Orchestrating UAVs for Prioritized Data Harvesting: A Cross-Layer Optimization Perspective

Bharath Keshavamurthy,Nicolo Michelusi

2024-04-01

Abstract:This work describes the orchestration of a fleet of rotary-wing Unmanned Aerial Vehicles (UAVs) for harvesting prioritized traffic from random distributions of heterogeneous users with Multiple Input Multiple Output (MIMO) capabilities. In a finite-horizon offline setting, the goal is to optimize the beam-forming design, the 3D UAV positioning and trajectory solution, and the user association/scheduling policy, to maximize the cumulative fleet-wide reward obtained by satisfying the quality-of-service mandates imposed on each user uplink request, subject to an average per-UAV mobility power constraint. With a probabilistic air-to-ground channel model, a multi-user MIMO uplink communication model with prioritized traffic, and a novel 3D mobility model for rotary-wing UAVs, the fleet-wide reward maximization problem is solved via a cross-layer optimization framework: first, K-means clustering is employed to obtain user clusters; then, equipped with a zero-forcing beam-forming design, the positions of the UAVs are optimized via two-stage grid search; next, treating these optimal positions as the graph vertices of a fully-connected mesh, the 3D UAV trajectories (i.e., graph edges) are designed via a learning based competitive swarm optimization algorithm, under an average UAV power consumption constraint, coupled with projected subgradient ascent for dual optimization; consequently, the user association/scheduling strategy is solved via a graphical branch-and-bound method on the underlying multiple traveling salesman problem. Numerical evaluations demonstrate that the proposed solution outperforms static UAV deployments, adaptive Voronoi decomposition techniques, and state-of-the-art iterative fleet control algorithms, with respect to user quality-of-service and per-UAV average power consumption.

Systems and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to optimize the position, trajectory design, and user association/scheduling strategy of a group of rotor - wing unmanned aerial vehicles (UAVs) with multi - input multi - output (MIMO) capabilities within a specific time to efficiently collect data from ground nodes (GNs) with different priority requirements. Specifically, the goal of the paper is to maximize the cumulative reward of the entire UAV formation while meeting the quality - of - service requirements of each ground node's uplink requests, taking into account the average mobile power constraints of each UAV.

The main challenges in the paper include:
1. **Quality - of - service requirements**: Handling different quality - of - service requirements arising from different types of uplink requests in typical deployment sites.
2. **Optimal positioning**: Determining the optimal position of the UAVs considering air - to - ground channel characteristics.
3. **Energy - aware trajectory design**: Designing energy - saving 3D trajectories for UAVs considering their limited on - board energy capabilities.

To address these challenges, the authors propose a cross - layer optimization framework. This framework decomposes the global formation reward maximization problem in an offline centralized setting from the perspectives of the radio layer and the vehicle layer, and solves the optimal position of the UAVs, their energy - saving trajectories, the optimal beamforming design, and the user association/scheduling strategy respectively. This framework is implemented by combining the K - means clustering algorithm, zero - forcing beamforming design, learning - based competitive swarm optimization algorithm, and graph - theory branch - and - bound method, aiming to improve data collection efficiency and quality - of - service while reducing energy consumption.