Tensor-Based Joint Channel and Symbol Estimation for Two-Way MIMO Relaying Systems

In this letter, we present two closed-form semiblind receivers for a two-way amplify-and-forward relaying system. The proposed receivers jointly estimate the symbol and channel matrices involved in the two-way relaying system by exploiting tensor structures of the received signals at the relay and the destination, and without using training sequences, in contrast to previous works. Differently from competing receivers, one of the proposed receivers does not require channel reciprocity between uplink and downlink phases, which can be of interest in frequency division duplexing relaying systems. Parameter identifiability and computational complexity are analyzed, and simulation results are provided to corroborate the effectiveness of the proposed semiblind receivers in scenarios with and without channel reciprocity.