Semi-blind channel estimation for WCDMA systems with parallel data and pilot signals

Maximum-likelihood (ML) based semi-blind estimators with Gaussian assumptions can provide improvement in performance compared to channel estimation schemes using the pilot signal only. In the third generation UTRA-FDD standard context this improvement can be even larger due to fact that the pilot and the data signals are sent simultaneously. However, the Gaussian ML approach results in very large complexity. In this paper, assumptions which reduce the computational complexity of the method significantly are proposed, without causing performance degradation. The assumptions are justified through simulations and the performance improvement over estimation schemes using pilot signal solely is verified