Prioritization Based Task Offloading in UAV-Assisted Edge Networks.

Area of Science:

• Computer Science
• Network Engineering
• Wireless Communications

Background:

• Terrestrial networks face limitations during high traffic, coverage issues, and disasters, impacting service levels and emergency communications.
• Fast-deployable, auxiliary networks are crucial for maintaining connectivity and capacity during such demanding operational conditions.
• Unmanned Aerial Vehicle (UAV) networks present a viable solution due to their inherent mobility and flexibility.

Purpose of the Study:

• To investigate prioritization-based task offloading in UAV-equipped edge networks.
• To develop an optimization model for managing task offloading and minimizing priority-weighted delay penalties.
• To address the NP-hard nature of the assignment problem by proposing heuristic and quasi-optimal algorithms.

Main Methods:

• Constructed an offloading management optimization model to minimize priority-weighted delay against task deadlines.
• Developed three heuristic algorithms and a branch and bound style quasi-optimal task offloading algorithm.
• Conducted simulation-based experiments to evaluate system performance under various operating conditions.
• Contributed open-source code to Mininet-WiFi for independent Wi-Fi mediums, enabling simultaneous packet transfers.

Main Results:

• The proposed optimization model and algorithms effectively manage task offloading in UAV edge networks.
• Simulation results demonstrate the system's performance under different operational scenarios.
• The open-source contribution enhances network simulation capabilities for UAV-based research.

Conclusions:

• Prioritization-based task offloading is effective for supporting critical services in UAV edge networks.
• The developed algorithms provide efficient solutions for the NP-hard task assignment problem.
• UAV networks offer a robust and adaptable solution for enhancing wireless connectivity and capacity in challenging environments.