Resource Allocation Strategy of the Educational Resource Base for MEC Multiserver Heuristic Joint Task

Area of Science:

• Computer Science
• Educational Technology
• Network Engineering

Background:

• Educational resource databases face challenges in efficient resource allocation.
• Mobile Edge Computing (MEC) offers potential for optimizing distributed resource management.
• Existing resource allocation schemes may not adapt well to dynamic content popularity.

Purpose of the Study:

• To analyze the application of MEC multiserver heuristic joint tasks for educational resource database allocation.
• To develop and evaluate a mathematical model for optimizing educational resource allocation via MEC.
• To assess the performance of a DQN-based unloading scheme in terms of load balancing, delay, and calculation cost.

Main Methods:

• Construction of an educational resource database scenario.
• Development of a mathematical model incorporating local execution, unloading execution, and resource allocation strategies.
• Implementation and comparison of a Deep Q-Network (DQN)-based unloading scheme against benchmark schemes (random, SNR-based).

Main Results:

• The DOOA scheme demonstrates good performance in calculation cost and timeout rates.
• The DQN-based unloading scheme shows superior performance in balancing load, reducing delay, and minimizing calculation cost compared to benchmarks.
• The proposed scheme effectively adapts caching strategies to changing content popularity, improving cache hit rates.

Conclusions:

• MEC multiserver heuristic joint task allocation provides a more reasonable and optimal solution for educational resources.
• The DQN-based approach enhances adaptability to dynamic network conditions and content popularity.
• The developed model and scheme contribute to efficient and intelligent resource management in educational databases.