Uplink capacity of MIMO cellular systems with multicell processing

Abstract

Multiple antennas are known to increase the link throughput by providing a multiplexing gain which scales with the number of antennas. Especially in cellular systems, multiple antennas can be exploited to achieve higher rates without the need for additional base station (BS) sites. In this direction, this paper investigates the multi-antenna capacity scaling in a cellular system which employs multicell processing (hyper-receiver). The model under investigation is a MIMO Gaussian cellular multiple-access channel (GCMAC) over a planar cellular array in the presence of power-law path loss and flat fading. Furthermore, the considered cellular model overcomes the assumption of user collocation utilized by previous models by incorporating uniformly distributed user terminals (UTs). The asymptotic eigenvalue distribution (a.e.d.) of the covariance channel matrix is calculated based on free-probabilistic arguments. In this context, we evaluate the effect of multiple BS/UT antennas on the optimal sum-rate capacity by considering a variable-density cellular system. Finally, the analytical results are interpreted in the context of a typical real-world macrocellular scenario.