Spatio-Temporal Analysis of D2D Networks in Finite Areas

This correspondence investigates the spatio-temporal performance of device-to-device (D2D) networks in finite areas. The locations of transmitters are modeled as a binomial point process with dedicated receivers at fixed distance. The evolution of discrete-time packets at each D2D pair is modeled as a Bernoulli process. Using the tools from stochastic geometry and queuing theory, we analyze the interactions between the spatial locations of nodes and their temporal traffic dynamic. We first derive the distance-dependent meta distribution of signal-to-interference-plus-noise ratio for both reference receiver and random receiver, based on which the packet throughput and ε-stability region are obtained. Simulations validate the analytical results and reveal the existence of an optimal number of D2D pairs that maximizes the packet throughput.