Blind channel estimation in GSM receivers: a comparison of HOS and SOCS based approaches

Due to the solution of an eigenvalue problem, a common drawback of all blind estimation algorithms is the interference of the estimated channel impulse response caused by an unknown complex factor. Since maximum likelihood sequence estimation (MLSE) requires the knowledge of the exact channel characteristics, we present in this paper a method based on a linear mean square error (MSE) solution which enables the estimation of this factor using a-priori information of the source signal. On the assumption of COST-207 mobile radio channels, we compare the blind channel estimation approaches with the non-blind cross-correlation method (CC) used in state-of-the-art GSM receivers (Global System for Mobile communications) and give the bit error rates (BER) after Viterbi detection in terms of the signal-to-noise ratio (SNR). Finally, we also study the influence of non-Gaussian co-channel interference (CCI) on the blind and non-blind estimation schemes.