DUCTM: An Online Resource Allocation Algorithm for Throughput Maximization in Cooperative NOMA-Enabled WPT-MEC Networks

This paper addresses the problem of throughput utility maximization in a non-orthogonal multiple access (NOMA)-enabled wireless power transfer mobile edge computing (WPT-MEC) network with dynamic task arrivals and user cooperation. To promote fairness and effectively handle random task arrivals and time-varying channels, we model the system utility as a nonlinear function of time-averaged throughput. We then formulate a stochastic optimization problem aimed at maximizing utility while strictly maintaining sensor queue stability. By leveraging the Lyapunov optimization framework, the long-term network-wide utility maximization is decomposed into efficient, slot-wise convex subproblems that operate online without requiring prior knowledge of future task arrivals or channel states. We develop a Dynamic User Cooperation Throughput Maximization (DUCTM) algorithm that enables adaptive resource allocation and cooperative computation offloading in an online manner. Theoretical analysis establishes a provable [O(1/V),O(V)] trade-off between utility optimality and queue backlog. Extensive simulations demonstrate that our approach consistently outperforms baseline methods, providing robust and stable performance even under bursty traffic and highly dynamic environmental conditions.