Side Information Effect on Semi-Blind Channel Identification for MIMO-OFDM Communications Systems

This paper investigates the impact of different priors, related to communications channels, on semi-blind channel estimation performance for Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM) communications systems. For an estimator-independent study, the Cramér-Rao Bound (CRB) is considered to analyze the performance limits and to quantify the channel estimation gain obtained by properly exploiting side information about the propagation channel. The analysis is carried out by considering three cases: (i) no modeling while estimating the channel coefficients in the frequency domain (for each frequency bin); (ii) a finite memory linear time invariant channel model while estimating the channel taps in the time domain; and finally (iii) a specular channel model while estimating the propagation channel parameters, i.e. the fading, the delay and the angle of arrival of each path. In the latter case, the impact of a model misspecification (caused for example by array calibration errors) on the channel estimation performance is investigated by using the misspecified CRB (MCRB).