Low complexity variable step size mechanisms for stochastic gradient algorithms in minimum variance CDMA receivers

In this paper, the performance of blind adaptive receivers for direct sequence code division multiple access (DS-CDMA) systems that employ stochastic gradient (SG) algorithms with variable step size mechanisms is investigated. Two low complexity variable step size mechanisms are proposed for estimating the parameters of linear CDMA receivers that operate with SG algorithms. For multipath channels the novel adaptation mechanisms are also incorporated in the channel estimation algorithms, whereas for the single-path case the novel techniques are restricted to the linear receiver parameter vector estimation. Analytical expressions for the excess mean squared error (MSE) are derived and a convergence analysis of the proposed adaptation techniques is carried out for both frequency selective and flat scenarios. Finally, numerical experiments are presented for nonstationary environments, showing that the new mechanisms achieve superior performance to previously reported methods at a reduced complexity.