Energy-efficient UAV movement and user-UAV association in multi-UAV networks
Subhadip Ghosh, Priyadarshi Mukherjee, Sasthi C. Ghosh
TL;DR
The paper addresses energy-efficient UAV trajectory planning and capacity-constrained, mobility-aware user association in multi-UAV mmWave networks, where LoS conditions critically affect performance. It decouples the problem into two optimizations: $P1$, a minimum-weight matching for UAV movement to minimize energy-time cost, and $P2$, a capacity-constrained generalized assignment problem for user-UAV pairing, solved greedily with priority-aware clustering and the Hungarian method. The main contributions are a priority-aware initial clustering to form balanced, throughput-focused UAV clusters, an energy-efficient UAV relocation scheme using the Hungarian algorithm, and a greedy, capacity-aware user assignment that accounts for mobility and delay-priority. The results show reduced unserved users, lower delay variability, and higher energy efficiency compared with benchmark schemes, highlighting practical gains for LoS-dominated mmWave UAV networks.
Abstract
These days, unmanned aerial vehicle (UAV)-based millimeter wave (mmWave) communication systems have drawn a lot of attention due to the increasing demand for faster data rates. Given the susceptibility of mmWave signals to obstacles and high propagation loss of mmWaves, ensuring line-of-sight (LoS) connectivity is critical for maintaining robust and efficient communication. Furthermore, UAVs have limited power resource and limited capacity in terms of number of users it can serve. Most significantly different users have different delay requirements and they keep moving while interacting with the UAVs. In this paper, first, we have provided an efficient solution for the optimal movement of the UAVs, by taking into account the energy efficiency of the UAVs as well as the mobility and delay priority of the users. Next, we have proposed a greedy solution for the optimal user-UAV assignment. After that, the numerical results show how well the suggested solution performs in comparison to the current benchmarks in terms of delay suffered by the users, number of unserved users, and energy efficiency of the UAVs.