Effective implementation of spiking neuron models in hardware is crucial for real systems. Utilizing the main capabilities of FPGAs, this paper introduces a highly precise method for evaluating nonlinear functions. The approach relies on effectively matching trigonometric-based functions to approximate the nonlinear terms of a Fitzhugh-Rinzel neuron model uses the electromagnetic flux coupling with a focus on cost-effectiveness and high-speed digital implementation using the CORDIC algorithm and multiplierless design. The close correspondence between the approximate functions and the nonlinear functions of the original model results in minimal errors in the outputs of the proposed model compared to the original model which reduces the lead and lag of signals between the original model and the proposed models. For the digital FPGA implementation of the FHR neuron model, we employed the Virtex-5 board to validate and synthesize the suggested method. In this scenario, the proposed FHR model demonstrates superior performance in terms of speed and cost compared to the original model. The speed-up of our proposed model is about 6 times faster than the original model (414.86 MHz compared to 69.232 MHz) and also, the number of fitted neurons for our proposed approach is about 6.66 times (20 compared to 3).

The neuron can be called the main cell of a nervous system that can transmit messages from one neuron to another neuron or another cell through electrical signals. In neuromorphic engineering, the hardware realization and simulation of these neurons are crucial. To accomplish a proper digital implementation (i.e. reducing hardware resources and increasing speed and accuracy) of three important neuron models including Hodgkin-Huxley, Morris-Lecar, and Izhikevich, this study proposes a set of multiplier-less mathematical equations based on converting nonlinear functions to$2^{x}$functions. Then optimizes the proposed equations based on reducing the number of different$2^{x}$terms. The suggested model can accurately recreate the behavioral characteristics of the original neuron models. The suggested model was synthesized and implemented on the Zynq XC7Z010 (3CLG400) reconfigurable board (FPGA) to validate the mathematical simulation findings. The results of hardware synthesizing and implementations of the proposed model show that different biological behavior can be replicated with greater efficiency and at substantially reduced implementation costs. This method (implemented on the zynq board) can raise the frequency of the proposed models at least by up to 3.5 times that of the original model and reduce power consumption between 20% and 60% for different proposed models. Also, due to the reduction of hardware resources in the proposed model, it is possible to implement a much larger number of neurons (between 4 and 12 times) relative to the original model on a single zynq board.

In recent years, hardware modeling for various parts of the body’s sensitive organs, including the brain and nervous system, heart and eyes, has been considered for the treatment of diseases and rehabilitation, as well as for moving towards the construction of artificial prostheses. The retina is a thin layer that is the innermost layer of the human eye. In this paper, low-cost hardware implementation for retinal cone cells is performed. Existing mathematical models for implementing the behavior of these cells include a series of nonlinear functions that, if implemented directly, would require a large amount of hardware and, in addition, would not have the desired speed. The proposed model uses multi-linear functions to approximate the nonlinear terms and eliminate the multiplication expressions. The simulation results show that the proposed model tracks the behavior of the original model with high precision. There is also a good match between the main model and the proposed model in terms of dynamic behaviors. The results of hardware implementation using the virtex5 XC5VLX20T (2FF323) reconfigurable board (FPGA) show that the proposed model is fully valid and has a lower hardware volume as well as a 4 times higher frequency, and 22% less power consumption than the original model.

Neurons are the basic blocks in the Central Nervous System (CNS). Simulation and hardware realization of these blocks are vital in neuromorphic engineering. This paper presents a set of multiplierless mathematical equations based on 2Xterms to achieve a low-cost, high-speed, and high-accuracy digital implementation of Hodgkin-Huxley (HH) neuron model. The HH model is the most complicated and high-accuracy among the mathematical neuron models. The proposed model can reproduce spiking behaviors of the original HH model with high precision. To validate the mathematical simulation results, the proposed model has been synthesized and implemented on Field-Programmable Gate Array (FPGA) development board. Hardware synthesis and physical implementations reveal that the biological behavior of different spiking patterns can be reproduced with higher performance and significantly lower implementation costs compared with the original HH model. Also, in this approach the maximum frequency of 200 MHz is achievable which is valuable in comparison with other similar works.