Fuzzy Filtering-Based Segment Grouping for User-Centered Multimedia Streaming Service in P2P Distribution Mobile Networks

Abstract

Segment-based proxy caching schemes are recently proposed to provide direct services of required media objects from the proxy without the help of media servers in P2P distribution mobile networks. These segment-based schemes, however, have such problems as jitter latency, congestion, interference, and re-transmission due to large data size, startup latency, and continuous synchronization of media objects. Thus this paper proposes a new FFSG (Fuzzy Filtering-based Segment Grouping) scheme for user-centered media streaming services to solve problems. The proposed scheme manages media objects in segments of FPRB (Fixed Partition Reference Block) or VPRB (Variable Partition Reference Block). Media objects are grouped as blocks of same fixed size in FPRB and as blocks of variable size in VPRB. The proposed scheme used semantic relationship for the grouping of divided media objects and used fuzzy relevance to filter out less relevant segments. The simulation result showed the improvements in average response ratio, cache hit ratio, startup latency rate, and service quality of streaming service.