.png){kind=icon}
.png){kind=icon}
Return
Distributed Dynamic Consensus (DDC) Protocol for Multi-UAV 3D Trajectory Planning and Resource Allocation
.png){kind=icon}
PRE
AI
DOI：10.1016/j.vehcom.2025.100969.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In a Multiple Unmanned Air Vehicle (m-UAV) system, employing a centralized communication approach poses many challenges such as communication range limitations, energy efficiency, latency, etc. due to limited UAV resources. On the other hand, a distributed consensus approach has the ability to overcome these limitations and possesses numerous advantages if appropriate coordination mechanism among the UAVs is employed. Therefore, in this paper, we investigate joint optimization of 3D trajectory and UAV resources using a distributed consensus approach. We assume that User Devices (UDs) compute a portion of the task locally and offload the remaining part to the nearby Mobile Edge Computing (MEC) based UAV. Considering UAV dynamics and environmental constraints, a Deep Deterministic Policy Gradient (DDPG) is presented based on Distributed Dynamic Consensus (DDC) approach that utilizes consensus theory for distributed computing. We classified DDC into three cases namely; Distributed Velocity Consensus (DVC), Distributed Error Consensus (DEC), and Distributed Dynamic Velocity Consensus (DDVC). The performance of all three cases based on cost percentage (cost is the sum of normalized time delay and normalized energy consumption) and observed that DEC achieves minimum cost i.e., 40.62 whereas DVC and DDVC settled at 48.18 and 44.06 respectively. We further investigate the performance of DEC in partially connected, moderately connected, and fully connected networks. With centralized and autonomous decision-making scenario as a benchmark, results show that the DEC in the partially connected scenario converges faster with a lower cost.
Journal
IF:
6.5
Papers:
786
Citations:
3.2K
Organization
View more organizations