Device-to-device modeling and analysis with a modified Matern hardcore BS location model

Device-to-device (D2D) communication is emerging as a potentially attractive way to increase cellular network capacity and flexibility. However, as we describe, analyzing the performance of D2D schemes is non-trivial. In this paper, we consider a D2D overlaid cellular network model, that incorporates many of the leading models for the D2D links and access control. We show that using a Poisson point process (PPP) model for the base stations (BSs), and also a modified Matern hardcore point process (MHC) that can capture BS repulsion, we can extend tools and ideas from stochastic geometry to give compact expressions for important performance metrics, including outage probability. Extending the scope of stochastic geometry is itself an important goal, as this has proven difficult in the past. This allows a tractable approach to understand the performance of D2D overlaid cellular networks. Our simulation results subsequently demonstrate that both the PPP and the modified MHC give good approximations of true BS deployments. In particular, our simulations show the modified MHC process is more accurate than the PPP BS model, at the cost of some additional computational effort.