Performance Analysis of an Intelligent Association Scheme in Ultra-Dense Networks Using Matern Cluster Process

Future cellular system is converging toward a heterogeneous network mixing Macro cells and high-density small cells clustering around hotspots. Meanwhile, in some hotspots, the density of active users may be inferior to the density of cells, which will bring a severe interference environment and cause waste of spectrum and energy. A more intelligent association scheme is needed to catch up with the dense reuse of spectrum. In this paper, we investigate the downlink association scheme of a user can be served by multiple small cells or only one Macro cell. Consequently, to capture the clustering feature, we model the distribution of random small cells with Matern Cluster Process (MCP) and Macro cells with Poisson Point Process (PPP). Under general assumptions, we derive the distribution properties of interference and based on that we derive an expression for the average downlink rate in two different serving modes. For efficient computation, we develop credible upper and lower bounds for its Laplace functional in different scheme. Then, we study the relation between average downlink rate and the main system parameters, namely: small cells density, clustering number and radius. Numerical simulation confirms the accuracy of analytical results. Our derived model can be used to guide the configuration of ultra-dense heterogeneous networks.