Search

For joint blind channel estimation and multiuser detection in direct-sequence code division multiple access (DS-CDMA) systems, a combined scheme is proposed, which couples the recursive least-squares (RLS) adaptations and the PASTd subspace tracking algorithm. A major advantage of the RLS-PASTd algorithm is that it admits an adaptive implementation with low computational complexity instead of singular value decomposition. Simulation results demonstrate the fast tracking capability and high steady-state estimation accuracy of the proposed algorithm in both the time-invariant and time-variant environments [ABSTRACT FROM PUBLISHER]

For code division multiple access (CDMA) systems operating over dispersive channels, the nested recursive least-squares (RLS)-type algorithm, based on minimum output energy (MOE) criterion, attracts much attention as an efficient interference suppression scheme for its simple adaptive recursion. In this letter, we investigate its performance analysis, Special attention is paid to theoretical issues concerning the global convergence and convergence dynamics performance. It is shown that this algorithm exhibits a global minimum attained, and the convergence is immune to the eigenvalue spread of the correlation matrix of received signal vector. Simulation results demonstrate that the nested RLS-type algorithm has indeed the expected convergence performance. [ABSTRACT FROM AUTHOR]