Machine Learning, Deep Learning and Computational Intelligence for Wireless Communication: Proceedings of MDCWC 2020 (Lecture Notes in Electrical Engineering, 749)

Description

This book is a collection of best selected research papers presented at the Conference on Machine Learning, Deep Learning and Computational Intelligence for Wireless Communication (MDCWC 2020) held during October 22nd to 24th 2020, at the Department of Electronics and Communication Engineering, National Institute of Technology Tiruchirappalli, India. The presented papers are grouped under the following topics (a) Machine Learning, Deep learning and Computational intelligence algorithms (b)Wireless communication systems and (c) Mobile data applications and are included in the book. The topics include  the latest research and results in the areas of network prediction, traffic classification, call detail record mining, mobile health care, mobile pattern recognition, natural language processing, automatic speech processing, mobility analysis, indoor localization, wireless sensor networks (WSN), energy minimization, routing, scheduling, resource allocation, multiple access, power control, malware detection, cyber security, flooding attacks detection, mobile apps sniffing, MIMO detection, signal detection in MIMO-OFDM, modulation recognition, channel estimation, MIMO nonlinear equalization, super-resolution channel and direction-of-arrival estimation. The book is a rich reference material for academia and industry. Preface Organization Contents About the Editor Machine Learning, Deep Learning and Computational Intelligence Algorithms Deep Learning to Predict the Number of Antennas in a Massive MIMO Setup Based on Channel Characteristics 1 Introduction 2 Contributions of the Paper 3 Mutual Orthogonality 3.1 Trend Analysis 3.2 Deep Learning Architecture to Predict Number of Antennas Required for Orthogonality 4 Perfect CSI 4.1 Signal-to-Interference-Noise-Ratio (SINR) Analysis 4.2 Deep Learning Architecture to Predict Number of Antennas Required for Convergence of SINR 5 Imperfect CSI: An SINR Analysis 6 Results 6.1 Mutual Orthogonality Simulation Data 6.2 Predicting Number of Base Station Antennas Required for Orthogonality 6.3 Perfect CSI-SINR Convergence Simulation Data 6.4 Perfect CSI-Predicting Number of Base Station Antennas Required for Convergence of SINR 6.5 Imperfect CSI—Analysing the Number of Antennas Required for Convergence of SINR 7 Conclusions References Optimal Design of Fractional Order PID Controller for AVR System Using Black Widow Optimization (BWO) Algorithm 1 Introduction 2 Overview of Automatic Voltage Regulator (AVR) System 3 Fractional Calculus and Fractional Order Controllers 3.1 Fractional Calculus 3.2 Fractional Order Controller 4 Black Widow Optimization 4.1 Initial Population 4.2 Procreate 4.3 Cannibalism 4.4 Mutation 5 Proposed BWO-FOPID Controller 6 Results and Discussions 6.1 Step Response 6.2 Robust Analysis 7 Conclusion References LSTM Network for Hotspot Prediction in Traffic Density of Cellular Network 1 Introduction 2 Dataset Collection and Representation 3 Hotspot Prediction Using LLR Method 3.1 Algorithm to Find Hotspot Using LLR 3.2 LSTM Architecture to Predict Future Hotspot Using LLR 4 Hotspot Prediction Using CDF Method 4.1 Algorithm to Find Hotspot Using Cumulative Distribution Function 4.2 LSTM Architecture to Predict Future Hotspot Using CDF 5 Results 6 Conclusions References Generative Adversarial Network and Reinforcement Learning to Estimate Channel Coefficients 1 Introduction 2 Contributions of the Paper 3 Signal Source Separation 3.1 Using Generative Adversarial Networks and Reinforcement Learning 3.2 Generative Adversarial Networks for Data Space Distribution Modelling 3.3 Reinforcement Learning-Based Sampling Technique for Signal Estimation 4 Results 4.1 Extraction of Distributions Using GAN 4.2 Estimating True Values of Channel Coefficients 5 Conclusions References Novel Method of Self-interference Cancelation in Full-Duplex Radios for 5G Wireless Technology Using Neural Networks 1 Introduction 2 Signal Modeling 3 Solutions for Self-Interference (SI) Cancelation 3.1 Outline of Hybrid SI Cancelation 3.2 Proposed Solution for Implementing Digital Cancelation Using Neural Networks 4 Results and Discussions 5 Conclusions References Dimensionality Reduction of KDD-99 Using Self-perpetuating Algorithm 1 Introduction 1.1 Feature Selection Methods 2 Related Work 3 Proposed Work 3.1 Basic Idea Behind Self-perpetuating Algorithm 3.2 The Proposed Algorithm 4 Experimental Setup 5 Conclusion and Future Work Reference Energy-Efficient Neighbor Discovery Using Bacterial Foraging Optimization (BFO) Algorithm for Directional Wireless Sensor Networks 1 Introduction 1.1 Problem Identification 2 Related Works 3 Energy-Efficient Neighbor Discovery Using BFOA 3.1 Overview 3.2 Fundamentals of Optimization Algorithm 3.3 Estimation of Metrics 3.4 Energy-Efficient Neighbor Discovery 4 Simulation Results 4.1 Simulation Setup 4.2 Simulation Results and Analysis 5 Conclusion References Auto-encoder—LSTM-Based Outlier Detection Method for WSNs 1 Introduction 2 Related Literature Work 3 Proposed Model 3.1 Auto-encoder Preliminaries 3.2 BLSTM-RNN Preliminaries 4 Experimental Results 5 Conclusion References An Improved Swarm Optimization Algorithm-Based Harmonics Estimation and Optimal Switching Angle Identification 1 Introduction 2 Harmonic Estimation and Switching Angles Identification 3 Improved Particle Swarm Optimization Algorithm 4 Simulation Results 5 Conclusions References A Study on Ensemble Methods for Classification 1 Introduction 2 Related Work 3 Ensemble Learning Approaches 3.1 Bagging 3.2 Boosting 3.3 Stacking 3.4 Random Forest 4 Application of Ensemble Techniques 5 Deep Learning and Ensemble Techniques 6 Experimentation 7 Conclusion References An Improved Particle Swarm Optimization-Based System Identification 1 Introduction 2 Problem Formulation 3 Improved Particle Swarm Optimization Algorithm 4 Simulation Results 5 Conclusions References Channel Coverage Identification Conditions for Massive MIMO Millimeter Wave at 28 and 39 GHz Using Fine K-Nearest Neighbor Machine Learning Algorithm 1 Introduction 2 Network Architecture 3 Simulation Methodology 4 Simulation Measurements 5 Pathloss 6 Power Delay Profile 7 Fine-KNN 8 Conclusion References Flip Flop Neural Networks: Modelling Memory for Efficient Forecasting 1 Introduction 2 Previous Work 2.1 Long Short-Term Memory (LSTM) 3 Model Architecture 4 Experiments 4.1 Household Power Consumption 4.2 Flight Passenger Prediction 4.3 Stock Price Prediction 4.4 Indoor Movement Classification 5 Conclusion References Wireless Communication Systems Selection Relay-Based RF-VLC Underwater Communication System 1 Introduction 1.1 Paper Structure 2 Proposed System Model 2.1 Source-Relay (s-r) Hop 2.2 Relay-Destination (r-d) Hop 2.3 Underwater Attenuation Coefficient Model 2.4 Water Turbidity Channel Modeling 2.5 Pointing Error in Underwater VLC Link 3 BER Performance of the System 4 Numerical Results 5 Conclusion References Circular Polarized Octal Band CPW-Fed Antenna Using Theory of Characteristic Mode for Wireless Communication Applications 1 Introduction 2 Theory of Characteristics Modes Analysis (TCMs) 3 Antenna Design Procedure 4 Experimental Results and Discussion 5 Conclusion References Massive MIMO Pre-coders for Cognitive Radio Network Performance Improvement: A Technological Survey 1 Introduction 2 System Background 2.1 Cognitive Radio Network (CRN) 2.2 MIMO System 2.3 CRN MIMO 2.4 Massive MIMO Systems 3 Pre-coding in Massive MIMO 3.1 Linear Pre-coding Techniques 3.2 Nonlinear Pre-coding Techniques 3.3 Constant Envelope Pre-coding Techniques 3.4 Pre-coding in MIMO CRN 4 Conclusion Reference Design of MIMO Antenna Using Circular Split Ring Slot Defected Ground Structure for ISM Band Applications 1 Introduction 2 Antenna Design 3 Results and Discussions 4 Conclusion References Performance Comparison of Arduino IDE and Runlinc IDE for Promotion of IoT STEM AI in Education Process 1 Introduction 2 Significance of IoT and AI in Education: Opportunities and Challenges 3 System Model: Smart Home 4 Results and Discussion 5 Microcontrollers 5.1 Arduino UNO and Arduino IDE 5.2 STEMSEL and Runlinc IDE 6 Implementation 7 Conclusion and Future Works References Analysis of Small Loop Antenna Using Numerical EM Technique 1 Introduction 2 Loop Antennas 3 Finite Element Mesh 4 Methodology 5 Antennas-Gradient Methods 6 Conclusion and Discussion References A Monopole Octagonal Sierpinski Carpet Antenna with Defective Ground Structure for SWB Applications 1 Introduction 2 Antenna Design and Analysis 2.1 Methodology 2.2 Configuration 2.3 Operating Principle 3 Results and Discussion 4 Conclusion References DFT Spread C-DSLM for Low PAPR FBMC with OQAM Systems 1 Introduction 2 System Model 2.1 FBMC with OQAM System Model 2.2 Overlapping Structured FBMC with OQAM Signals 2.3 The PAPR Description in FBMC with OQAM System 3 Selective Mapping (SLM) Scheme 3.1 Conventional SLM Scheme 3.2 The SLM with Converse Vectors (C-SLM) 3.3 Conversion Vector-Based Dispersive SLM (C-DSLM) Schemes with FBMC with OQAM 4 Proposed DFT Spread C-DSLM for FBMC—OQAM System 4.1 DFT Spreading 4.2 The Conversion Vectors and Its Design 4.3 Proposed DFT Spread Converse Vectors with DSLM Method (C-DSLM) 4.4 Analysis of Computational Complexity 5 Performance Evaluation 5.1 Simulation Environment 5.2 Calculation of Computational Complexity 5.3 Transmission and Reception of DFT Spread C-DSLM Scheme 6 Conclusion References Secure, Efficient, Lightweight Authentication in Wireless Sensor Networks 1 Introduction 2 Authentication in WSN 3 Related Work 4 Proposed Authentication Mechanism 4.1 Simple Cluster Head (CH) Development 4.2 Proposed Protocol 5 AVISPA Tool Simulation Results, Security Analysis, and BAN Logic 5.1 BAN Logic Analysis: Logical Rules of BAN Logic 6 Conclusion References Performance Evaluation of Logic Gates Using Magnetic Tunnel Junction 1 Introduction 2 Magnetic Tunnel Junction 3 Proposed Approach 4 Simulation of Logic Circuits and Comparison 5 Conclusion References Medical IoT—Automatic Medical Dispensing Machine 1 Introduction 2 Challenges Faced During COVID-19 Situation 3 Need for Automatic Medical Dispensing Machine 4 Existing Solutions 4.1 Semi-automated Dispensing Machine Using Barcode 4.2 Drug Data Transfer System 4.3 Computerized Physician Order Entry (CPOE) for Neonatal Ward 5 Proposed Solution 5.1 Modules in Our Solution 6 Conclusion and Future Scope References Performance Analysis of Digital Modulation Formats in FSO 1 Introduction 2 Implementation of Modulation 3 Digital Schemes 4 Differential Phase-Shift Keying 4.1 DPSK System 4.2 DPSK Optical System 5 Offset Quadrature Phase-Shift Keying 5.1 OQPSK System 5.2 OQPSK Optical System 6 Results and Discussions 7 Conclusion References High-Level Synthesis of Cellular Automata–Belousov Zhabotinsky Reaction in FPGA 1 Introduction 1.1 Cellular Automata 1.2 High-Level Synthesis 2 Belousov Zhabotinsky Reaction 2.1 Simplified Reaction Mechanism 2.2 Reaction Surface 3 Programming the Automata in FPGA 3.1 Optimizations Specific to FPGA 4 Results 4.1 Simulation 4.2 Synthesis 5 Conclusions Reference IoT-Based Calling Bell 1 Introduction 2 Proposed Method 3 Arduino Specific Instructions 4 Screens 5 System Test 6 Conclusion References Mobile Data Applications Development of an Ensemble Gradient Boosting Algorithm for Generating Alerts About Impending Soil Movements 1 Introduction 2 Background 3 Methodology 3.1 Data 3.2 Measures for Evaluating ML Algorithms 3.3 Different Algorithms Used for Classification 3.4 Model Calibration 4 Results 5 Discussion and Conclusion References Seam Carving Detection and Localization Using Two-Stage Deep Neural Networks 1 Introduction 2 Related Work 3 Seam Carving and Seam Insertion 4 Detection of Seam Carving 5 Experiments 5.1 Experimental Setup 5.2 Learning 5.3 Detection Heatmaps 5.4 Robustness to Percentage of Seams Removed 5.5 Robustness to JPEG Compression 5.6 Explainability on Object Removed Images 5.7 Extension to Seam Insertion Detection 6 Conclusion and Future Work References A Machine Learning-Based Approach to Password Authentication Using Keystroke Biometrics 1 Introduction 2 Keystroke Dynamics 2.1 Types of Authentication Systems 2.2 Timing Features 2.3 Evaluation Parameters 3 Models Used 3.1 Support Vector Machines (SVM) 3.2 Random Forest Algorithm (RF) 3.3 Artificial Neural Network (ANN) 4 Dataset 5 Results 6 Conclusion References Attention-Based SRGAN for Super Resolution of Satellite Images 1 Introduction 1.1 Deep Learning for Super Resolution 1.2 Generative Adversarial Network-Based Deep Learning for SR 1.3 Motivation and Contribution 2 Attention-Based SRGAN Model 2.1 Network Structure 2.2 Generator 2.3 Discriminator 2.4 Loss Function 3 Results and Discussion 4 Conclusion References Detection of Acute Lymphoblastic Leukemia Using Machine Learning Techniques 1 Introduction 2 Datasets 3 Proposed Method 4 Result and Discussion 5 Conclusion References Computer-Aided Classifier for Identification of Renal Cystic Abnormalities Using Bosniak Classification 1 Introduction 2 Methodology 2.1 Pre-processing 2.2 Kidney Segmentation 2.3 Feature Extraction 2.4 Classification 3 Results and Discussion 3.1 Database 3.2 Experiment Setup 4 Conclusions References Recognition of Obscure Objects Using Super Resolution-Based Generative Adversarial Networks 1 Introduction 1.1 Image Super Resolution 1.2 Deep Learning for Super Resolution 2 Proposed Methodology 2.1 Stages of Recognition 3 Results and Discussion 3.1 Database for Stage I RCNN: 3.2 Stage II-GAN 3.3 Stage III-ALEXNET 3.4 RCNN-Based Object Detection 3.5 GAN-Based Super Resolution 4 Conclusion References Low-Power U-Net for Semantic Image Segmentation 1 Introduction 1.1 Convolutional Neural Networks 1.2 Quantized Neural Network 1.3 Need of Field Programmable Gate Array (FPGA) for inference of CNNs 2 Related Works 3 Methodology 3.1 Vitis™ AI Development Kit 3.2 Development Flow 3.3 Deep Learning Processing Unit 3.4 Network Architecture 3.5 Dataset 3.6 Training 4 Experiments and Results 4.1 Processing System—Programmable Logic System 4.2 Network Inference and Quantization 5 Discussion 6 Conclusion References Electrocardiogram Signal Classification for the Detection of Abnormalities Using Discrete Wavelet Transform and Artificial Neural Network Back Propagation Algorithm 1 Introduction 2 Proposed Method for Classifying ECG Signals 3 Collection of Database 4 Preprocessing of Electrocardiogram Signal 5 Feature Extraction of ECG Signal 6 Artificial Neural Network Classifier Using Back Propagation Algorithm 7 Simulated Results and Discussions of ANN Classifier Using Back Propagation Algorithm 8 Conclusion References Performance Analysis of Optimizers for Glaucoma Diagnosis from Fundus Images Using Transfer Learning 1 Introduction 2 Related Work 3 Methodology 3.1 Transfer Learning 3.2 Optimization Algorithms 4 Results and Discussion 5 Conclusion References Machine Learning based Early Prediction of Disease with Risk Factors Data of the Patient Using Support Vector Machines 1 Introduction 2 Literature Review 3 Proposed System 3.1 Overview of the Proposed System 3.2 Proposed System Implementation 4 Experiments and Results 4.1 Mobile Application for Health Monitoring 4.2 R-Studio Data Visualization 5 Discussion 6 Conclusion References Scene Classification of Remotely Sensed Images using Ensembled Machine Learning Models 1 Introduction 2 Related Works 3 Proposed Works 3.1 Speed-Up Robust Feature (SURF) Extraction 3.2 Ensemble Classifier Learning Systems 4 Performance Evaluation Metrics 4.1 Precision 4.2 Recall 4.3 Accuracy 4.4 F1-Score 5 Results and Discussions 5.1 Dataset Description 5.2 Experimental Analysis of Base Classifiers 6 Conclusion References 40 Fuzziness and Vagueness in Natural Language Quantifiers: Searching and Systemizing Few Patterns in Predicate Logic Abstract 1 Introduction 2 Related Works 3 Objectives of the Study 4 Quantifiers in Punjabi and Hindi 4.1 Punjabi Quantifiers 4.2 Hindi Quantifiers 5 Vague Nature for Punjabi Quantifiers: Some Investigations in a Predicate Logic 5.1 Structuring Fuzzy Quantifiers 5.2 Mapping Plan for Fuzzy Quantifiers 6 Discussions and Results 7 Conclusion and Future Endeavors Acknowledgements References An Attempt on Twitter ‘likes’ Grading Strategy Using Pure Linguistic Feature Engineering: A Novel Approach 1 Introduction 2 Related Work 3 Dataset 3.1 Collection 3.2 Filtering 3.3 Preprocessing 4 Methodology 4.1 Linguistic Features 4.2 Problem Formulation 4.3 Machine Learning Models 5 Feature Engineering 6 Results 7 Conclusion 8 Limitations and Future Work References Groundwater Level Prediction and Correlative Study with Groundwater Contamination Under Conditional Scenarios: Insights from Multivariate Deep LSTM Neural Network Modeling 1 Introduction 2 Literature Review 3 Multivariate LSTM Modeling Under Circumstantial Scenarios of Groundwater Prediction 3.1 Simulation Results of Conditional Scenarios 4 Groundwater Contamination and Predictive Modeling 4.1 Data Preprocessing and Empirical Modeling 4.2 Model Simulation and Discussion 5 Correlated Study Between Groundwater Level and Groundwater Contamination 6 Concluding Remarks and Future Scope of Study References A Novel Deep Hybrid Spectral Network for Hyperspectral Image Classification 1 Introduction 2 Related Works 3 Methodology 4 Dataset 5 Results 6 Conclusion References Anomaly Prognostication of Retinal Fundus Images Using EALCLAHE Enhancement and Classifying with Support Vector Machine 1 Introduction 2 Related Literary Work 3 Proposed System 4 Description of the Schematic Diagram 4.1 Dataset 4.2 Stage I—Preprocessing 4.3 Stage II—Segmentation 4.4 Stage III—Feature Extraction 4.5 Stage IV—Classification 4.6 Results and Conclusion 5 Conclusion and Future Scope References Analysis of Pre-earthquake Signals Using ANN: Implication for Short-Term Earthquake Forecasting 1 Introduction 1.1 Involving Concepts 1.2 Introduction to Neural Networks 1.3 Elman Backpropagation Neural Network 2 Study Area 3 Methodology 3.1 Anomalous Outgoing Longwave Radiation 3.2 Elman Backpropagation Neural Network 4 Findings and Discussions 4.1 Training Function 4.2 Layers and Description of Nodes 4.3 Input Nodes 4.4 Variables Involved 4.5 Spatial Parameters 4.6 Time Variable 5 Conclusion References A Novel Method for Plant Leaf Disease Classification Using Deep Learning Techniques 1 Introduction 2 Literature Review 3 Materials and Methods 3.1 Preprocessing 3.2 Training 3.3 Testing 4 Results and Discussion 5 Conclusion References