Measurement-based peer selection for P2P-IPTV services in campus networks

Abstract

P2P-based IPTV has been a gradual hot topic both for the academic institutions and service providers, which offers residential and commercial users digital television services through IP protocol and media sharing among peers at low cost. A rational peer selection algorithm plays an essential role on offering reliable media streaming transmission and excellent user experience. Meanwhile, convenient and effective quality of experience (QoE) assessment mechanism is not only an important evaluation index for IPTV service providers, but also can supply high-efficient performance feedback for them, especially under a limited network resource environment. Our contributions in this paper are three-fold: 1) we confirm the key parameters which affect QoE among P2P live application according to our 6-month trace, and establish a rational peer selection algorithm by virtue of the parameters above; 2) we set up a novel and passive model for QoE assessment representation; 3) during the popularization in campus networks, we approve the effectiveness of our new QoE evaluation method and the improvement of users' video quality caused by peer selection strategy. Whether or not, we hope our work can supply reliable references for IPTV service providers, moreover, actual improvement for future IPTV deployment.