Integrating uplink scheduling and power control into block coordinate descent with convergence guarantees

Develop and validate an integration of the joint user scheduling and power control subproblem into the block coordinate descent framework for the uplink multi-UAV enabled wireless network in which ground users transmit to UAVs, ensuring that the resulting alternating optimization algorithm provably converges.

Background

The paper addresses downlink communications in multi-UAV enabled wireless networks, proposing a block coordinate descent algorithm that alternates among user scheduling/association, UAV trajectory, and transmit power control, coupled with successive convex optimization. Convergence is established despite solving approximate convexified subproblems.

In the conclusions, the authors note that a similar alternating optimization approach could be pursued for the uplink (ground users to UAVs). However, they explicitly flag a key unresolved challenge: integrating the solution of the joint optimization of user scheduling and power control into the block coordinate descent framework while guaranteeing convergence. This raises the need for an algorithmic design and analysis that preserves convergence when the uplink scheduling-power control subproblem is embedded in the alternating scheme.