1st Annual Conference on Broadband Networks

Research Article

Power-controlled media streaming in the interference-limited wireless networks

  • @INPROCEEDINGS{10.1109/BROADNETS.2004.68,
        author={Yan  Li and Nicholas Bambos},
        title={Power-controlled media streaming in the interference-limited wireless networks},
        proceedings={1st Annual Conference on Broadband Networks},
        publisher={IEEE},
        proceedings_a={BROADNETS},
        year={2004},
        month={12},
        keywords={},
        doi={10.1109/BROADNETS.2004.68}
    }
    
  • Yan Li
    Nicholas Bambos
    Year: 2004
    Power-controlled media streaming in the interference-limited wireless networks
    BROADNETS
    IEEE
    DOI: 10.1109/BROADNETS.2004.68
Yan Li1,*, Nicholas Bambos1,*
  • 1: Dept. of Electrical Engineering, Stanford University, 350 Serra Mall Stanford, CA 94305-9505
*Contact email: liyan@stanford.edu, bambos@stanford.edu

Abstract

We investigate a power-controlled transmission scheme for streaming media traffic in an interference-limited wireless environment, where many data-packet links (transmissions) coexist with the streamed media link. The transmissions from individual links in the system interfere with each other. The media server (source) transmits encoded media packets to the client (user) over the media link. The client stores the received packets into a fixed-size buffer and plays them out at a constant rate. Due to random and response-induced interference, the source needs to adapt its transmission power and rate to guarantee a certain level of received media quality (QoS) and to minimize the average power (energy consumption), thus to reduce the interference perceived by other links in the system. The QoS is primarily measured by the distortion due to packets missing their playtime deadlines at the receiver. We develop a modelling framework to identify the characteristics of the optimal power/rate control policy. Based on these characteristics, we propose two practical heuristic algorithms designed to operate in realistic interference-limited environments. Through simulation, the proposed algorithms show significant performance improvements when compared to standard benchmarks.