A robust quadratically constrained adaptive blind multiuser receiver for cellular DS/CDMA systems

Abstract

A new low-computational complexity robust adaptive blind multiuser receiver based on a constrained optimization approach with multiple constraints and a quadratic inequality (QI) constraint on the weight vector norm is developed. A diagonal loading technique is required to apply the quadratic constraint where the diagonal loading value is adjusted to satisfy the constrained value. Implementing the quadratic constraint into adaptive update algorithms such as least-mean-square (LMS) or recursive-least-mean (RLS) seems to be a difficult task. The minimum output energy (MOE) detector is implemented using a new adaptive algorithm based on the partition linear interference canceller (PLIC) structure with multiple constraints and a QI constraint on the weight vector norm. A new optimal variable loading (VL) technique, which is capable of providing robust control against signal mismatch with low computational complexity, is adopted. Simulations in a rich multipath environment with near-far effect are presented to demonstrate the robustness and fidelity of the proposed detector.