Abstract
The growing demand for low-latency content delivery in indoor environments has driven the need for efficient caching strategies in Mobile Edge Computing (MEC) systems. Visible Light Communication (VLC), combined with MEC, offers a promising solution for enhancing content delivery performance by utilizing high-speed, light-based communication. However, traditional caching strategies that rely solely on content popularity often fail to account for user mobility and preferences, which are critical in dynamic indoor settings. In this paper, a hybrid location-aware caching strategy is proposed, integrating user location prediction with content preference modelling to enhance cache efficiency in MEC-enabled VLC networks. Received Signal Strength (RSS)-based multilateration and Bidirectional Gated Recurrent Unit (BiGRU) models are implied to predict user movement patterns and Enhanced Collaborative Filtering (ECF) to anticipate content requests based on historical interactions. By combining these techniques, the system dynamically caches popular content at VLC-based Next Generation Node B small cells (VLC-gNB) near predicted user locations, improving cache hit ratios and reducing content retrieval latency. Additionally, a comprehensive system model is presented incorporating VLC and mmWave technologies, coordinated by Software-Defined Networking (SDN) technology, to manage content delivery and cache placement. Extensive simulation results demonstrate that the proposed caching strategy significantly improves cache hit ratios and reduces latency compared to traditional methods. This study highlights the importance of incorporating both mobility and user preferences in designing efficient caching mechanisms for next-generation MEC-enabled networks.
DOI: 10.61416/ceai.v27i2.9367