Analysis on the capacity and optimization for D2D underlaying cellular communications using geostatistical methods -

Abstract

Device-to-device (D2D) communication has gained a lot of attention in the last few years, and is considered to be a crucial part of the 4G Long Term Evolution - Advanced (LTE-A) cellular system and Beyond 4G (B4G) systems, introducing flexible, cost and energy efficient solutions. However, one of the limiting factors in such communication scenario is the interference introduced due to the reuse of cellular resources for D2D links, degrading the network’s quality of service (QoS) and the transmission capacity. In this thesis, we consider a non-cooperative deployment for the LTE and D2D links and derive closed-form expressions of the ergodic capacities for both systems based on their Signal-to-Interference Ratio (SIR) distributions. We show that the capacities of both D2D and LTE are concave and convex respectively w.r.t the transmitted power ratio. We then propose two optimization algorithms of the power allocation based on the convex concave procedure (CCCP) and a Geo-statistical approach. The CCCP procedure is used to obtain the optimal transmitted power ratio that satisfies the target Quality of Service (QoS) under limited number of D2D links. In crowded D2D environment, we propose to exploit a spatial statistics technique to estimate the Signal-to-Interference plus Noise Ratio (SINR) and interference distributions based on the spatial correlation of the D2D links and cellular links using Kriging interpolator. The resulting information is then incorporated in the power optimization problem to relax the constraints and provide interference-aware solution for such scenario. The analytical and numerical results demonstrate the effectiveness of the introduced algorithms and show that there exists an optimal power ratio that satisfies the transmission constraints in both D2D and cellular systems.