Exploiting Channel Angular Domain Information for Precoder Design in Distributed Antenna Systems

In this paper, the benefit of linear precoding along with distributed orthogonal space time block codes (OSTBCs) is exploited to improve the performance of distributed antenna system (DAS). In particular, we propose a novel precoder design method based on channel angular domain parameters, i.e., average angle of arrival (AOA), angular spread (AS), normalized antenna spacing (NAS) and Ricean K-factor, instead of the conventional channel correlation matrix which has been widely used to characterize MIMO channel. Built on the distributed characteristics of DAS, a more realistic non-Kronecker channel model is established, and then the optimal precoder is designed as a function of the above angular domain parameters. Asymptotic analysis reveals that, the optimal precoder acts as a combination of channel precoder and power allocator. Furthermore, the impact of the above angular domain parameters on system performance, including diversity order and coding advantage, is investigated in detail. Specifically, we find that better gain can be achieved with larger K-factor. Finally, our theoretical claims are confirmed by the numerical results.