Performance and Optimized Design of Space-Time Codes for MIMO Wireless Systems With Imperfect Channel Estimates

Abstract

Multiple-antennas constitute an effective mean to achieve spatial diversity in emerging bandwidth-efficient multiple-input multiple-output (MIMO) wireless systems. Until now, most contributions in this area are focused on the two limit cases of fully coherent and fully incoherent decoding, which, in turn, occur when perfect channel estimates and no channel estimates are available at the receiver. However, very accurate channel estimates typically demand long training sequences that reduces spectral efficiency. Therefore, testing capabilities of multiple-antenna systems with partially coherent decoding may be appealing for improving power-versus-bandwidth tradeoff. In this contribution, we focus on the optimized design and performance evaluations of multiple-antenna block-coded systems with partially-coherent maximum-likelihood (ML) decoding. After considering emerging fourth-generation WLANs (4GWLANs) as a possible application scenario, we present new performance bounds and optimized design criteria together with a new family of robust space-time unitary block codes that "self-match" to channel-estimation errors. We name these codes "self matching.".