A fundamental problem arising in dense wireless networks is the highco-channel interference. Interference alignment (IA) was recently proposed asan effective way to combat interference in wireless networks. The concept ofIA, though, is originated by the capacity study of interference channels and assuch, its performance is mainly gauged under ideal assumptions, such asinstantaneous and perfect channel state information (CSI) at all nodes, andhomogeneous signal-to-noise ratio (SNR) users, i.e., each user has the sameaverage SNR. Consequently, the performance of IA under realistic conditions hasnot been completely investigated yet. In this paper, we aim at filling this gapby providing a performance assessment of spatial IA in practical systems.Specifically, we derive a closed-form expression for the IA average sum-ratewhen CSI is acquired through training and users have heterogeneous SNR. A maininsight from our analysis is that IA can indeed provide significant spectralefficiency gains over traditional approaches in a wide range of dense networkscenarios. To demonstrate this, we consider the examples of linear, grid andrandom network topologies.
展开▼
