Joint Optimization of Data Caching and Task Scheduling based on Information Entropy for Mobile Edge Computing

The rise of 5G technology has prompted the emergence of mobile edge computing (MEC), which deploys edge servers at the edge of networks to process task requests. Most work on MEC considers either data caching or task scheduling, but not their combination. However, the two are related, and should be considered together. We do so in this work, with the aim to reduce the average response delay of user task requests. A multi-index collaborative cache replacement (MCCR) algorithm based on the theory of information entropy is proposed, which considers the priority, replacement cost, and size of cached data, so as to replace data with low caching value. An effective task scheduling algorithm, NHSA-MCCR, based on the nearest hierarchical scheduling strategy (NHSA) and MCCR, is then proposed. Experiments are conducted to evaluate the performance of our proposed algorithms, whose results show that MCCR is superior to other caching algorithms in terms of cache hit rate, and that NHSA-MCCR can achieve lower response delay and energy consumption than other scheduling algorithms.