Today, Peer-to-Peer SIP based communication systems have attracted much attention from both the academia and industry. The decentralized nature of P2P might provide the distributed peer-to-peer communication system without help of the traditional SIP server. However, the decentralization features come to the cost of the reduced manageability and create new concerns. Until now, the main focus of research was on the availability of the network and systems, while few attempts are put on protecting privacy. In this chapter, we investigate on P2PSIP security issues and introduce two enhancement solutions: central based security and distributed trust security, both of which have their own advantages and disadvantages. After that, we study appropriate combination of these two approaches to get optimized protection. Our design is independent of the DHT (Distributed Hash Table) overlay technology. We take the Chord overlay as the example, and then, analyze the system in several aspects: security & privacy, number-of the hops, message flows, etc.
TopIntroduction
Peer-to-Peer (P2P) computing has attracted great attention in both academia and industry. Compare with traditional server-based system architecture in which most of functionality is executed in server side, P2P-based computing allocates computing task to all participating peers. This might eliminate/reduce functionality of server and therefore provides better robustness on system level. Today, P2P computing has been widely implemented in many kinds of networking systems and applications.
In communication field, one of the most well-known P2P applications is Skype (Skype, n.d.), which offers free Voice-over-IP (VoIP) and Instant Messaging (IM) services for computer-to-computer and charged services for computer-to-PSTN. Additionally, Skype service has been extended to mobile world. Many mobile platforms today (e.g. Symbian S60 (Nokia, n.d.), iPhone OS (Apple, n.d.), Android (Android.com, n.d.), Windows Mobile (Microsoft, n.d.), etc), have been embedded with Wi-Fi/3G connection based Skype application. According to eBay statistics (eBay, 2009), the number of Skype users has reached 521 million until Q3, 2009, and it is still growing fast.
However, Skype protocol has been monopolized and made unpublic. Although part of its functionality (e.g. login, NAT traversal, media transfer, codec, etc) has been understood via analyzing Skype network traffic (Postel, 1981; Song et al., 2005), researchers (outside Skype project) are still uncertain about its core technical specification, disadvantages, and required improvement. Besides, Skype does not provide interoperability with other open applications, such as SIP based WLM (Windows Live Messenger) (Windows Live Messenger, n.d.), Yahoo IM (Yahoo!, n.d.), etc. This is partly because of technical difficulty in seamless interconnecting among different protocols, and partly because of unwillingness of cooperation with its competitors.
The success of Skype greatly inspired the research on peer-to-peer based communication systems. Researchers were trying to find an alternative solution where on one hand, decentralized nodes are capable to auto-configure themselves in IP-based Ad-Hoc style; and on the other hand, the designed protocol supports fast location of nodes, optimized route selection, secure and reliable service delivery. Let us take a look at Figure 1, which illustrates an example of P2P based communication paradigm. Each node in the system has a few connections with its neighbours, and these neighbours act as intermediate nodes to deliver requests and responses. A few routing mechanisms, negotiation protocols are supposed to be implemented so that session between source peer (for instance, A in Figure 1) and destination peer (B) can be established in optimized way. The designed protocol is assumed to be an open standard so that everyone could develop applications on it.
However, reality is different. After a few studies, researchers began to recognize that it was not trivial to realize this type communication paradigm. P2P protocol lacks of session level description and negotiation mechanisms, which make some functionalities (such as optimized route selection, media codec negotiation, etc) difficult to achieve. At this moment, Session Initiation Protocol (SIP) comes into the sight.
Session Initiation Protocol (SIP) is a transaction-oriented, text-based protocol that inherits the simplicity from Hypertext Transfer Protocol (HTTP) and Simple Mail Transfer Protocol (SMTP) (Sparks, 2007). It is designed to create, modify, and terminate sessions with one or more participants. Because of its characteristics (e.g. Simplicity, extensibility, flexibility, etc), SIP is chosen by 3rd Generation Partnership Project (3GPP) as the main protocol for the IP Multimedia Subsystems (IMS)-based future All-IP network (The SIP Center, n.d.). SIP to telecommunication systems is regarded as important as HTTP to Internet.