Spatial coverage cross-tier correlation analysis for heterogeneous cellular networks
In the search for improved coverage and capacity, cellular networks are currently undergoing a major transformation. A thoroughly planned architecture comprised of macrocells served by large-coverage expensive base stations (BSs) is evolving toward a muchmore heterogeneous architecture where the macrocell network is underlaid by one or several tiers of small cells deployed in an irregular and unplanned fashion using universal frequency reuse. Major challenges of this new scenario are the problems of intercell interference (ICI) and cell association. In this paper, these two problems are tackled in a novel way by analyzing and exploiting the inherent spatial cross-tier coverage correlation due to the cochannel ICI for a two-tier network. A mathematical framework for the representation of the two-tier coverage maps and their correlation is developed based on the spatial statistical properties of the signal quality measurements reported by the users to the base station. Several semivariogram-based estimation models are applied and cross-validated. Furthermore, a closed-form expression for the cross-tier coverage correlation function depending only on the estimator's parameters is obtained. In addition, a practical application of this framework is proposed. Cross-tier correlation information is exploited in the design of a new cell association policy based on cell-specific biasing for small cells. Numerical results show that the mathematical framework can provide accurate representations of the coverage fields and their correlation. Moreover, the performance of our proposed correlation-aware cell association policy is shown to be promising enough to encourage further research in this direction.