The hardware structure design of perceptron with FPGA implementation

Most commonly, neural networks's models or algorithms are simulated and implemented by computer programming in neural networks theory research. But in many practical applications, it is necessary to consider essential issues such as hardware implementation. Specific application of neural networks hardware has the advantages of high speed, small in size, good performance and low cost. Thus, the implementation of high performance neural networks hardware is the final target in some actual applications. In this paper, a hardware structure of single perceptron that serves as the basic nerve cell and its implementation method with FPGA is introduced. It is based on VLSI implementation approach for the standard neural networks. The method proposed is a primary discussion and research for the hardware implementation of artificial neural networks.