Energy-Aware Management in Multi-UAV Deployments: Modelling and Strategies

Area of Science:

• Computer Science
• Network Engineering
• Robotics

Background:

• Unmanned Aerial Vehicles (UAVs) are increasingly used in civilian applications, including providing network services.
• A key challenge for multi-UAV systems is the limited battery life of individual UAVs, restricting mission duration.
• Maintaining continuous connectivity for ground users requires addressing the UAV battery life-time versus mission duration discrepancy.

Purpose of the Study:

• To introduce and evaluate a UAV replacement procedure for sustained connectivity in multi-UAV aerial networks.
• To address the NP-hard practical UAV replacement problem in moderately large UAV swarms.
• To assess the effectiveness of heuristic approaches for solving the UAV replacement problem.

Main Methods:

• Formulation of the practical UAV replacement problem.
• Proposing the Betweenness Centrality Heuristic Algorithm (BETA), a graph theory-based approach.
• Evaluating BETA's performance against optimal solutions and existing methods across various scenarios.

Main Results:

• The UAV replacement problem is proven to be NP-hard, indicating exponential complexity for optimal solutions.
• The proposed BETA algorithm generates solutions with high similarity (up to 99%) to the exact optimal results.
• BETA demonstrates significant improvements over two ground-truth solutions, particularly in resource-constrained environments.

Conclusions:

• Heuristic approaches, like BETA, are suitable for addressing the practical UAV replacement problem in multi-UAV systems.
• BETA offers an efficient and effective method for managing UAV battery limitations and ensuring persistent network connectivity.
• The algorithm shows promise for real-world applications requiring long-duration aerial network services.