Stochastic Geometry Modeling and System-Level Analysis & Optimization of Relay-Aided Downlink Cellular Networks

In this paper, a tractable mathematical framework for the analysis and optimization of two-hop relay-aided cellular networks is introduced. The proposed approach leverages stochastic geometry for system-level analysis, by modeling the locations of base stations, relay nodes and mobile terminals as points of homogeneous Poisson point processes. A flexible cell association and relay-aided transmission protocol based on the best biased average received power are considered. Computationally tractable integrals and closed-form expressions for coverage and rate are provided, and the performance trends of relay-aided cellular networks are identified. It is shown that coverage and rate highly depend on the path-loss exponents of one- and two-hop links. In the interference-limited regime, in particular, it is shown that, if the system is not adequately designed, the presence of relay nodes may provide negligible performance gains. By capitalizing on the proposed mathematical framework, a system-level and interference-aware optimization criterion of the bias coefficients is proposed. Numerical results confirm the effectiveness of the proposed system-level optimization to enhance the coverage probability in the interference-limited regime. The presence of relays, on the other hand, is shown to have a limited impact on average/coverage rate under the same assumptions.