Distributed Resource Allocation and Computation Offloading Scheme for Cognitive Mobile Edge Computing Networks with NOMA

Mobile edge computing (MEC) has become a key component of 5G, via bringing computation services to the edge of the mobile network to enhance the computational capacity of users. To improve spectrum efficiency, we adopt cognitive radio (CR) technology in the MEC system, which makes the resource allocation and computation offloading problem more complex. In this paper, we consider a multi-MEC and multi-user system combined with CR, where each user has a computation task to complete. To further improve the system efficiency, non-orthogonal multiple access (NOMA) is adopted by users. We formulate the resource allocation and computation offloading problem in our scenario as an optimization problem with multiple constraints which aims to maximize the system utility. Alternating direction method of multipliers (ADMM) is adopted to solve this problem in a distributed and efficient way. Simulation results show that our algorithm could achieve great performance comparing to state-of-art distributed and centralized approaches.