Prioritized Delay Optimization for NOMA-Based Multi-UAV Emergency Networks

In emergency networks, it is significant to jointly consider users' priorities and communication delays for post-disaster users. To this end, we propose an essential metric by integrating these two factors, namely prioritized delay. Moreover, we formulate a min-max prioritized delay problem for non-orthogonal multiple access (NOMA)-based multi-unmanned aerial vehicle (UAV) emergency networks, where the powers of all users are concurrently optimized to balance prioritized delay among users while suppressing multi-user interference introduced by NOMA and multiple UAVs. The problem is found to be non-convex, and we propose the substitution-decomposition-scaling-block successive upper bound minimization (SDS-BSUM) algorithm to address this issue. Specifically, the originally-formulated problem is first transformed into a standard form through variable substitution, followed by problem decomposition into single-UAV subproblems through exploiting its separable structure, and the scaling of non-convex constraints into tight convex ones. Finally, the converted problem is fitted into BSUM for solving. Simulation results reveal that compared with the existing schemes, the proposed SDS-BSUM not only reduces the prioritized delay by up to 70%, but also achieves an excellent delay fairness commensurate with users' priorities.