Joint Task Offloading and Resource Allocation with Data Caching in UAV-Aided Mobile Edge Computing Networks for...

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Growing demand for low-latency, energy-efficient processing for applications like AR, VR, and self-driving cars.
• Mobile Edge Computing (MEC) shifts computation to the network edge.
• Unmanned Aerial Vehicles (UAVs) enhance MEC with flexible deployment and mobility for high-demand scenarios.

Purpose of the Study:

• To address the challenge of jointly optimizing task offloading, resource allocation, and data caching in UAV-aided MEC networks.
• To propose an integrated framework, JORC, for efficient task management and resource utilization.
• To reduce communication overhead and computational workload through intelligent data caching.

Main Methods:

• Formulated offloading and resource allocation as a Markov decision process.
• Employed the soft actor-critic reinforcement learning algorithm for problem-solving.
• Implemented a dynamic, adaptive caching strategy using a hybrid Least-Frequently Used (LFU) and Least-Recently Used (LRU) policy.

Main Results:

• The JORC framework demonstrated substantial reductions in latency, energy consumption, and overall system cost.
• Achieved a higher successful task completion ratio compared to existing baseline approaches.
• Validated the effectiveness of the integrated approach in managing limited storage and reducing redundant computations.

Conclusions:

• The proposed JORC framework effectively optimizes UAV-aided MEC systems by integrating task offloading, resource allocation, and data caching.
• The intelligent caching strategy significantly minimizes computational redundancy and improves system efficiency.
• The study provides a robust solution for supporting latency-sensitive applications in edge computing environments.