Digital multiplierless realization of two-coupled biological Hindmarsh-Rose neuron model

Abstract

The efficient modeling, simulation, and implementation of biological neural networks are key objectives of the neuromorphic research field, leading to potential applications, such as assisting the search for new solutions to cure brain diseases, improved performance of robots, and the fundamental study of neural network behavior. This brief proposes a modified biological Hindmarsh-Rose (HR) neuron model that is more suited for efficient implementation on digital platforms. Simulation results show that the model can reproduce the desired behaviors of the neuron. The proposed model is investigated, in terms of digital implementation feasibility and cost, targeting a low-cost hardware implementation. Hardware implementation on a field-programmable gate array shows that the modified model mimics the biological behavior of different types of neurons, with higher performance and considerably lower hardware overhead cost compared with the original HR model.