Intelligent computational methods for multi-unmanned aerial vehicle-enabled autonomous mobile edge computing systems.

Area of Science:

• Computer Engineering
• Robotics
• Wireless Communications

Background:

• Mobile edge computing (MEC) systems are increasingly deployed using unmanned aerial vehicles (UAVs) to serve user devices (UDs).
• Optimizing UAV trajectories is crucial for minimizing energy consumption in autonomous MEC systems.
• Traditional methods struggle with the complexity of joint optimization of stop point deployment, user association, and UAV path planning.

Purpose of the Study:

• To propose a novel algorithm for optimizing multi-UAV trajectories in MEC systems.
• To minimize the overall energy consumption of autonomous MEC systems.
• To address the complex problem of joint optimization for stop point deployment, user-UAV association, and trajectory planning.

Main Methods:

• A variable-length trajectory planning algorithm (VLTPA) is proposed, comprising three phases.
• Phase 1: Genetic algorithm (GA) with variable-length individuals for updating stop point (SP) deployment.
• Phase 2: Close rule for addressing the association between UDs and SPs.
• Phase 3: Multi-chrome GA for jointly handling SP-UAV association and UAV ordering.

Main Results:

• Extensive experiments were conducted on eight instances with 60 to 200 UDs.
• The proposed VLTPA demonstrated superior performance compared to existing state-of-the-art algorithms.
• The algorithm effectively optimizes UAV trajectories for reduced energy consumption.

Conclusions:

• The VLTPA provides an effective solution for optimizing multi-UAV trajectories in MEC systems.
• The proposed method successfully minimizes energy consumption in autonomous MEC operations.
• The VLTPA offers a significant advancement over traditional approaches for complex trajectory planning problems.