No abstract

Different architectures and techniques have developed in the neuromorphic field to mimic and investigate the activity of biological neural networks. This paper presents a set of piece-wise linear approximations of a two-dimensional Hindmarsh-Rose neuron model for digital circuit implementation to achieve higher speeds and lower hardware costs in large-scale implementation of the biological neural networks. The performance of the model was evaluated with a time domain signal error. Synthesis and hardware implementation on a field-programmable gate array, as a proof of concept, indicates that the proposed model reproduces several neuronal behaviors similar to the original model with higher performance and considerably lower implementation costs.

No abstract