Memristive Wilson–Cowan Neuron Models: Innovative Conceptual Framework and Programmable Analog and Digital Implementations

ABSTRACT Since general purposes of memristors are not yet widely available as standard, off‐the‐shelf components, finding a reliable hardware equivalent is essential for accurately implementing a memristive neural system. With this motivation, we improve the piecewise‐linear, piecewise quadratic, and cubic monotone‐increasing functions‐based memristive Wilson–Cowan (W‐C) neuron models in this paper. After determining the control parameters of these three memristive W‐C neuron models, we calculate their equilibrium points and Lyapunov exponents and then, we also illustrate the influence of the control parameters through bifurcation diagrams. Thus, the W‐C neuron model is re‐defined as including the essential characteristics of the memristors. Following the theoretical analysis of the memristive neuron models, we develop an alternative framework for spike‐timing‐dependent plasticity (STDP)‐based network structures. Finally, the hardware confirmations of these models are built separately by using field programmable gate array (FPGA) and field programmable analog array (FPAA).