Mem-elements for Neuromorphic Circuits with Artificial Intelligence Applications (Advances in Nonlinear Dynamics and Chaos (ANDC))

Front Cover
Mem-elements for Neuromorphic Circuits with Artificial Intelligence Applications
Copyright
Contents
List of contributors
Preface
Part 1 Mem-elements and their emulators
	1 The fourth circuit element was found: a brief history
		1.1 Memristor – the first step
		1.2 Properties of memristor
			1.2.1 Passivity criterion
			1.2.2 Closure theorem
			1.2.3 Existence and uniqueness theorem
		1.3 Memristive systems
		1.4 The first physical model of the memristor
		1.5 Memristor\'s applications
		1.6 Conclusion
		References
	2 Implementing memristor emulators in hardware
		2.1 Introduction
		2.2 Memristor modeling framework
		2.3 The fingerprints of a memristor
			2.3.1 The classical requirements: explicit conditions
			2.3.2 Additional requirements: implicit conditions
		2.4 Designing memristor emulators
			2.4.1 The diode bridge
			2.4.2 A short-term memory emulator
				2.4.2.1 Nonlinear resistive two-port for memristor implementation
			2.4.3 A long-term memory emulator
				2.4.3.1 Circuit description and operation
				2.4.3.2 Physical layout
				2.4.3.3 Results
		2.5 Conclusion
		References
	3 On the FPGA implementation of chaotic oscillators based on memristive circuits
		3.1 Introduction
		3.2 3D, 4D, and 5D memristive systems
		3.3 Numerical methods
			3.3.1 One-step methods
			3.3.2 Multi-step methods
			3.3.3 Stability of the numerical methods
		3.4 Analysis of memristive-based chaotic oscillators
		3.5 FPGA implementation of memristive systems
		3.6 Memristive-based secure communication system
		3.7 Conclusions
		References
	4 Microwave memristive components for smart RF front-end modules
		4.1 RF/microwave model of memristive switch and PIN diode
		4.2 Memristive phase-shifter realization
			4.2.1 Planar main-line memristor mounted type loaded-line phase shifter
			4.2.2 Implementation of main-line memristor mounted type loaded-line phase shifter and results
		4.3 Reconfigurable dual-band bandpass microwave filter
		4.4 Dual-band bandpass filter with multilayer dual-mode resonator enhanced with RF memristor
			4.4.1 Dual-mode resonator with memristor
			4.4.2 Dual-band bandpass filter with multilayer dual-mode resonator enhanced with RF memristor
			4.4.3 Conclusion
		Acknowledgments
		References
	5 The modeling of memcapacitor oscillator motion with ANN and its nonlinear control application
		5.1 Introduction
		5.2 Chaotic memcapacitor oscillator and its dynamical analysis
			5.2.1 Equilibrium points
			5.2.2 Bifurcation analysis
		5.3 Nonlinear feedback control
		5.4 Chaotic motion extraction from video and ANN
			5.4.1 Perceptron architecture
			5.4.2 Multilayer artificial neural networks
			5.4.3 Delayed artificial neural networks
			5.4.4 Training artificial neural networks
		5.5 Identification of memcapacitor system with ANN
		5.6 Conclusions
		References
	6 Rich dynamics of memristor based Liénard systems
		6.1 Introduction
		6.2 Model system
			6.2.1 Stability analysis
			6.2.2 Hidden attractors
		6.3 Mixed-mode oscillations
			6.3.1 Frequency scanning
			6.3.2 Amplitude scanning
			6.3.3 Successive period-adding sequence of MMOs
		6.4 Higher dimensional torus and large expanded chaotic attractor
		6.5 Conclusion
		Acknowledgment
		References
	7 Hidden extreme multistability generated from a novel memristive two-scroll chaotic system
		7.1 Introduction
		7.2 Memristive two-scroll chaotic system and its basic properties
		7.3 Dynamical analysis of memristive two-scroll chaotic system
			7.3.1 Two-parameter Lyapunov exponents analysis
			7.3.2 One parameter bifurcation analysis
			7.3.3 Emergence of hidden extreme multistability
			7.3.4 Remerging period-doubling bifurcation
			7.3.5 Offset boosting control
		7.4 Circuit design and experimental measurements
		7.5 Conclusion
		References
	8 Extreme multistability, hidden chaotic attractors and amplitude controls in an absolute memristor Van der Pol–Duffing circuit: dynamical analysis and electronic implementation
		8.1 Introduction
		8.2 Theoretical analysis of an absolute memristor autonomous Van der Pol–Duffing circuit
		8.3 Electronic implementation of an absolute memristor autonomous Van der Pol–Duffing circuit
		8.4 Conclusion
		References
	9 Memristor-based novel 4D chaotic system without equilibria
		9.1 Introduction
			9.1.1 Literature survey
			9.1.2 Application of memristor and memristive circuit
		9.2 Brief introduction to flux- and charge-controlled memristor models and novel chaotic system
			9.2.1 Introduction to flux- and charge-controlled memristor models
			9.2.2 Flux-controlled memristor-based novel chaotic system
		9.3 Properties and behaviors of memristor-based novel chaotic system
			9.3.1 Symmetry and invariance
			9.3.2 Dissipation
			9.3.3 Lyapunov spectrum
			9.3.4 Bifurcation diagram
			9.3.5 Kaplan–Yorke dimension
			9.3.6 Poincaré section
		9.4 Projective synchronization between the memristor-based chaotic systems
		9.5 Simulation results and discussion
		9.6 Conclusions and future scope
		References
	10 Memristor Helmholtz oscillator: analysis, electronic implementation, synchronization and chaos control using single controller
		10.1 Introduction
		10.2 Design and analysis of the proposed memristor Helmholtz oscillator
			10.2.1 Equilibrium points and their stabilities
			10.2.2 Dynamical analysis of memristor Helmholtz oscillator
		10.3 Electronic circuit simulations of the proposed memristor Helmholtz oscillator
		10.4 Chaos synchronization of unidirectional coupled identical chaotic memristor Helmholtz oscillators
		10.5 Chaos control of memristor Helmholtz oscillator using single controller
		10.6 Conclusion
		References
	11 Design guidelines for physical implementation of fractional-order integrators and its application in memristive systems
		11.1 Introduction
		11.2 Fractional-order calculus preliminaries
			11.2.1 Grünwald–Letnikov definition
		11.3 Fractional-order memristive systems
		11.4 Continued fraction expansion (CFE)
			11.4.1 CFE error analysis
		11.5 Implementation of fractional-order integrators using FPAAs
			11.5.1 Fractional-order integrator based on a first order transfer function
		11.6 Electronic implementation of a fractional-order memristive system
		11.7 Conclusions
		Acknowledgments
		References
	12 Control of bursting oscillations in memristor based Wien-bridge oscillator
		12.1 Introduction
		12.2 Mathematical model of LC network based diode bridge memristor
		12.3 Memristive Wien-bridge oscillator
		12.4 Chaotic and periodic bursting oscillations (BOs)
		12.5 Control of active states and quiescent states in BOs
		12.6 Control of amplitude in BOs
			12.6.1 Amplitude increasing in BOs
			12.6.2 Amplitude decreasing in BOs
		12.7 Conclusion
		Acknowledgments
		References
Part 2 Applications of mem-elements
	13 Memristor, mem-systems and neuromorphic applications: a review
		13.1 Introduction
		13.2 Memristor and mem-systems
			13.2.1 Memristor
			13.2.2 Memristive systems
		13.3 Neuromorphic systems
			13.3.1 Neuron model
			13.3.2 Synapse
			13.3.3 Neural network
		13.4 Reservoir computing
		13.5 Conclusion
		Acknowledgment
		References
	14 Guidelines for benchmarking non-ideal analog memristive crossbars for neural networks
		14.1 Introduction
		14.2 Basic concepts
			14.2.1 Memristor simplified
			14.2.2 Crossbar simplified
			14.2.3 Analog neural networks simplified
		14.3 Non-idealities of memristors
			14.3.1 Aging
			14.3.2 Electromagnetics and signal integrity
			14.3.3 Conductance variabilities
			14.3.4 Noise effects
			14.3.5 System implementations
		14.4 Applications
			14.4.1 Programmable logic
			14.4.2 Neuromorphic accelerators
		14.5 Conclusions
		References
	15 Bipolar resistive switching in biomaterials: case studies of DNA and melanin-based bio-memristive devices
		15.1 Introduction
		15.2 Brief overview of resistive switching and memristive devices
		15.3 Materials for resistive switching application
		15.4 Biomaterial-based memristive devices
			15.4.1 Case study 1: DNA-based memristive device
				15.4.1.1 Materials and methods
				15.4.1.2 Results and discussion
				15.4.1.3 Summary
			15.4.2 Case study 2: melanin-based memristive device
				15.4.2.1 Materials and methods
				15.4.2.2 Results and discussion
				15.4.2.3 Summary
		15.5 Conclusion and future outlook
		Acknowledgments
		References
	16 Nonvolatile memristive logic: a road to in-memory computing
		16.1 Introduction
		16.2 Memristive logic gates in crossbar array
		16.3 R-R logic gate
			16.3.1 Material implication logic (IMPLY or IMP)
			16.3.2 Variants of IMP — NOR logic gate
			16.3.3 Variants of IMP – NAND and AND logic gates
			16.3.4 Memristor-aided logic (MAGIC)
			16.3.5 Hyperdimensional computing in 3D RRAM
			16.3.6 R-R logic based on neural networks
			16.3.7 Conclusion of memristive R-R logic
		16.4 Memristive V-R logic
			16.4.1 Memristive sequential logic based on a single device
			16.4.2 Four-variables methods
				16.4.2.1 A four-variable method using a single BRS in a crossbar array
				16.4.2.2 A four-variable method using a one-transistor–one-resistor (1T1R) cell
			16.4.3 Other memristive V-R logic methods
			16.4.4 Conclusion of the memristive V-R logic
		16.5 Challenges and outlooks
		Acknowledgments
		References
	17 Implementation of organic RRAM with ink-jet printer: from design to using in RFID-based application
		17.1 Introduction
		17.2 Design process
		17.3 Fabrication process
		17.4 A practical application example
		17.5 Conclusion
		References
	18 Neuromorphic vision networks for face recognition
		18.1 Introduction
		18.2 Preliminaries
		18.3 Model description
		18.4 Template formation
		18.5 Face recognition
		18.6 Memristive threshold logic (MTL)
		18.7 Edge detection with memristive threshold logic (MTL) cells
		18.8 Circuit realization
			18.8.1 Edge detection and template formation
		18.9 Experimental setup
		18.10 Results and discussion
		18.11 Future research directions
		18.12 Conclusion
		18.13 Key terms and definitions
		References
	19 Synaptic devices based on HfO2 memristors
		19.1 Introduction
		19.2 HfO2-based resistive switching structures
		19.3 Demonstration of learning rules in memristor devices to mimic biological synapses
		19.4 Stability and reliability issues of resistive synaptic devices
		19.5 Physical simulation of memristors
			19.5.1 Simulation schemes for the description of memristor physics
			19.5.2 Kinetic Monte Carlo simulation approach
			19.5.3 Kinetic Monte Carlo simulation of conductive-bridge RAMs
			19.5.4 Valence change memories kinetic Monte Carlo simulation
		19.6 Memristor compact modeling
			19.6.1 VCM compact modeling
			19.6.2 Compact modeling of CBRAMs
		19.7 Memristor random telegraph noise
			19.7.1 Numerical procedures to analyze random telegraph signals
				19.7.1.1 Current versus time trace representation
				19.7.1.2 Time lag plot (TLP) representation
				19.7.1.3 Color code time lag plot (CCTLP)
				19.7.1.4 Radius time lag plot (RTLP)
				19.7.1.5 Weighted time lag plot (WTLP)
				19.7.1.6 Locally weighted time lag plot (LWTLP)
				19.7.1.7 Differential locally weighted time lag plot (DLWTLP)
		19.8 Conclusion
		Acknowledgments
		References
	20 Analog circuit integration of backpropagation learning in memristive HTM architecture
		20.1 Introduction
		20.2 What is HTM?
		20.3 Memristive HTM architectures
		20.4 Analog backpropagation circuit integration in memristive HTM
		20.5 Discussion and open problems
		20.6 Conclusions
		References
	21 Multi-stable patterns coexisting in memristor synapse-coupled Hopfield neural network
		21.1 Introduction
		21.2 Memristor synapse-coupled HNN with three neurons
		21.3 Bifurcation behaviors with multi-stability
			21.3.1 Dynamics depended on the coupling strength k
			21.3.2 Dynamics depended on the synaptic weights
				21.3.2.1 2-dimensional bifurcation diagrams
				21.3.2.2 Dynamics depended on the synaptic self-connection weight w11
				21.3.2.3 Dynamics depended on the synaptic inter-connection weight w23
			21.3.3 Dynamics depended on the initial conditions
			21.3.4 Long-term transient chaotic behaviors
		21.4 Circuit synthesis and PSIM simulation
		21.5 Conclusion
		Acknowledgment
		References
	22 Fuzzy memristive networks
		22.1 Introduction
		22.2 Requirement
			22.2.1 Fuzzy calculus concepts
			22.2.2 Fractional calculation concepts
		22.3 Memristive fuzzy logic systems
		22.4 Delay memristive fuzzy systems
		22.5 Fractional memristive fuzzy systems
		22.6 General overview of the area and future trends
		References
	23 Fuzzy integral sliding mode technique for synchronization of memristive neural networks
		23.1 Introduction
		23.2 Model description
		23.3 Controller design
			23.3.1 ISMC
			23.3.2 FISMC
		23.4 Numerical results
			23.4.1 Memristive neural network without controller
			23.4.2 Stabilization and comparison of the proposed method with the ISMC
			23.4.3 Synchronization of uncertain memristive neural networks
		23.5 Conclusions
		References
	24 Robust adaptive control of fractional-order memristive neural networks
		24.1 Introduction
		24.2 Model of the system and preliminary concepts
		24.3 Controller design
		24.4 Simulation results
			24.4.1 Synchronization of fractional-order memristive neural networks
			24.4.2 Comparison of the proposed adaptive controller method with PI control
		24.5 Conclusion
		References
	25 Learning memristive spiking neurons and beyond
		25.1 Introduction
		25.2 Spike domain data processing and learning
		25.3 Learning in memristive neuromorphic architectures
			25.3.1 Spiking neural networks
			25.3.2 Complex SNN typologies: long-short term memory and hierarchical temporal memory
			25.3.3 Recent advances in memristor-based SNN architectures
		25.4 Open problems and research directions
		25.5 Conclusion
		References
Index
Back Cover