OPTIMUM SELECTION OF ERROR CONTROL CODING USING NEURAL NETWORKS

Significant performance improvements may be obtained in digital commun ication system if error control coding is properly applied. However, s election of a coding scheme for specific applications is often a compl icated task. The choice is affected by a set of system design goals. S ome of these goals impose case-dependent conflicting requirements. Sim ilar scheme selection problems exist in many engineering system design processes. Here, a knowledge-combined neural network approach is deve loped and applied to optimum coding selection. The proposed approach u tilizes a neural network trained not only by precedent examples but al so by knowledge rules to draw conclusions. It is shown that artificial neural networks can provide effective solutions to the problems encou ntered in building systems that emulate a coding specialist's expertis e.