Common Caching Replacement Algorithm for Video-on-Demand System
Fengbin Li, Jun Li, Zhong Hu, Jun Zhou · 2009
As the streaming media files growing larger and larger in size, it inevitably aggravates the network congestion and user perceive latency, to settle problem lots of caching algorithms have been applied in video-on-demand (VOD)system. However, different algorithm is correspondence to an unique caching replacement policy, which limits its applications and the effects are not very satisfied, so in VOD aspect it is necessary to find a common replacement algorithm. In this paper, we propose a Common Caching Replacement Algorithm (CCRA). As a unification replacement algorithm, it sufficiently considers the recent visit and segment size, effectively solves the disk I/O bandwidth bottleneck limitation and improves byte hit rate. We respectively replace the previous caching replacement policy in Uniform segmentation algorithm and Exponential segmentation algorithm with CCRA. From the experimental comparison, Uniform segmentation with CCRA improves almost 5\% in disk reduce ratio, and exponential segmentation with CCRA increases near 8\% in byte hit ratio.