A joint data rate-error rate analysis in correlated space-time wireless channels.

摘要

Future generations of wireless communications demand (a) high data rate and (b) low error rate. However, information theory has revealed that in general the two performance metrics can not be optimized simultaneously. In other words, there is a tradeoff between data rate and error rate. Recently, this tradeoff is studied in multi-input-multi-output (MIMO) fading channels, where multi-antenna is employed at both transmitter side and receiver side.; Original works on the tradeoff assumes rich scattering environments, i.e., fadings across different antenna pairs are independent. But, in real applications, scattering sources can be poor. For example, rural, on highway, or on the top of a skyscraper. Therefore, practically fadings across different antenna pairs are not independent: they are correlated. It is motivated to study the tradeoff in poor scattering environments. However, solving the interested problem is not as trivial as it looks like. In fact, it has to be decomposed into several simpler problems.; First, the tradeoff is investigated in single-input-single-output (SISO) channels, where temporal correlations are incorporated. A novel concept, spike sharpening effect, is introduced to interpret the tradeoff in SISO channels.; Second, the single-input-multi-output (SIMO) channels are studied, where multiple antennas are employed at the receiver side. Error performance in SIMO channels is analyzed and simulated under perturbations and non-stationarity considerations.; Thirdly, correlated MIMO channel modeling is studied, i.e., spatially correlated fadings are considered. A unified framework is proposed to accommodate several stochastic models that are presented in the literature. Moreover, under this framework, it is depicted that different models are feasible in different physical environments.; Finally, an innovative channel capacity analysis reveals that, in high signal-to-noise ratio regimes, a correlated MIMO channel can be approximately decomposed into a linear summation of multiple parallel SISO and SIMO channels. Then, combining the previous results, the tradeoff in poor scattering MIMO channels is established.; The above results are not only meaningful for abstract analysis, but also useful for practical algorithm development in wireless communications. They help wireless engineer to incorporate practical considerations (Doppler spread, perturbation, stationarity, poor scattering, etc) into their design, and aim to build a system that can approach its ultimate performance in real environments.