Direct neural-network hardware-implementation algorithm

An algorithm for compact neural network hardware implementation is presented, which exploits special properties of the Boolean functions describing the operation of artificial neurones with step activation function. The algorithm contains three steps: ANN mathematical model digitisation, conversion of the digitised model into a logic gate structure, and hardware optimisation by elimination of redundant logic gates. A set of C++ programs automates algorithm implementation, generating optimised VHDL code. This strategy bridges the gap between ANN design software and hardware design packages (Xilinx). Although the method is directly applicable only to neurones with step activation functions, it can be extended to sigmoidal functions.