A new subspace method for blind estimation of selective MIMO-STBC channels

Abstract

In this paper, a new technique for the blind estimation of frequency and/or time-selective multiple-input multiple-output (MIMO) channels under space-time block coding (STBC) transmissions is presented. The proposed method relies on a basis expansion model (BEM) of the MIMO channel, which reduces the number of parameters to be estimated, and includes many practical STBC-based transmission scenarios, such as STBC-orthogonal frequency division multiplexing (OFDM), space-frequency block coding (SFBC), time-reversal STBC, and time-varying STBC encoded systems. Inspired by the unconstrained blind maximum likelihood (UML) decoder, the proposed criterion is a subspace method that efficiently exploits all the information provided by the STBC structure, as well as by the reduced-rank representation of the MIMO channel. The method, which is independent of the specific signal constellation, is able to blindly recover the MIMO channel within a small number of available blocks at the receiver side. In fact, for some particular cases of interest such as orthogonal STBC-OFDM schemes, the proposed technique blindly identifies the channel using just one data block. The complexity of the proposed approach reduces to the solution of a generalized eigenvalue (GEV) problem and its computational cost is linear in the number of sub-channels. An identifiability analysis and some numerical examples illustrating the performance of the proposed algorithm are also provided