Interference scaling in multi-cell network with rate guarantees

Heterogeneous networks are expected to become prevalent in the near future due to introduction of close proximity communication in micro cells in conjunction with large-scale coverage by macro cells. In two-tier heterogeneous networks, it is important to analyze performance degradation in macro cells caused by micro cells. For that reason, we analyze the interference to macro cell caused by micro cell single-hop networks where micro cells have QoS requirement, i.e., minimum average requirement of each micro cell. The resource allocation policies considered here are joint power control and node scheduling for downlink and uplink scenarios. We assume that each micro cell operates in orthogonal channels, i.e., no inter-cell interference between micro cells. In this paper, we use extreme value theory to provide scaling laws for the interference to macro cell, when the number of nodes per micro cell N is allowed grows large. For downlink transmission, interference scales according to 1/log(N) rule. A key result is that 1/log(N) scaling behavior exhibited in downlink transmission is transformed into a 1/N behavior in uplink transmission due to additional diversity introduced in interference channel, i.e., the channel between transmitter and macro cell base station.