Morris-Lecar model neurons and neural networks: FPGA implementation and analysis

Based on the dynamics of neurons, we use the FPGA to realize Morris-Lecar (ML) model, and to achieve dynamic analysis such as bifurcation.etc in this paper, obtaining the corresponding firing patterns; then we use the FPGA to achieve the ML neuron network which is connected by chemical synapse, and analyze the affection of parameters on the neural network dynamic characteristics; At the same time, the FPGA implementation is compared with the MATLAB simulation. We found that, the characteristics of neuron FPGA and network meet the requirements of computational neuroscience and, the simulation of a single FPGA neuron is about six times faster than that of MATLAB in the same simulation accuracy and step conditions. In addition, as the network size increases, the speed advantage of FPGA neural network relative to the MATALB simulation will continue to expand.