Optimal Task Assignment and Power Allocation for Downlink NOMA MEC Networks

The application of downlink non-orthogonal multiple access (NOMA) in multi-access edge computing (MEC) systems can significantly improve the spectrum efficiency and reduce the delay. In this paper, we investigate the optimal resource allocation to minimize the energy consumption in downlink NOMA MEC networks. The formulated problem is nonconvex. A bilevel programming method is proposed to obtain the optimal solution. By exploiting the monotonicity and convexity properties of subproblems, the optimal closed-form expressions of offloading task ratios and offloading power are derived to achieve the minimum energy consumption in the downlink NOMA MEC network. Moreover, the analysis of the optimal solution indicates different offloading schemes, i.e. pure NOMA and OMA, which are determined by the energy consumption efficiency (ECE) of offloading phase and computing phase at the MEC servers. Simulations demonstrate the optimality and low complexity of the proposed scheme.