Spectrum Utility Driven WiFi Offloading over NOMA-based Cellular Network Edges

Xi Zhang,Fan Yang

DOI: https://doi.org/10.1109/ciss.2017.7926136

2017-01-01

Abstract:We propose a spectrum utility based WiFi offloading scheme on the edge of cellular networks, implemented with non-orthogonal multiple access (NOMA) techniques. In our proposed scheme, we aim to optimize the spectrum utility, obtained by service providers, through power allocation and offloading ratio. With practical constructions of the cost functions on spectrum, we define the spectrum utility mathematically, and then formulate the problem by considering relevant resource constraints and quality of service (QoS) requirements. To solve the formulated problem, we first employ the Dinkelbach's parametric approach to transform. Then, we decompose the optimization with a sub-optimal manner into two subproblems, i.e., AP-subproblem and BS-subproblem. Applying the convex optimization method and greedy algorithm to AP-subproblem, we effectively obtain power allocations for WiFi AP, and acquire the solution of offloading ratio. Through discussion on BS-subproblem, we then have power allocations for cellular BS. The simulation results finally validate our theoretical analysis, and evaluate the effectiveness of our proposed offloading scheme.