This paper presents a hardware implementation of a compartmental spiking neuron model based on RC circuits, designed for physical modeling of control and information processing systems using compartmental spiking neurons. The study developed circuit designs for individual functional elements of the CSNM neuron model. Simulation experiments demonstrated that the proposed implementation accurately reproduces the required temporal characteristics of pulse stream conversion compared to the mathematical model. Furthermore, compared to the previously proposed operational amplifier-based solution, the presented circuits are significantly simpler to implement in hardware. The solution allows flexible modification of the neuron’s dendritic and synaptic structure. A limitation of the proposed approach is its relatively low energy efficiency, though this is not critical in the context of modeling relatively small neural structures.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

The Hardware Implementation of the Compartmental Spiking Neuron Model (CSNM) Based on RC Circuits

  • Alexander Boiko,
  • Aleksandr Bakhshiev

摘要

This paper presents a hardware implementation of a compartmental spiking neuron model based on RC circuits, designed for physical modeling of control and information processing systems using compartmental spiking neurons. The study developed circuit designs for individual functional elements of the CSNM neuron model. Simulation experiments demonstrated that the proposed implementation accurately reproduces the required temporal characteristics of pulse stream conversion compared to the mathematical model. Furthermore, compared to the previously proposed operational amplifier-based solution, the presented circuits are significantly simpler to implement in hardware. The solution allows flexible modification of the neuron’s dendritic and synaptic structure. A limitation of the proposed approach is its relatively low energy efficiency, though this is not critical in the context of modeling relatively small neural structures.