UPR-AECS: An Active Edge Caching Strategy Integrating User Preference Representation
Hao Wen, Lei Zhang · 2023
Emerging edge caching solutions deploy edge servers (ES) at small base stations (SBS) and store content within these edge devices to offer caching services to end users. This approach aims to improve the quality of service (QoS) for users while reducing backhaul traffic. However, the heterogeneous edge servers with limited storage resources and dynamic user content requests present challenges for edge caching strategies. To address these challenges, we proposed UPR-AECS, an active edge caching strategy that leverages the user preference representation model for collaborative edge caching. This enhancement improves the prediction of dynamic user content preferences, leading to more precise and efficient caching decisions. The primary goal is to maximize the cache hit rate by considering user preferences. Using this method, the proactive edge caching strategy pre-caches expected future content requests to enhance response times. The experimental results show that UPR-AECS outperforms the compared baseline methods in cache hit rate, cache utilization efficiency, and reduction in user content retrieval latency.