P2P applications have become enormously popular and currently take into account a large majority of the traffic transmitted over the Internet. A unique characteristic of P2P networks is their flexible and robust operation, which is enabled by the peers’ ability to serve as both servers and clients. Thus, P2P networks are able to provide a cost effective and easily deployable solution for sharing large files among participating peers with no significant help from a de facto, centralized infrastructure. Due to these advantages, P2P networks have also recently become popular for multimedia streaming. The requirements for general file sharing and real-time media streaming are very different and thus, we discuss in this chapter solutions for both these applications. We begin the chapter with an overview of various P2P network structures and their advantages and disadvantages. We then present in detail the BitTorrent system, which is one of the most popular file sharing protocols. We then overview existing P2P-based media streaming applications, and discuss mechanisms that have been developed to support such applications. We also discuss state-of-the-art research in P2P networks which is based on several game theoretic approaches.
Top1. Introduction
Peer-to-peer (P2P) networks connect many end-hosts (also referred to as peers) in an ad-hoc manner. P2P networks have been typically used for file sharing applications, which enable peers to share digitized content such as general documents, audio, video, electronic books, etc. Recently, more advanced applications such as real-time conferences, online gaming, and media streaming have also been deployed over such networks. Unlike traditional client-server networks, where servers only provide content, and clients only consume content, in P2P networks, each peer is both a client and a server.
It has been observed that P2P file sharing applications dominate Internet traffic usage. In fact, a wide range of measurements, which were performed in 8 different geographic regions during the years of 2008-2009, show that P2P networks generated most of the traffic in all monitored regions, ranging from 43% in Northern Africa to 70% in Eastern Europe (http://www.pplive.com), and Internet video broadcasting (e.g., AOL broadcast, MSNBC, CBS, etc.) have emerged. These services have become very popular, as they can deliver video to a large number of receivers simultaneously at any given time. In order to reduce infrastructure, maintenance, and service costs, and provide more reliable services, the content providers often implement their services using P2P network.
While several designs for P2P systems have been successfully deployed for file sharing and real-time media streaming, key challenges such as the design of optimal resource reciprocation strategies among self-interested peers still remain largely unaddressed. For example, pull-based techniques (Cohen, 2003., Pai et al., 2005., Zhang et al., 2005) are designed assuming that peers are altruistic and are willing to provide their available data chunks (pieces) whenever requested. However, such assumptions may be undesirable from the perspective of a self-interested peer, which aims to maximize its own utility. Thus, efficient resource reciprocation strategies need to be deployed, which can also provide incentives to the peers for their contributions.
In BitTorrent systems, incentive strategies are based on the so-called tit-for-tat (TFT) strategy, where a peer selects some of its associated peers (i.e., leechers), which are currently uploading at the highest rates, and provides them its content for downloading (Cohen, 2003). This simple strategy is currently implemented in BitTorrent systems, and provides good performance. However, a key disadvantage of this resource reciprocation strategy is that peers decide how to determine their resource reciprocation based on only the current upload rates that it receives from its associated peers, and does not consider how this reciprocation will impact their upload rates in the future. In other words, the resource reciprocation based on the TFT is myopic. Since peers in P2P networks are generally involved in repeated and long-term interactions, such myopic resource reciprocation strategy can result in a suboptimal performance for the involved peers.