Channel Estimation and Symbol Detection for Communications on Overlapping Channels

Abstract

In this paper, we investigate the joint channel estimation, interference cancellation and data detection problem for a general setting in which a desired communication is interfered by another communication having a different symbol rate and being asynchronous with the considered communication. The fast fading channel gains of the desired communication and the effective interference coefficients (EIC) induced by filters of the interfering signal must be estimated to enable reliable detection of the desired data. However, the joint estimation of fast fading channel and EIC from communications with different bandwidths has not been studied in literature. Toward this end, we propose a two-phase strategy for channel estimation and data detection. In the first phase, we derive the closed-form maximum-likelihood estimator of the EIC. Then, the interference is subtracted and the desired channel gains corresponding to pilot symbols are estimated. In the second phase, with the knowledge of desired channel information obtained from the previous phase, we derive the posterior probability for data symbols for the soft data detection. Via numerical studies, we demonstrate that our design can effectively cancel the interference and the soft detection approach can achieve better symbol error rate compared to the existing message passing based detection approach. We show that our design can perform well for a wide range of interference power and frequency spacing and it has lower complexity than the existing technique.