Subchannel Assignment and Power Allocation for Time-Varying Fog Radio Access Network with NOMA

Abstract

To satisfy future wireless network's requirements of huge capacity, ultra low delay and supermassive connectivity, it is urgent to study novel wireless communication network architecture and technology. Fog-computing radio access network (F-RAN) is a newly developed network paradigm, in which the edge devices perform the storage, communication, control, configuration and management. Meanwhile, non-orthogonal multiple access (NOMA) has been considered to be a hopeful multiple access mechanism for future radio access networks (RANs). The distinctive feature of NOMA is to assign the same channel for various users simultaneously by utilizing the power domain. With the deployment of NOMA, F-RAN can further promote system throughput, time latency, access capability and spectrum efficiency. The coordination between NOMA and F-RAN provides powerful support to extensive application of augmented/virtual reality (AR/VR), vehicular networking, intelligent medical and other emerging applications. In this paper, we focus on a noncooperative game resource optimization to decrease the complexity while the dynamic optimization problem is decoupled to subchannel assignment and power allocation. Matching theory is applied to propose a subchannel assignment scheme. Then, we convert and decouple the power allocation problem into three subproblems and solve them separately at each slot. Simulation results illustrate the potency of the proposed noncooperative resource allocation scheme.