Effects of Training Parameters on Neural Network Parallel Interference Canceller for CDMA

Abstract

Abstract-Reduction of multiple access interference is occupying great importance in wireless communication using code division multiple access (CDMA). Different interference cancellation schemes were developed and neural network based canceller is one of them. This paper introduces the effect of training parameters on achieving optimum performance on a multi stage parallel interference canceller (PIC) based on neural network.Index Terms-BER, CDMA, LM algorithm, neural network.