, TER performance achieves asymptotic levels even at high primary transmit power.

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

857

Fig. 3 Probability of detection at different CR speed for σ 2 = 0.1

Fig. 4 Area under the ROC curve for different CR speeds at σ 2 = 0.1

Figure 5 shows the TER performance for both perfect and imperfect CSI at various σs2 /σn2 values. The graph shows that when σt2 /σn2 increases, the error rate experiences an asymptotic floor. When σs2 /σn2 rises, and the error rate increases as well. From Fig. 5 at (σt2 /σn2 = 10 dB, v = 20 mph and χ = 0.1), As σs2 /σn2 rises from 10 to 20 dB, the TER rises from 0.121 to 0.216 for imperfect CSI and from 0.091 to 0.189 for perfect CSI.

858

A. K. Rao et al.

Fig. 5 TER versus PT power σt2 /σn2 for various CR power σs2 /σn2 at v = 20 mph

Fig. 6 TER versus threshold at the different number of cooperative CR node and their speed for σ 2 = 0.1

Full-Duplex Spectrum Sensing with Imperfect CSI and Mobile Nodes

859

The plot of the cooperative CR threshold versus the global total error rate is shown in Fig. 6. It is clear from the plot that for a fixed number of CR (M), TER drops with an increase in CR number while increasing with an increase in CR speed (M). The findings above show that the CR node mobility and channel estimate error considerably negatively impact the FD-CRN. When the reporting channel is considered perfect to imperfect, TER increases while the Pd and AUC decreases. Additionally, with increased CR node speed, the probability of detection and AUC decreased, and the error rate increased.

5 Conclusion This paper examines the impact of imperfect CSI and mobility of CR nodes on the sensing performance of FD-CRN. The probability of detection and false alarm expressions are obtained in closed form. Performance metrics ROC, AUC, and TER, are calculated. It is determined that a large number of CR nodes improves FD-CR sensing performance, but a higher CR node mobility degrades it. With channel estimate error, the FD-CR sensing performance also suffers. Therefore, when designing an FD-CRN, the impact of imperfect CSI and the CR node mobility are crucial factors to consider.

References 1. Nguyen, BC, Hoang TM, Kim T (2021) On performance of two-way full-duplex communication system with reconfigurable intelligent surface. IEEE Access 9:81274–81285 2. de Melo Guimarães L, Bordim JL (2021) A full-duplex mac technique to improve spectrumefficiency on 5g mobile wireless networks. Comput Commun 166:216–225. 3. Rao AK, Singh RK, Srivastava N (2020) Full-duplex wireless communication in cognitive radio networks: a survey. In: Advances in VLSI, communication, and signal processing. Springer, Singapore, pp 261–277 4. Boulogeorgos A-A, Chatzidiamantis ND, Karagiannidis GK (2016) Spectrum sensing with multiple primary users over fading channels. IEEE Commun Lett 20(7):1457–1460 5. Gahane L, Sharma PK, Varshney N, Tsiftsis TA, Kumar P (2017) An improved energy detector for mobile cognitive users over generalized fading channels. IEEE Trans Commun 66(2):534– 545 6. Rao AK, Sabat S, Srivastava N, Singh RK (2022) Mobile FD-CR with High-Speed VTFET CMOS SOI switch under channel estimation error. Silicon 1–12 7. Rawat DB, Alsabet R, Bajracharya C, Song M (2018) On the performance of cognitive internetof-vehicles with unlicensed user-mobility and licensed user-activity. Comput Netw 137:98–106 8. Rao AK, Sabat S, Singh RK, Srivastava N (2021) Cooperative spectrum sensing using mobile full-duplex cognitive radio and non-time-slotted primary user activity. Trans Electr Electron Mater 22(5):679–686 9. Min AW, Shin KG (2009) Impact of mobility on spectrum sensing in cognitive radio networks. In: Proceedings of the 2009 ACM workshop on cognitive radio networks, pp 13–18

860

A. K. Rao et al.

10. Rao AK, Singh S, Srivastava N, Singh RK (2022) Impact of CSI on Single channel In-band Full-duplex Radio implemented with 60 nm Vertical TFET-based CMOS structure. Silicon 1–11 11. Patel A, Ram H, Jagannatham AK, Varshney PK (2017) Robust cooperative spectrum sensing for MIMO cognitive radio networks under CSI uncertainty. IEEE Trans Signal Process 66(1):18–33 12. Khattabi YM, Matalgah MM (2015) Performance analysis of multiple-relay AF cooperative systems over Rayleigh time-selective fading channels with imperfect channel estimation. IEEE Trans Veh Technol 65(1):427–434 13. Khattabi YM, Matalgah MM (2017) Alamouti-OSTBC wireless cooperative networks with mobile nodes and imperfect CSI estimation. IEEE Trans Veh Technol 67(4):3447–3456 14. Simon MK, Alouini M-S (2001) Digital communication over fading channels. Wiley, New York 15. Zhao Y, Paul P, Xin CS, Song M (2014) Performance analysis of spectrum sensing with mobile SUs in cognitive radio networks. In: 2014 IEEE international conference on communications (ICC). IEEE, pp 2761–2766 16. Gahane L, Sharma PK (2017) Performance of improved energy detector with cognitive radio mobility and imperfect channel state information. IET Commun 11(12):1857–1863 17. Gradshteyn IS, Ryzhik IM (2007) Table of integrals, series, and products. In: Seventh, vol 48. Elsevier/Academic Press, Amsterdam, p. 1171 18. Atapattu S, Tellambura C, Jiang H (2010) Analysis of area under the ROC curve of energy detection. IEEE Trans Wireless Commun 9(3):1216–1225

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon Resonance Sensor with 2D Nanomaterial: MXene Rajeev Kumar, Maneesh Kumar Singh, Sarika Pal, and Alka Verma

Abstract We present a SPR sensor consisting of CaF2 prism, Au metal layer and monolayer of MXene (Ti3 C2 Tx ) for biomolecule sensing. The optimized SPR sensor configuration enhanced the sensitivity as well as other performance parameters such as FWHM, DA, and FoM. The theoretical analysis was carried out using the MATLAB and the COMSOL Multiphysics simulation software. The CaF2 prism is coated with metal layer for generation of surface plasmons (SPs), followed by MXene, for attachment of biochemical due to its larger surface area and (H, OH, F) functionalized groups. The simulation result shows that an enhanced maximum sensitivity of 292.55°/RIU for the variation in RIs of sensing medium (Δns = 0.005) at the 633 nm wavelength. Keywords Surface plasmon resonance sensor · CaF2 prism · Gold layer · MXene layer · Sensitivity · Penetration depth

1 Introduction Over the last four-decade, surface plasmon resonance (SPR) sensor among optical sensor category has been extensively favored for sensing of biomolecules and biochemicals [1]. The SPR biosensor has attracted the attention of researcher because R. Kumar (B) · S. Pal National Institute of Technology, Srinagar, Uttarakhand 246174, India e-mail: [email protected] S. Pal e-mail: [email protected] M. K. Singh Department of CSE, Dev Bhoomi Uttarakhand University, Dehradun, Uttarakhand 248007, India e-mail: [email protected] A. Verma Institute of Engineering and Rural Technology, Prayagraj, U.P. 211002, India e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_66

861

862

R. Kumar et al.

of their attractive sensing capability, compactness, robustness, high precision, and reliability. SPR sensing principle is based on refractive index (RI) sensing using attenuated total reflection (ATR) technique. Surface plasmon polaritons (SPPs) are electron density wave (collective oscillation of free electron gas density) that propagates along the interface of dielectric (positive permittivity) and metal (negative permittivity) [2, 3]. Surface plasmon wave (SPW) is generated by interaction between transverse magnetic (TM) polarized light incident from prism coupler in Kretschmann configuration and free electrons on the metal surface. SPR phenomenon is achieved when evanescent wave vector of transverse magnetic (TM) polarized incident light matches with wave vector of SPPs wave [4]. At resonance condition, electrons are strongly resonating at metal–dielectric interface, transverse wave exponentially fall off; result in a sharp dip in reflectivity measured from photo detector [5]. The angle of incidence at which the reflection curve shows the maximum dip (i.e. minimum reflectance value or Rmin .) is called an SPR resonance condition. Most of the SPR biosensors have need of a high sensitivity for proper detection of biomolecules. Therefore, enhancement of sensitivity of biosensor has become attractive research window for researchers to contribute their unique ideas. Various methods are proposed to improve the sensitivity of biosensors like colloidal gold nanoparticle, nanoslits, bimetallic SPR active metal, hybrid layer structure, and 2D material [6–9]. 2D materials have given the new dimension in the field of material science, optoelectronics, and biosensing. Over the decade, numbers of emerging 2D materials like graphene, black phosphorous (BP) [10], transition metal dichalcogenides (TMDCs) [6], metal oxides and antimonene [1] have been attracted noteworthy interest owing to their unique electrical and optical properties desired to enhance the sensitivity of biosensors [7]. Most of the 2D material has certain drawbacks such as weak biomolecules interaction or poor chemical stability. Recently explored 2D material, MXene has exhibited prospective claim for sensing of biochemical and biomolecule when used as biorecognition element (BRE) layer in SPR configuration [11, 12]. MXene has unique properties such as larger surface area, hydrophilic surface terminations, chemical and mechanical stability, smaller work function, strong carrier confinement, and higher binding energies for biomolecules [10, 13, 14]. MXene plasmonic property may get changed via altering their surface terminations [13]. Thus, they may be proficiently used in SPR sensors for biomolecule sensing with appropriate control of their surface termination. Wu et al. [15] have presented a SPR sensor with MXene to increase the sensitivity and maximum sensitivity 160°/RIU is obtained. They observed that MXene along with other metal such as Ag, Cu, and Al which shows the sensitivity comparison for proposed SPR sensor at a characteristics wavelength 633 nm. Srivastava et al. [16] theoretically presented MXene and BP along with TMDC material to enhance the sensitivity. They obtained the highest sensitivity 190.22°/RIU using single layer of each 2D material. Kumar et al. [17] proposed a SPR sensor based on MXene and silicon with maximum sensitivity 231°/RIU. Xu et al. [18] theoretically investigated of TMDCs and MXene-based SPR sensor with the structure of BK7/Au/TMDCs/Au/MXene at fixed wavelength 633 nm. The maximum sensitivity of 198°/RIU was obtained with Au/WS2 /Au/MXene.

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon …

863

In the proposed SPR sensor, 2D nanomaterial MXene as a BRE layer is coated over the Gold (Au) plasmonic metal, and CaF2 prism is used for coupling of incident light. The sensor performance is also compared for different metals like Ag, Cu, and Al also. Section 2 clarifies the theoretical model, design consideration, and performance parameter; Sect. 3 includes the results discussion, and finally, Sect. 4 concludes the work presented here.

2 Theoretical Model, Sensor Design, and performance parameters The proposed SPR sensor structure based on the MXene is shown in Fig. 1. The Kretschmann-based sensor configuration with CaF2 prism is used for the proposed SPR sensor, operating at characteristic wavelength of 633 nm. The CaF2 prism is used as coupling prism due to its low RI which helps in achieving high sensitivity. The 2D MXene layer is used as the biological recognition for attachment of biomolecules on sensor surface and for enhancing sensor sensitivity. The RI of CaF2 prism is 1.4329 [19]. The RI of SPR active metal (Au) can be easily computed using Drude–Lorentz model [20]. √ n m = εm =

/ 1−

λ2 λc λ2p (λc + iλ)

(1)

where λc and λp denote the collision and plasma wavelength, respectively, for the Au, Ag, Cu, and Al metal layers, and their values are given in Table 1. The RI of MXene is 2.38 + 1.33i [15, 20] with 0.993 nm monolayer thickness. The last layer is used as sensing medium for binding the biomolecules with RI of [1.33 to 1.335]. Fig. 1 Proposed SPR biosensor

864 Table 1 λc and λp of Au, Ag, Cu, and Al metal layers as per the Drude model at λ = 633 nm

R. Kumar et al. S. No.

Metals

Collision wavelength (λc )

Plasma wavelength (λp )

1

Gold (Au)

8.9342 × 10−6

1.6826 × 10−7

1.7614 ×

10−5

1.4541 × 10−7

10−5

1.3617 × 10−7 1.0657 × 10−7

2

Silver (Ag)

3

Copper (Cu)

4.0852 ×

4

Aluminum (Al)

2.4511 × 10−5

In mathematical modeling of proposed 4-layer SPR sensor configuration, the reflectance intensity (Rp ) of the p-polarized light is calculated using Fresnel reflection coefficient (Rp ) and transfer matrix method (TMM). And, for simulation of results of proposed SPR sensor, MATLAB software is used [15–21]. | |2 Rp = |rp |

(2)

Further, for evaluation of transverse field and penetration depth (PD), COMSOL simulation software is used which is based on finite element method (FEM). The performance parameters are calculated using the reflectance curve (plot of reflected intensity vs. incident angle) of the proposed SPR sensor. Sensitivity (S) is the ratio of the change in resonance angle (ΔθRes. ) and the change in RI of the sensing layer (Δn s ); it indicates the sensor’s sensing capability. S=

ΔθRes. ◦ ( /RIU) Δn s

(3)

Full Width at Half Maximum (FWHM) is difference of resonance angles at 50% reflection intensity, i.e., Rmin = 0.5 a.u. It signifies about the angular width of the resonance curve. Inverse of FWHM is called detection accuracy. DA =

1 (1%) FWHM

(4)

Figure of merit (FOM) indicates overall performance of the SPR sensor and calculated by the multiplication of the sensitivity and detection accuracy. FoM = S ∗ DA (1/RIU)

(5)

3 Results and analysis Metal layer in SPR sensing plays a very important role as it is responsible for generation of SPs. Optimization of metal layer thickness is necessary for maintaining the

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon …

865

balance between photon absorption efficiency and energy loss. Finally, the optimized metal thickness leads to achieve the minimum value of reflected intensity (i.e., Rmin ) which shows extreme interaction of evanescent field with the biomolecules contained in sensing medium. In this work, we have optimized thickness of the metal layer (Au, Ag, Cu, Al) in terms of maximum sensitivity and minimum Rmin . Figure 2a, b is plotted for optimization of metal layers thickness in terms of sensitivity and minimum Rmin considering monolayer MXene. The CaF2 prism is used as a coupling prism for enhanced sensitivity because it’s of low RI [19]. It is clearly indicated that the maximum sensitivity is obtained for Au layer because of its smaller ohmic losses and chemical stability [11]. From Fig. 2a, the variation in sensitivity is observed from 184.49°/RIU to 292.55°/RIU for Au-based SPR sensor, up to 38 nm Au thickness, and thereafter, it decreases to 59.25°/RIU. In other three Ag, Cu, and Al-based SPR configurations, the sensitivity is found to increase with increase in the thicknesses of Ag, Cu, and Al layer. It is observed that the sensitivity variation is from 164.55°/RIU to 251°/RIU, from 159.62°/RIU to 211.76°/RIU, from 144.72°/RIU to 152°/RIU on using Ag, Cu, and Al layer, respectively. Now, the variation of Rmin vs metal layer thickness is shown in Fig. 2b. It is noted that the Rmin firstly decreases up to certain metal thickness, and thereafter, it increases. The optimized thicknesses of metal layers corresponding to maximum sensitivity at minimum value of Rmin is given in Table 2.

Fig. 2 Variation in a sensitivity, b Min. reflectance versus thickness of Au, Ag, Cu, and Al metal layers

Table 2 Performance parameter evaluated for Au, Ag, Cu, and Al-based proposed sensor Metal

Thick.(nm)

Sensitivity (°/RIU)

Rmin . (a.u.)

DA (/°)

FoM (/RIU)

Ag

42

223.91

4.8 ×

Au

38

290.6

7.44 × 10–5

Cu

43

199.60

1.25 × 10–6

0.359

71.62

Al

33

149.19

3.44 × 10–4

0.617

92

10–4

0.22

49.26

0.11

31.96

866

R. Kumar et al.

Now, reflectance curves are plotted for Au, Ag, Cu, and Al-based proposed sensor at optimized thickness of metals layer and monolayer MXene in Fig. 3a– d. First, we studied performance for Ag-based SPR sensor (Structure: CaF2 /Ag (42 nm)/MXene/Sensing medium) as shown in Fig. 3a. We found the sharp reflectance curve, which show sensitivity, FWHM, DA, and FoM as 223.91°/RIU, 4.54°, 0.22/°, and 49.26/RIU, respectively, for the sensing medium RI range from 1.33 to 1.335. Narrower reflectance curve is due to smaller SPs damping for Ag metal. In the second case, we studied the performance of Au-based SPR sensor (Structure: CaF2 /Au (38 nm)/MXene/Sensing medium) as shown in Fig. 3b. Much higher sensitivity (290.60°/RIU) is obtained than previous case with compromise of FWHM (8.84°), DA (0.11/°) and FoM (31.96/RIU). In the third and fourth case, we replaced the Au layer to Cu and Al layer, respectively, for performance comparison. The performance parameters evaluated for Au, Ag, Cu, and Al metal layer-based proposed sensor are presented in Table 2 also. It may be analyzed from Table 2 that lesser sensitivity and better DA, FoM are obtained on use of Cu and Al metal layers due to weak generation and larger damping of SPs for Cu and Al metals.

Fig. 3 Reflectance curve for proposed sensor based on metal layers a Ag, b, Au, c Cu, and d Al, with monolayer MXene

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon …

867

To analyze how much effectively the biomolecules are attached on sensor surface which leads to RI shift of sensing medium, different performance parameters are studied corresponding to alteration in RI of the sensing medium. Figure 4a–d indicates the variation in sensitivity, Rmin, DA, and FoM for change in sensing medium RI from 1.330 to 1.335, respectively. The sensitivity enhancement for all three cases discussed above is observed from 2.13.14°/RIU to 223.91°/RIU, 278.46°/RIU to 290.60°/RIU, 192.52°/RIU to 199.50°/RIU, and 145.52°/RIU to 149.19°/RIU, respectively. It is noted that the maximum sensitivity (290.60°/RIU) is obtained for Au layer-based proposed sensor which is maximum of all three cases. Figure 4b shows the variation in Rmin vs variation in RIs of sensing medium for Au, Ag, Cu, and Al mediated proposed sensor. It is clearly observed that the Rmin increases with increase in the RI of sensing medium. The Rmin increment for Au mediated proposed sensor is much higher than for Ag, Cu and Al mediated sensor. Figure 4c–d demonstrates almost constant values of DA and FoM with respect to sensing medium RI variation for all four cases. However, maximum sensitivity is obtained with the compromise of DA and FoM for the Au-based proposed sensor. In Fig. 5a–d 1D plot for the transverse component of electric field along normal to the layer interface is given for Ag, Au, Cu, and Al mediated proposed sensor, respectively. The TM field variation was plotted using COMSOL Multiphysics software. The penetration depth is an important parameter, especially for sensing of biological molecule. The distance from prism to the interface at which the electric field amplitude decrease by a factor of 1/e from its maximum value is called the penetration

Fig. 4 Variation in a sensitivity, b Rmin . c DA, and d FoM versus RI change in sensing medium for Au, Ag, Cu, and Al mediated proposed sensor

868

R. Kumar et al.

depth (PD) [21]. We have calculated PD for all four cases. The maximum PD of 312.5 nm is obtained on use of Al layer at 33 nm thickness. The comparison of the PD for all four cases is given in Table 3. Figure 6a, b indicates that the 2D variation of field intensity for all four cases. It is clearly observed that larger evanescent field in sensing medium is observed for Al and Cu layer mediated proposed SPR sensor. This is due to lesser damping SPs on use of Al and Cu metal layers which leads to higher evanescent field in the sensing medium.

Fig. 5 1D plot for normal component of transverse electric field along normal to layer interface for a Ag, b Au, c Cu, and d Al layer mediated proposed sensor design

Table 3 PD of proposed SPR sensors with different plasmonic metal at optimized thickness

Sensor configuration

Penetration depth (nm)

CaF2 /Ag(42 nm)/MXene/sensing medium

224

CaF2 /Au(38 nm)/MXene/sensing medium

198

CaF2 /Cu (43 nm)/MXene/sensing medium

246

CaF2 /Al(33 nm)/MXene/sensing medium

312.5

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon …

869

Fig. 6 a Distribution of normalized electric field for proposed SPR sensor at the resonance angle and b distribution of the SPPs for proposed SPR sensor at the resonance angle

Finally, a comparative analysis of the proposed sensor with recently investigated MXene mediated SPR sensor is reported in Table 4. It clearly reflects much higher sensitivity for Au mediated proposed SPR sensor.

4 Conclusion In this work, the comparative analysis on using different metal layers for MXene and CaF2 prism-based SPR sensor is presented. The CaF2 prism is fluoride glass with low RIs. The MXene layer shows better adsorption of biomolecules when used

870 Table 4 Sensitivity analysis of the proposed sensor with MXene-based SPR sensor investigated recently

R. Kumar et al. Sensor structure

Sensitivity (º/RIU)

Prism/Au/MXene/sensing medium [15]

160

Prism/Au/MXene/WS2 /BP/sensing medium [16]

190.2

Prism/Ag layer/MXene/BlueP-MoS2 /BP/sensing medium [17]

231

Prism/WS2 /MXene/sensing medium [18]

198

Proposed-Prism-CaF2 /Au/MXene/sensing 292.55 medium

as BRE layer due to its unique optical and electrical sensing properties. The monolayer MXene was employed as BRE layer to increase the sensitivity. The highest sensitivity of the optimized configuration is 292.55°/RIU for Δns = 0.005, with the Au plasmonic metal. The sensor performance such as sensitivity, FWHM, DA, and FoM are compared for Au, Ag, Cu, and Al metal layers. As per results analysis, it is expected that the Au-based sensor (CaF2 /Au/MXene/SM) is suitable for the detection of biomolecule with reasonable sensitivity and other performance parameters.

References 1. Singh MK, Pal S, Prajapati YK. Design and analysis of an SPR sensor based on antimonene and platinum for the detection of formalin. IEEE Trans Nano Biosci. https://doi.org/10.1109/ TNB.2022.3159532. 2. Homola J, Yee SS, Gauglitz G (1999) Surface plasmon resonance sensors: review. Sens Actuators B Chem 54:3–15 3. Hasib MHH, Nur N, Rizal C, Shushama KN (2019) Improved transition metal dichalcogenidesbased surface plasmon resonance biosensors. Condensed Matter 4(2):49 4. Zhu J, Ke Y, Dai J, You Q, Wu L, Li J, Guo J, Xiang J, Dai X (2019) Topological insulator over-layer to enhance the sensitivity and detection limit of surface plasmon resonance sensor. Nanophotonics (2019) 5. Moznuzzaman M, Islam MR, Hossain MB, Mehedi IM (2020) Modeling of highly improved SPR sensor for formalin detection. Results Phys 16:102874 6. Singh MK, Pal S, Verma A, Prajapati YK, Saini JP (2020) Highly sensitive Antimonene coated Black Phosphorous based surface plasmon resonance biosensor for DNA hybridization: design and numerical analysis. J Nanophoton SPIE 14(4):046015. https://doi.org/10.1117/1.JNP.14. 046015 7. Chaudhary VS, Kumar D, Mishra GP, Sharma S, Kumar S (2022) Plasmonic biosensor with gold and titanium dioxide immobilized on photonic crystal fiber for blood composition detection. IEEE Sens J 22(9):8474–8481 8. Wang X, Sun X, Hu Y, Zhang L, Zeng L, Liu Q, Duan JA (2022) A dual-parameter optical fiber SPR sensor for simultaneous measurement of glucose and cholesterol concentrations. IEEE Sens J (2022) 9. Srivastava R, Prajapati YK, Pal S, Kumar S (2022) Micro-channel plasmon sensor based on a D-shaped photonic crystal fiber for malaria diagnosis with improved performance. IEEE Sens J 22(15):14834–14841

Sensitivity Enhancement of Kretschmann Configured Surface Plasmon …

871

10. Maurya JB, Prajapati YK, Raikwar S, Saini JP (2018) A silicon-black phosphorous based surface plasmon resonance sensor for the detection of NO2 gas. Optik 160:428–433 11. Gupta BD, Sharma AK (2005) Sensitivity evaluation of a multi-layered surface plasmon resonance-based fiber optic sensor: a theoretical study. Sens Actuators B Chem 107:40–46 12. Yu X, Yuan Y, Xiao B, Li Z, Qu J, Song J (2018) Flexible plasmonic pressure sensor based on layered two-dimensional heterostructures. J Lightwave Technol 36(23):5678–5684 13. Cherifi A, Bouhafs B (2017) Potential of SPR sensors based on multilayer interfaces with gold and LHM for biosensing applications. Photonic Sensors 7(3):199–205 14. Pal S, Verma A, Prajapati YK, Saini JP (2017) Influence of black phosphorous on performance of surface plasmon resonance biosensor. Opt Quant Electron 49(12):403 15. Wu L, You Q, Shan Y, Gan S, Zhao Y, Dai X, Xiang Y (2018) Few-layer Ti3 C2 Tx MXene: a promising surface plasmon resonance biosensing material to enhance the sensitivity. Sens Actuators B Chem 277:210–215 16. Srivastava A, Verma A, Das R, Prajapati YK (2020) A theoretical approach to improve the performance of SPR biosensor using MXene and black phosphorus. Optik 203:163430 17. Kumar R, Pal S, Prajapati YK, Saini JP (2021) Sensitivity enhancement of MXene based SPR sensor using silicon: theoretical analysis. SILICON 13(6):1887–1894 18. Xu Y, Ang YS, Wu L, Ang LK (2019) High sensitivity surface plasmon resonance sensor based on two-dimensional MXene and transition metal dichalcogenide: a theoretical study. Nanomaterials 9(2):165 19. Nisha A, Maheswari P, Subanya S, Anbarasan PM, Rajesh KB, Jaroszewicz Z (2021) Ag–Ni bimetallic film on CaF2 prism for high sensitive surface plasmon resonance sensor. Photonics Lett Poland 13(3):58–60 20. Kumar R, Pal S, Prajapati YK, Kumar S, Saini JP (2022) Sensitivity improvement of a MXene-immobilized SPR sensor with Ga-doped-ZnO for biomolecules detection. IEEE Sens J 22(7):6536–6543 21. Peterson AW, Halter M, Tona A, Plant AL (2014) High resolution surface plasmon resonance imaging for single cells. BMC Cell Biol 15(1):1–14

Metasurface-Based Tunable Radar Absorbing Structure for Broadband Applications Hrishit Mohan Das , Syed Tabassum Nazeer , Shrikrishan Baghel , Vineetha Joy , and Hema Singh

Abstract In this paper, two potential configurations of active tunable metasurfacebased radar absorbing structures (RAS) are presented for broadband applications. The proposed meta-atoms have a thin profile with a total thickness of 3.146 mm (approximately λ/10 at 10 GHz) and are composed of resistive patterns etched on RO4835 substrate. Active tuning has been accomplished via PIN diodes loaded at appropriate points on the metasurface. The number of diodes and their locations directly vary the impedance of the unit cell, and consequently, the absorptivity changes due to alteration in the overall quality factor. The proposed RAS models provide superior power absorption over a broad frequency range of 5.5–16.2 GHz (approximately covering C, X, and Ku bands) by adjusting the ON and OFF states of PIN diodes. They have excellent angular stability for oblique incident angles from 0° to 40° and are polarization independent as well due to the symmetric configuration. Further, an insight into the working principle of RAS is also included via an equivalent circuit model (ECM). Keywords Tunable radar absorbing structures · Metasurface · PIN diode · Absorption · Frequency reconfigurable

H. M. Das (B) Birla Institute of Technology and Science, Pilani, Rajassthan 333031, India e-mail: [email protected] S. T. Nazeer · S. Baghel · V. Joy · H. Singh Centre for Electromagnetics, CSIR-NAL, Bengaluru 560017, India e-mail: [email protected] H. Singh e-mail: [email protected] © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 R. K. Nagaria et al. (eds.), VLSI, Communication and Signal Processing, Lecture Notes in Electrical Engineering 1024, https://doi.org/10.1007/978-981-99-0973-5_67

873

874

H. M. Das et al.

1 Introduction Metasurfaces are periodic entities which consist of arrays of two-dimensional subwavelength unit cells arranged on a dielectric material [1]. They exhibit unusual electromagnetic properties which are not normally found in nature and can be used for controlling the amplitude, phase, and polarization of incident electromagnetic (EM) waves [2]. These properties can be advantageous for the design and development of metasurface-based radar absorbing structures (RAS) which can be used as a coating on any object of interest to absorb EM signals. They have several added advantages like ease of manufacturing, minimal weight penalty, etc. [3]. Thus, metasurfaces have been of immense research interest in the military and defense industry for implementing stealth technology in a cost effective and efficient manner. However, in the case of passive metasurfaces, their properties cannot be changed once manufactured. The geometry of the unit cell and material constitution/thickness have to be changed for getting different absorption characteristics [4]. For this reason, reconfigurable metasurfaces with active tuning capabilities are currently being explored. Active devices like PIN diodes provide modified internal resistance with change in bias current and hence can be used to change the characteristics of the metasurface like switching between perfect absorption and reflection modes [5], or for changing the frequency of operation [6]. Similarly, varactor diodes with their capability of having variable capacitance via modified bias voltages are also used for metasurface tuning [7]. However, they are difficult to tune as they are more susceptible to parasitic capacitances. Micro-electromechanical systems (MEMS)-based frequency tuning in X-band has been demonstrated in [8], but they require precision manufacturing capabilities. Non-uniform tuning with two different PIN diodes for simultaneous manipulation of magnitude as well as phase of reflection coefficient has been explored in [9]. Further, an intelligent metasurface with self-adaptive EM wave manipulation capability along with FPGA-based feedback mechanism to actively change bias current to the PIN diodes is described in [10]. Nevertheless, these models require complicated biasing circuitry. Recently, an active metamaterial absorber has been demonstrated, where PIN diodes [11] were utilized for getting an overall bandwidth from 0.78 GHz to 4.62 GHz. However, the absorber has very high thickness of 25 mm making it unsuitable for airborne applications [12]. An active FSS, with ultra-thin packaging, capable of dual-polarized shielding and transmission capabilities has been proposed in [13]. Further, a wideband tunable metamaterial absorber, loaded with both varactors and PIN diodes, for fine as well as coarse adjustment between S and C bands, i.e., band splicing has been proposed in [14]. In this regard, this paper presents a novel metasurface-based tunable RAS with an actively switchable bandwidth of operation by using PIN diodes. Two configurations of tunable resistive-sheet-based RAS with different number of PIN diodes are presented. Section 2 consists of detailed EM performance analysis of the active unit cell with subsequent subsections dealing with power absorption, surface current analysis, and equivalent circuit modeling (ECM). All the EM simulations have been carried out using full wave simulation software (CST Studio suite).

Metasurface-Based Tunable Radar Absorbing Structure for Broadband …

875

2 EM Performance Analysis of Active Unit Cell Based on PIN Diodes The proposed meta-atom has a thin profile with a total thickness of 3.146 mm (approximately λ/10 at 10 GHz) and is composed of resistive pattern etched on RO4835 substrate. The details of three different layers are given below: Layer#1: Resistive sheet (Rs = 100 Ω/sq; thickness (t res ) = 0.063 mm) Layer#2: RO4835 (Er = 3.48; tanδ = 0.0037; thickness (t sub ) = 3.048 mm) Layer#3: Perfect electric conductor (PEC) (thickness (t ground ) = 0.035 mm) Total thickness of RAS = 3.146 mm. Here, Rs and t res denote resistivity and thickness of resistive sheet, respectively. The optimized parameters for the unit cell are given below (annotated in Fig. 1): (i) a = 5.74 mm, (ii) b = 2.58 mm, (iii) c = 3.60 mm, (iv) d1 = 2.60 mm, (v) d2 = 2.20 mm, (vi) e = 2.7 mm, (vii) r = 3.41 mm, (viii) dw = 0.50 mm, (ix) dx = 0.10 mm, (x) dy = 0.54 mm, (xi) β = 16.26°, (xii) λ = 90°, and (xiii) p = 17.00 mm. The original design without PIN diodes is shown in Fig. 1. Two different active configurations with four and eight PIN diodes per unit cell are shown in Fig. 2a, b respectively. The equivalent circuits corresponding to the PIN diode in ON and OFF states are shown in Fig. 2c, d, respectively [5].

2.1 Power Absorption Characteristics

p

The primary objective here is to have a compact RAS model with the reflection coefficient less than −10 dB (i.e., greater than 90% power absorption) over the

(a)

(b)

Layer#1

Layer#2

Layer#3

Fig. 1 Unit cell configuration for Design#1 (without diodes). a Top view and b 3D view

876

H. M. Das et al.

Fig. 2 Unit cell configurations with diodes. a Design#2 (with 4 PIN diodes), b Design#3 (with 8 PIN diodes), c ECM of PIN diode in ON state, and d ECM of PIN diode in OFF state

desired frequency range (2–18 GHz). A comparative analysis of the power absorption characteristics for the unit cells with and without diodes is presented in this section. There are three RAS configurations: Design#1 (without diode), Design#2 (4 diodes in ON state), and Design#3 (8 diodes in ON state). The reflection coefficient of the designs is illustrated in Fig. 3. It can be observed that the absorption bandwidths achieved are 7 GHz (6–13.1 GHz), 8.5 GHz (6–14.5 GHz), and 10.7 GHz (5.5– 16.2 GHz) for Design#1, Design#2 (ON state), and Design#3 (ON state), respectively. It is apparent that the introduction of PIN diodes in the unit cell designs changed the bandwidth of operation by almost 1.5 GHz and 4 GHz, respectively, albeit with some reduction in power absorption but still well within the desired limits. Further, the frequency dependent variation in the phase of the reflection coefficient can be observed as approximately similar for all the three designs from 2 to 6 GHz and from 16 to 18 GHz. However, slight shifts can be observed in the frequency range from 6 to 16 GHz. It can be noted that the phase varies from +180° to −180° hence crossing 0° between 5.5 and 16.2 GHz in all the three cases. This allows the proposed RAS to show maximum absorption performance in this frequency range. Furthermore, percentage of power absorbed for TE polarized normally incident EM waves are compared for Designs #2 and #3 for ON and OFF states in Fig. 4. It can be seen that there is a drastic shift in operational bandwidth, reduced to 6 GHz for Design #2 and 5 GHz for Design #3. Explanation for the same has been explored in Sect. 2.3. Further to establish the performance stability of the proposed models, the absorption characteristics have been evaluated for oblique angles of incidence up to 60°

877

0

180

-5

120

-10

60

-15

0

-20

-60

Mag. (D#1- W/o diodes) Mag. (D#2- 4 diodes) Mag. (D#3- 8 diodes) Phase (D#1- W/o diodes) Phase (D#2- 4 diodes) Phase (D#3- 8 diodes)

-25

-120

-30 2

4

6

8

10

12

14

16

Phase (Deg.)

Reflection Coefficient (dB)

Metasurface-Based Tunable Radar Absorbing Structure for Broadband …

18

-180 22

20

Frequency (GHz)

Fig. 3 Reflection coefficient of designed RAS configurations (Design#1(w/o diode), Design #2 & #3 in ON state) under normally incident TE plane wave

100 90

Power Absorbed (%)

80 70 60 50 40 30

Design #2 (ON state) Design #3 (ON state) Design #2 (OFF state) Design #2 (OFF state)

20 10 0 2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

Frequency (GHz)

Fig. 4 Comparison of power absorbed by Design #2 and #3 for TE polarized normally incident plane wave

for both TE as well as TM polarizations for all three designs. Figure 5 depicts the angular stability for TE and TM polarization for Design#1. For TM polarization in Fig. 5, the percentage of power absorbed is greater than 90% in the frequency range of 6–13 GHz for Design#1, whereas for TE, it is 6–12 GHz till 40°.

878

H. M. Das et al.

Fig. 5 Power absorbed (in %) at oblique incidence for Design#1. a TE, b TM

Moreover, for Design #2 and #3, the characteristics have been compared for both ON and OFF states in Figs. 6 and 7, respectively. In both the designs, for TE polarized wave, greater than 90% absorption is achieved till 40° angle of incidence. On the other hand, for TM polarization, the results are considerably better for angles of incidence up to 60°. Notably for TE polarization, as one simulates for higher angles of incidence, the absorption decreases but for TM it is the opposite, RAS shows highest absorption bandwidth for 60°. Further on comparing ON and OFF state, there is obvious increase in bandwidth for both polarizations with negligible change

Metasurface-Based Tunable Radar Absorbing Structure for Broadband … 100

100

ON

90

80

70

Power Absorbed (%)

Power Absorbed (%)

OFF

90

80

60 50 40 30 20

70 60 50 40 30 20

10

10

0 2

4

6

8

10

12

14

16

18

20

0

22

2

4

6

8

(a)

Frequency (GHz)

100

90

90

80

80

70 60 50 40 30 20

10

12

14

16

18

20

22

16

18

20

22

Frequency (GHz)

100

Power Absorbed (%)

Power Absorbed (%)

879

70 60 50 40 30 20

ON

10 2

4

OFF

10

0 6

8

10

12

14

Frequency (GHz)

16

18

20

0

22

2

(b)

4

6

8

10

12

14

Frequency (GHz)

Fig. 6 Power absorption characteristics (in %) at oblique angles of incidence for Design #2 for ON (left) and OFF (right) state. a TE polarization and b TM polarization

in absorption percentage. Thus, tuning of RAS has been achieved with different bandwidth for different biasing of the diodes. It is clear from the results that there is a trade-off between bandwidth and absorptivity. With greater number of diodes switched on, absorption bandwidth is higher but at the cost of performance at higher angles of incidence (>50°), especially for TE polarization. Bandwidth for absorption greater than 90% at normal incidence is summarized in Table 1. The polarization independence of the RAS performance is evident. For better visualization of power absorption characteristics w.r.t. frequency and angle of incidence, contour plots are shown in Fig. 8 for both TE and TM polarization. It can be seen that as the number of diodes are increased, the absorption bandwidth increases.

2.2 Surface Current Distribution Over Unit Cells Analyzing the surface current and electric field distribution over the unit cell is a vital step for developing the equivalent circuit model as it helps to understand which part of the pattern acts as a capacitor or an inductor. Figure 9 presents the surface

H. M. Das et al.

100

100

90

90

Power Absorbed (%)

Power Absorbed (%)

880

80 70 60 50 40 30 20

80 70 60 50 40 30 20

ON

10

2

4

6

8

10

12

14

16

18

20

2

6

8

100

90

90

80 70 60 50 40 30

10

12

14

16

18

20

22

18

20

22

Frequency (GHz)

100

80 70 60 50 40 30 20

ON

10

4

(a)

Power Absorbed (%)

Power Absorbed (%)

0

22

Frequency (GHz)

20

OFF

10

0

OFF

10

0

0 2

4

6

8

10

12

14

16

18

20

22

Frequency (GHz)

2

4

6

8

10

12

14

16

Frequency (GHz)

(b)

Fig. 7 Power absorption characteristics (in %) at oblique angles of incidence for Design #3 for ON (left) and OFF (right) state. a TE polarization and b TM polarization

Table 1 Bandwidth for power absorption at normal incidence in the proposed RAS designs Design #2

Design #3

ON state (GHz)

OFF state (GHz)

ON state (GHz)

OFF state (GHz)

TE polarized

6.2–14.2

5.9–12

5.5–16.2

5.7–10.3

TM polarized

6.1–14.4

5.9–11.8

5.5–16.5

5.7–10.3

current and electric field distribution over the three proposed designs. It is clear that the diodes in ON state direct the current from sides of the outer circle to the main pattern in the middle. Higher number of diodes enhances this effect and reduces the capacitive effect of the shape. Here, the diode acts as a series combination of resistor and inductor, thus allowing current to flow through it. The electric field visibly accumulates at the outer circles with increase in the number of diodes.

Metasurface-Based Tunable Radar Absorbing Structure for Broadband …

881

Fig. 8 Contour plots showing variation in power absorbed w.r.t. both frequency and angle of incidence. a Design#1 (without diode), b Design#2 (4 diodes: ON), and c Design#3 (8 diodes: ON)

2.3 Equivalent Circuit Model (ECM) for Unit Cells In order to understand the electrical behavior of the unit cells, equivalent circuit models have been developed for all the configurations. ECM represents the unit cell

882

H. M. Das et al.

Fig. 9 Surface current (left hand side) and E-field distribution (right hand side) over the unit cells. a Design #1, b Design #2, (ON), and c Design #3 (ON)

in terms of simple circuit elements like resistors, capacitors, and inductors making it easier to analyze and visualize the design physically. On comparing the circuits in Fig. 10, one can observe that L1, C1, L2, and C2 in the ECM corresponding to Design#1 (without diode) and Design#2 (four diodes) are same. Accordingly, the first and second resonant dips are at the same frequencies (8.1 GHz and 12.5 GHz, respectively). The values agree theoretically as well as per the following formula for resonant frequency:

Metasurface-Based Tunable Radar Absorbing Structure for Broadband …

ω = 2π f = √

1 LC

883

(1)

For Design#2, due to slight shift in the resonant dips, the values of inductors and capacitors are different. The lowest magnitude of reflection coefficient for all three designs have been found to be governed by R4. It is −30, −20, and −14 dB for Design#1, Design#2, and Design#3, respectively. With increase in the number of

Fig. 10 Validation of the simulated reflection coefficient obtained from CST and ECM. a Design#1, b Design #2 (ON), c Design #3 (ON)

884

H. M. Das et al.

Reflection Coefficient (dB)

0

-5

-10

-15 Design #2 ON Design #3 ON Design#2 OFF Design#3 OFF

-20

-25 2

4

6

8

10

12

14

16

18

Frequency (GHz) Fig. 11 Variation in reflection characteristics with ON and OFF states of PIN diodes

diodes, the overall resistance decreases as all the diodes are effectively connected in parallel to each other. Moreover, R2 and R3 control the individual resonant dips and thus have a decrement in values with increase in number of diodes due to lower effective resistance. In Fig. 11, the reflection characteristics are presented for Design#2 and Design#3 with all diodes in ON/OFF states. The bandwidth has changed in both the cases as follows: Design#2: BW in ON state = 6–14.5 GHz; BW in OFF state = 5.9–12.5 GHz Design#3: BW in ON state = 5.5–16.2 GHz; BW in OFF state = 5.7–10.3 GHz The performance of the proposed resistive-sheet-based RAS models has been compared with earlier works reported in open domain, and the summary is presented in Table 2.

Metasurface-Based Tunable Radar Absorbing Structure for Broadband …

885

Table 2 Comparison of proposed RAS with other tunable absorbers S. No.

References

Thickness of RAS (nm)

Operating freq. regime (GHz)

Absorption BW (OFF state) (GHz)

Absorption BW (ON state) (GHz)

1

[12]

25.5

0.5–5



0.78–4.6

2

[15]

2

2–4

2.5–2.6

2.9–3

3

[16]

5.4

6–22



7–22

4

[17]

34.572

0.5–6

0.85–1.88

2.66–5.23

5

[18]

1

3–7

4–4.5 and 6–6.5

4.8–5.5

6

Design#2 (Proposed)

3.146

8–16

5.9–12.5

6–14.5

7

Design#3 (Proposed)

3.146

8–16

5.7–10.3

5.5–16.2

3 Conclusion Tunable absorbers catering to different frequency regions of operation are extensively used on stealth platforms. In this paper, two such models of metasurface-based RAS loaded with PIN diodes have been presented. Extensive parameter optimization w.r.t. dimensions, number of diodes, and location of diodes has been carried out in full wave simulation software to achieve greater than 90% power absorption in the widest possible bandwidths. The optimized designs have been found to exhibit superior power absorption characteristics over a broad frequency range of 5 GHz to 16.5 GHz (approximately covering C, X, and Ku bands) by adjusting the ON and OFF states of PIN diodes. A comparative analysis of the designed resistive-sheet-based RAS models with four and eight diodes has been done. Substantial bandwidth increase of 21.42% and 57.14%, respectively, has been achieved in ON state with negligible loss of power absorption capability and minimum bias complexity. The models also showed stable power absorption performance for oblique incident angles from 0° to 40° for both TE as well as TM polarization. Further, equivalent circuit models developed for all the configurations provide in-depth insight into the underlying physics. Hence, the proposed actively tunable RAS configurations can be used to switch between different absorption bands on low observable platforms subjected to radars operating in different frequency domains. They can be employed for usage in smart radomes as well.

References 1. Munk BA (2000) Frequency selective surfaces: theory and design. Wiley, New York 2. Wu T (2005) Ch. Frequency selective surfaces. In: Chang K (ed) Encyclopedia of RF and microwave engineering, 1st ed. Wiley, New York (2005)

886

H. M. Das et al.

3. Joy V, Dileep A, Abhilash PV, Nair RU, Singh H (2021) Metasurfaces for stealth applications: a comprehensive review. J Electron Mater 50:3129–3148 4. Shao L, Zhu W (2021) Electrically reconfigurable microwave metasurfaces with active lumped elements: a mini review. Front Mat 8:7p 5. Song X, Yang W, Qu K, Bai X, Chen K, Feng Y, Zhu W (2021) Switchable metasurface for nearly perfect reflection, transmission, and absorption using PIN diodes. Opt Express 29:29320–29328 6. Ghosh S, Srivastava K (2016) Polarisation-independent switchable absorber/reflector. Electron Lett 52:1141–1143 7. Qi R, Zhai H, Yang D, Xue K (2020) An angular-stable multi-layer reconfigurable frequency selective surface based on varactor with wide tuning range. Int J RF Microw Comput Aided Eng 30(1):8 p 8. Safari M, Shafai C, Shafai L (2015) X-Band tunable frequency selective surface using MEMS capacitive loads. IEEE Trans Ant Propag 63(3):1014–1021 9. Liao J, Guo S, Yuan L, Ji C, Huang C, Luo X (2022) Independent manipulation of reflection amplitude and phase by a single-layer reconfigurable metasurface. Adv Opt Mater 10:2101551 10. She Y, Ji C, Huang C, Zhang Z, Liao J, Wang J, Luo X (2022) Intelligent reconfigurable metasurface for self-adaptively electromagnetic functionality switching. Photon Res 10:769– 776 11. Helszajn J (1978) Passive and active microwave circuits. Wiley, New York 12. Wu Z, Zhao J, Chen K, Feng Y (2022) An active metamaterial absorber with ultrawideband continuous tunability. IEEE Access 10:25290–25295 13. Zhao Y, Fu J, Wang Z, Chen W, Lv B, Zhang Q (2022) Design of a broadband switchable active frequency selective surfaces based on modified diode model. IEEE Ant Wirel Propag Lett 21(7):1378–1382 14. Wu T, Li W, Chen S, Guan J (2020) Wideband frequency tunable metamaterial absorber by splicing multiple tuning ranges. Res Phys 20:7 p 15. Zhu B, Huang C, Feng Y, Zhao J, Jiang T (2010) Dual band switchable metamaterial electromagnetic absorber. Prog Electromag Res B 24:121–129 16. Fang J, Huang J, Gou Y, Shang Y (2020) Research on broadband tunable metamaterial absorber based on PIN diode. Optik Int J Light Electron Opt 200:163171 17. Zhang Q, Shen Z, Wang J, Lee KS (2012) Design of a switchable microwave absorber. IEEE Ant Wirel Propag Lett 11:1158–1161 18. Zeng X, Zhang L, Wan G, Gao M (2017) Active metamaterial absorber with controllable polarisation and frequency. Electron Lett 53:1085–1086