Interference Coordination: Random Clustering and Adaptive Limited Feedback

Interference coordination improves data rates and reduces outages in cellular networks. Accurately evaluating the gains of coordination, however, is contingent upon using a network topology that models realistic cellular deployments. In this paper, we model the base stations locations as a Poisson point process to provide a better analytical assessment of the performance of coordination. Since interference coordination is only feasible within clusters of limited size, we consider a random clustering process where cluster stations are located according to a random point process and groups of base stations associated with the same cluster coordinate. We assume channel knowledge is exchanged among coordinating base stations, and we analyze the performance of interference coordination when channel knowledge at the transmitters is either perfect or acquired through limited feedback. We apply intercell interference nulling (ICIN) to coordinate interference inside the clusters. The feasibility of ICIN depends on the number of antennas at the base stations. Using tools from stochastic geometry, we derive the probability of coverage and the average rate for a typical mobile user. We show that the average cluster size can be optimized as a function of the number of antennas to maximize the gains of ICIN. To minimize the mean loss in rate due to limited feedback, we propose an adaptive feedback allocation strategy at the mobile users. We show that adapting the bit allocation as a function of the signals´ strength increases the achievable rate with limited feedback, compared to equal bit partitioning. Finally, we illustrate how this analysis can help solve network design problems such as identifying regions where coordination provides gains based on average cluster size, number of antennas, and number of feedback bits.