Interference alignment in coordinated multipoint systems

It has been shown that in wireless networks, when receivers or transmitters are coordinated, it is possible to achieve more throughput than usual. In this paper, we consider a K-user multi-input multi-output (MIMO) interference channel for the case receivers are coordinated such that they can exchange the effective cross channel state information (CCSI) and the true data. We evaluate the performance of interference alignment via alternating minimization (IA-AM) in [1] over coordinated receivers and compare it with the original IA-AM algorithm without any coordination and the maximum-signal-to-interference-plus-noise-ratio IA algorithm (Max-SINR IA) in [2]. We show that even in the partially coordinated scenario when the k-th receive node knows the data and the effective CCSI of only K -k nodes, the total achievable rate becomes higher than that of the original IA-AM scheme at all signal-to-noise ratio (SNR) ranges. Meanwhile at high SNRs, the achievable sum rate of the proposed method becomes higher than that of the Max-SINR IA algorithm when K is small enough. We also show that the algorithmically arithmetic complexity of our proposed IA-AM with partially coordinated receivers is nearly half of that of the original IA-AM counterpart.