학술논문

An FPGA based Hardware Emulator for Neuromorphic Chip
Document Type
Conference
Source
2020 International Conference on Electronics and Sustainable Communication Systems (ICESC) Electronics and Sustainable Communication Systems (ICESC), 2020 International Conference on. :1131-1136 Jul, 2020
Subject
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Fields, Waves and Electromagnetics
Robotics and Control Systems
Signal Processing and Analysis
Mathematical model
Biological system modeling
Neurons
Computational modeling
Brain modeling
Field programmable gate arrays
Action potentials
Izhikevich Model
Vivado
LabVIEW
MATLAB
Spiking Neuron
cortical
Language
Abstract
The brain can be considered as the central processing unit (CPU) of the human body. The elementary units of the brain used for processing can be considered as the neurons fired in the brain. The aim is to reproduce the spiking and bursting of the specific type of neuron called the cortical neurons. The model used for reproducing the different types of behaviors of the neurons which is accurate and is functionally feasible and which correlates the most to the actual biological process is called the Izhikevich model which is a mathematical model. The focus of the work is to implement Simulated model on FPGA platform. The Look up table based approach is utilised to generate samples at the output of the 12-bit Digital to Analog Converter. The data samples are generated by implementing Izhikevich model in LabVIEW and Matlab on Console PC which are then transmitted to FPGA through communication interface. The data from the Console PC is stored in the Block Memory generator which is then applied to Digital to Analog Converter IP and waveform signal is generated by applying appropriate timing trigger signals. The gener-ated waveform signal are analyzed on the DSO at the sampling rate of 1 MHz using Artix-7 FPGA family and 12-bit AD5628 Digital to Analog Converter.