Reactive Video Caching via long-short-term fusion approach
This addresses the challenge of inefficient video caching in network architectures, offering a solution that enhances caching performance, though it appears incremental as it builds on existing learning-aided methods.
The paper tackles the problem of balancing long-term history and short-term events in video caching replacement algorithms by proposing LA-E2, a long-short-term fusion approach, which achieves state-of-the-art performance with improvements of 17.5%-68.7% in total hit rate, especially for small cache sizes.
Video caching has been a basic network functionality in today's network architectures. Although the abundance of caching replacement algorithms has been proposed recently, these methods all suffer from a key limitation: due to their immature rules, inaccurate feature engineering or unresponsive model update, they cannot strike a balance between the long-term history and short-term sudden events. To address this concern, we propose LA-E2, a long-short-term fusion caching replacement approach, which is based on a learning-aided exploration-exploitation process. Specifically, by effectively combining the deep neural network (DNN) based prediction with the online exploitation-exploration process through a \emph{top-k} method, LA-E2 can both make use of the historical information and adapt to the constantly changing popularity responsively. Through the extensive experiments in two real-world datasets, we show that LA-E2 can achieve state-of-the-art performance and generalize well. Especially when the cache size is small, our approach can outperform the baselines by 17.5\%-68.7\% higher in total hit rate.