Performance Analysis of Adaptive Arrays based on Minimum BER Algorithms

This paper derives theoretical performance of adaptive arrays based on minimum BER algorithms which achieve superior performance to that on MMSE based algorithms such as LMS (least mean square) in the presence of lots of interference. Especially, LBER (least BER) algorithm which has been proposed as one of minimum BER algorithms is applied for the derivation in this paper. It is shown that the LBER algorithm can be regarded as a variable gain adaptive algorithm where the variable stepsize is a function of eigenvalues of the channel matrix and transmission signal states. The theoretical performance is confirmed to fit that of computer simulation.