基于焦点修正的流媒体分段缓存策略

针对现有分段算法的缺点,提出一种基于用户随机行为的分段缓存策略,充分考虑不同流媒体对象及同一流媒体对象的不同段落在流行性方面的差异,以段落的流行度作为空间调度的依据,结合前缀缓存的思想,保证缓存效率,且改善字节命中率和延时启动率。实验表明,较之焦点缓存和适应性分段缓存,该策略的字节命中率分别提高了8%和18%左右。

Segment Cache Strategy for Stream Media Based on Focus Modification

Aiming at the shortage of existing segment argorithm,this paper propses a segment cache strategy based on user random behavior.It considers the differences influidities of different stream media objects and different segments in an object.The space schedule is according as the fluidity of segments,it combines with the prefix cache,guarantees the cache efficiency,improves the hit ratio and delay start ratio.The experiment indicates the hit ratio each raises 8% and 18% compared with focus cache and adaptive s...