Statistical characterization of P2P-TV services and users

P2P-TV systems will certainly contribute to innovate the broadcast TV concept by allowing ubiquitous access to a practically unlimited number of channels, representing an important step forward in the direction of the Anything/Anyone/Anywhere/Anytime communication paradigm of future Internet applications. Since its deployment, P2P-TV has already attracted millions of users and it is expected to grow exponentially in the near future, both in terms of the number and diversity of their clients. Analyzing the behavior of P2P-TV systems and users and characterizing the generated traffic are fundamental steps to understand and evaluate the potential impacts of these applications on the underlying transport network. In this paper, a set of intensive measurements was conducted in a residential Internet access environment in order to obtain a good insight over the mechanisms that govern data transfer on some of the most important P2P-TV applications (TVants, SOPCast and TVU Player) and understand the most important advances they have suffered in the latest years. Among other relevant findings, the analysis showed that peer location does not seem to be exploited by any of the studied applications and P2P-TV is already a worldwide phenomenon: using the three selected applications during 12 days of experiments, peers from 177 different countries all over the world were contacted, including countries where the interest for English-spoken channels was not so likely.