Distributed interference alignment with limited feedback for cellular networks

In theory, coordinated multipoint transmission (CoMP) promises vast gains in spectral efficiency. However, industrial field trials show rather disappointing throughput gains, whereby the major limiting factor is properly sharing channel state information. This so-called limited feedback problem has been considered in many recent papers in context of CoMP under the assumptions: 1) infinite SNR regime, 2) no user selection and 3) ideal link adaption; rendering the analysis overly optimistic. In this paper we make a step forward towards a more realistic assessment of the limited feedback problem by introducing an improved metric for the performance evaluation, which better captures the throughput degradation. We obtain the relevant scaling laws (lower and upper bounds) and show that they are, at finite SNR, fundamentally different from existing ones. Moreover we provide a robust iterative interference alignment algorithm and corresponding feedback strategies achieving the optimal scaling laws. The main idea is that instead of sending the complete channel matrix to each base station each user fixes a receive filter in each iteration and feeds back a quantized version of the effective channel. Finally we underline the findings with simulations for the proposed system.