论文部分内容阅读
提出的 ROC(resist- overload capability)缓存管理策略 ,包括缓存准入策略和替换算法 ,解决了变码率视频服务器间隔缓存的缓存管理问题。 ROC缓存准入策略基于统计缓存准入策略 ,能提供概率的缓存服务质量保证和较高的缓存利用率。实验结果表明 :在典型系统配置下 ,ROC可以提高约 2 5 %的系统吞吐量 ;相对确定性缓存准入策略和STP- L缓存替换算法 ,可以多服务约 17%的视频流 ,平均缓存利用率也要高出约 38%。仿真实验结果证明了 :ROC充分发挥了间隔缓存算法优势 ,是一种非常有效的变码率视频服务器缓存管理策略
The proposed ROC (resist-overload capability) cache management strategy includes a cache admission policy and a replacement algorithm to solve the cache management problem of variable-rate video server interval cache. The ROC cache admission policy is based on a statistical cache admission policy that provides probabilistic caching quality of service assurance and high cache utilization. The experimental results show that ROC can improve system throughput by about 25% under the typical system configuration. With the relative deterministic cache admission policy and STP-L cache replacement algorithm, about 17% of the video streams can be served and the average cache utilization The rate is also about 38% higher. Simulation results show that: ROC give full play to the advantages of the interval buffer algorithm is a very effective variable rate video server cache management strategy