Investigation on time-multiplexing cellular neural network simulation by RKAHeM(4,4) technique

Abstract

In practical sense owing to hardware limitations, it is not possible to have a one-one mapping between the CNN hardware processors and all the pixels of the image. The time-multiplexing approach plays a pivotal role in the area of simulating hardware models and testing hardware implementations of cellular non-linear networks (CNNs). In this framework, time-multiplexing scheme is used to process large images using small CNN arrays. Using a novel integration algorithm by formulating an embedded technique involving RK technique based on arithmetic mean (AM) and Heronian mean (HeM) with error control for general CNNs is presented. Simulation results and comparison have also been made to show the efficiency of the numerical integration algorithms. The analytic expression for local truncation error (LTE) has been derived. It is found that the RK-embedded HeM gives promising results in comparison with the Harmonic mean. A more quantitative analysis has been carried out to clearly visualise the goodness and robustness of the proposed algorithm.