Blind channel estimation based on subspace for multicarrier CDMA
Jun Wu, Yi Wang, Kuang‐Wei Cheng · 2002
The channel model and effective signature waveform are analyzed for synchronous multicarrier CDMA (MC-CDMA), and a blind channel estimation algorithm is proposed. We deduce the effective signature waveform in MC-CDMA systems, which is the convolution of signature and unknown channel vector and followed by OFDM demodulation. Thus, the effective signature estimation problem in a subspace-based multiuser detector is actually the channel vector estimation problem. Based on the analysis, a blind channel estimation algorithm is proposed, which provides a close-form signal estimate. Numerical results show that the analysis of effective signature waveform in MC-CDMA is exact and the proposed algorithm is robust in multiuser MC-CDMA system.