Congestion-aware user-centric cooperative base station selection in ultra-dense networks

In next generation cellular networks, small cell base stations (BSs) with ultra-dense deployments are likely to cooperate together to manage interference and provide users with high data rates. But users may still experience large drops in throughput due to lack of available resources even with high received signal strength. In this paper, we propose a congestion-aware user-centric cooperative BS selection method in ultra-dense networks. Each user chooses its cooperative BSs based on not only the received signal-to-interference-pulse-noise ratio (SINR), but also the predicted load of BS, which is called congestion metric. The congestion metric is calculated by each BS according to the current load of BS and the predicted load brought by the users who are ready to access the BS. In order to maximize the objective function which considers both received SINR and congestion metric, we propose two suboptimal greedy algorithms which have much lower computational complexity than exhaustive search procedure. Simulation results show that the user average data rate and system spectral efficiency of proposed congestion-aware schemes are much higher than that of the SINR-based user-centric clustering scheme. It is also shown that the greedy algorithms perform closely to exhaustive search procedure, while having much lower computational complexity.