Characterizing multi-cell cooperation via the outage-probability exponent

Multi-cell cooperation (MCC) is a promising approach for mitigating inter-cell interference in dense cellular networks. To study the MCC performance, existing work typically relies on the over-simplified Wyner-type models that fail to account for mobile spatial statistics, irregular locations of base stations (BSs) and the resultant highly variable path-loss. Unsurprisingly, real-world systems show gains far below those predicted using such idealized models. This paper adopts a stochastic-geometry model for a cellular downlink network with MCC where cells are modeled as a Poisson random tessellation generated by Poisson distributed BSs, these BSs are then clustered using a hexagonal lattice, and BSs in the same cluster mitigate mutual interference by spatial interference avoidance. We analyze the effects of scattering on network coverage as the average number of cooperative BSs, K, increases. For mobiles near the centers of cooperative BS clusters, we show that the outage probability diminishes with increasing K at least sub-exponentially for sparse scattering and following a power law for rich scattering where the exponent is proportional to the signal diversity order. For randomly located mobiles, the outage probability is shown to decrease with increasing K following a power law independent with scattering.