Multiple-antenna CDMA mobile reception using constrained normalized adaptive algorithms

Abstract

We analyze adaptive algorithms for a multiple-antenna code division multiple access (CDMA) mobile receiver. We study the constrained versions of the binormalized data-reusing LMS (BNDR-LMS) and the normalized LMS (NLMS). Their performance is compared with that of the classical Frost algorithm. Time-varying step-sequences are used to further enhance the performance of the receiver. The results show that considerable improvement in the convergence speed can be achieved with the use of the normalized algorithms compared with the conventional Frost algorithm. Using an optimal step-sequence in the update provides a fast convergence and a near-optimum steady-state performance.