A fuzzy algorithm for dynamically adaptive multimedia streaming

Abstract

The QoS-aware delivery model of multimedia is an interesting research area. The wireless networking systems connecting mobile clients and media servers have created the paradigm of mobile multimedia. In mobile multimedia systems, the media delivery model has to maintain two diagonally opposite objectives, such as maintaining QoS of playback and saving energy consumption of the mobile devices. The traditional pull, push, and the hybrid push-pull models of media delivery are not completely suitable to offer consistent QoS of playback while saving the energy consumptions at mobile devices. This article proposes a novel multimedia delivery system based on the Fuzzy Adaptive Buffering (FAB) algorithm using pull model. The FAB algorithm employs a fuzzy inference technique and dynamically adapts to the execution environments. The experimental results illustrate that the FAB algorithm successfully adapts to dynamic execution contexts while maintaining playback-QoS and saving the energy consumption of the mobile clients by keeping the data prefetching thread in the sleeping mode from 31.44% to 97.4% of streaming time depending on the execution environments.