Novel Space-Time Coded MIMO Receivers in the Presence of Channel Estimation Errors

Novel maximum likelihood (ML) receivers for multiple-input and multiple-output systems with imperfect channel estimation are proposed. Two different cases are considered. In the first case, the conditional probablity density function (PDF) of the channel estimate is assumed Gaussian and known. In the second case, the joint PDF of the channel estimate and the true channel gain is assumed Gaussian and known. In addition to optimal signal detection with channel estimation, optimal signal detection without channel estimation is also studied. To reduce the implementation cost of optimal receivers, two suboptimal structures are derived. As an example, the Alamouti space-time codes are used to examine the performances of the new receivers. Simulation results show that the new receivers can reduce the gap between the conventional receiver with channel estimation errors and the receiver with perfect channel knowledge at least by half in some cases.