A Subspace Semi-Blind Adaptive Multiuser Detection Algorithm Based on Kalman Filter for MC-CDMA Uplink

Abstract

In this paper, a new subspace semi-blind adaptive multiuser detection based on a hybrid of Kalman filter and subspace estimation is proposed for MC-CDMA uplink. It is shown that the detector can be expressed as an anchored signal in the signal subspace and coefficients can be estimated by the Kalman filter. The spreading sequences of all known users are exploited to develop a new efficient semi-blind adaptive multiuser detection requiring fewer adaptive coefficients. To reduce the computational complexity, an improved projecting approximation subspace tracking with deflation (PASTd) algorithm is developed for adaptive signal subspace estimation. Simulation results demonstrate that the proposed semi-blind multiuser detection offers substantial performance gains over the blind multiuser detection method in the MC-CDMA uplink environment