Characterizing SopCast client behavior

  • Authors:

  • Alex Borges;Pedro Gomes;José Nacif;Rodrigo Mantini;Jussara M. Almeida;Sérgio Campos

  • Affiliations:

  • Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil and Computer Science Department, Universidade Federal de Juiz de Fora, Juiz de Fora, Minas G ...;Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil;Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil and Universidade Federal de Viçosa, Florestal, Minas Gerais, Brazil;Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil;Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil;Computer Science Department, Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil

  • Venue:
  • Computer Communications
  • Year:
  • 2012

Quantified Score

Hi-index 0.24

Visualization

Abstract

Live streaming media applications are becoming more popular each day. Indeed, some important TV channels already broadcast their live content on the Internet. In such scenario, Peer-to-Peer (P2P) applications are very attractive as platforms to distribute live content to large client populations at low costs. A thorough understanding of how clients of such applications typically behave, particularly in terms of dynamic patterns, can provide useful insights into the design of more cost-effective and robust solutions. With the goal of extending the current knowledge of how clients of live streaming applications typically behave, this paper provides a detailed characterization of clients of SopCast, a currently very popular P2P live streaming application. We have analyzed a series of SopCast transmissions collected using PlanetLab. These transmissions are categorized into two different types, namely, major event live transmissions and regular (or non-event) live transmissions. Our main contributions are: (a) a detailed model of client behavior in P2P live streaming applications, (b) the characterization of all model components for two different types of transmissions in the SopCast application, (c) the identification of qualitative and quantitative similarities and differences in the typical client behavior across different transmissions, and (d) the determination of parameter values for the proposed client behavior model to support the design of realistic synthetic workload generators.