Approximating the interference distribution in large wireless networks

The spatial locations of transmitters play a cardinal role in evaluating the aggregate interference and hence the performance of large wireless networks. The most widely used approach for modeling the network geometry is the Poisson point process (PPP), mainly due to its analytical tractability. However, point process models with repulsion and inhibition, such as the Matérn hardcore process, the Strauss process, the Ginibre point process, and the Poisson cluster process, are more accurate and realistic models than the PPP for actual heterogeneous cellular network (HetNets) deployments. Nevertheless, the limited analytical tractability of the aforementioned point processes makes it difficult or even impossible to obtain closed-form analytical expressions for network performance. In this paper, we provide a framework for approximating the interference distribution in wireless networks with spatial randomness. Specifically, the distribution of aggregate interference is approximated by known probability density functions, as a means to provide simple and tractable expressions for key performance metrics, i.e. coverage probability, area spectral efficiency, and average rate. We show the effectiveness of the proposed approach in analyzing large wireless networks with nodes exhibiting attraction and inhibition.