Simulation of P2P TV system using OMNeT++

The use of peer-to-peer overlay networks (P2P) to deliver live television on the Internet (P2P TV) is gaining increasing attention. P2P TV systems are growing nowadays at increasing speed as a result of the opening of video content producers to new business models. This process is supported and accelerated by larger bandwidth availability of Internet and better algorithms for compression, reproducing, and receiving video streams. Traditional Internet TV services based on simple unicast approach are restricted to small group of clients. The overwhelming resource requirement makes this solution impossible when the number of user grows to thousands or millions. By multiplying the servers and creating a content distribution network (CDN) the solution will scale only to a larger audience with regards to the number of deployed servers, which may be limited by the infrastructure costs. Other solution is based on multicast protocols, unfortunately their deployment and availability is currently limited. Therefore the use of the P2P overlay networks to deliver live television on the Internet is achieving popularity and has been considered a promising alternative to IP unicast and multicast models. Nodes in the network, called peers, exchange TV content between themselves serving simultaneously as clients and servers. In a results, with increasing number of network peers, the number of servers in the network also increases leading to a smoother exchange process. Consequently, such an approach has the potential to scale with group size, as greater demand also generates more resources. The raising popularity of this solution is confirmed by the amount of new P2P TV applications that have became available: PPLive, SOPCast, TVants and TVUPlayer, Joost, Babelgum, Zattoo, and by the fact that the traffic generated by such applications has recently increased significantly. Although from the users as well as from the broadcasters’ perspective P2P applications may be very useful, network operators may not share this opinion due to their concerns about the capability of asymmetric underlying network to support large number of users acting as servers uploading their content. Therefore it seems important to gain an understanding of the potential influence that P2P TV applications may impose on the underlying network. Since the most widely deployed commercial systems cited above have closed architecture and are proprietary, only an experimental behavioural (blackbox) characterisation of traffic injected by such systems is in general possible. Reverse engineering of such systems may be costly and not give answers to all nurturing questions regarding their performance. To avoid those disadvantages we propose a simulation analysis of a P2P TV system which allows us fast prototyping, provides the possibility to perform large scale experiments, and offers a common reference platform for experimentation.