Abstract
Perhaps the most popular feature offered by Peer-to-Peer (P2P) networks is the possibility of having several replicas of the same content distributed among multiple nodes. Among the advantages of P2P networks, we emphasize the property of robust fault tolerance. Nevertheless, there are also some disadvantages (e.g. those derived from their decentralized and self-organized nature) such as the complexity of network management and the existence of new security vulnerabilities. In fact, a significant challenge for P2P file sharing systems is maintaining the correctness and consistency of their global data structures and shared contents, as peers independently and unpredictably join and leave the system. Similarly, access control to the shared resources is a noteworthy problem for users of such systems. Most of the difficulties found to apply classic security solutions (e.g. authentication and authorization services) are just related to the impossibility of deploying a public key infrastructure (PKI) in these env ronments, and new proposals have to deal with avoiding such a centralization by exploring alternative paradigms, which often require cooperation among peers. In this chapter, we describe the background and framework of content distribution in pure P2P networks and present the most representative schemes for providing a secure content replication.
TopBackground: Replication In Pure P2p And Ad Hoc Networks
One of the main advantages of P2P systems is their capability to offer replicas of the same content at various locations. Replication is a common approach to improve performance when distributed systems need to scale in number of users and objects in the system, and/or geographical area. Faced with different locations of the same content, an application can grant priority to that which offers a less expensive path (e.g. in terms of bandwidth.) To some extent, replication also guarantees some sort of fault tolerance, since information can be available even if some parts of the network are temporarily disconnected.
Key Terms in this Chapter
Cryptography for Distributed Networks: Cryptographic techniques and algorithms suitable for distributed environments, mainly due to the need of preventing Byzantine faults and replacing a centralized authority as well.
Content Authentication: It ensures the integrity of the shared content, and generally also the identity of its owner (or originator).
Ad Hoc Environment: A self-configuring environment formed spontaneously by mobile (in MANETs) nodes, which union creates an arbitrary topology. It involves the absence of a fixed infrastructure previously established.
Secure Content Replication: Process of sharing information provides guarantees of reliability, fault-tolerance, consistency, accessibility, and security (typically in the form of integrity, authentication and authorization services.)
P2P File Sharing System: In such a system, nodes communicate directly with each other to exchange information. This provides with the capability to offer replicas of the same content at various locations.
Access Control System: Determines the privileges of a particular user, i.e. what a user can do. In general, it includes authentication, authorization and auditory.
Pure Peer-to-Peer Network: A fully decentralized network typically made up of millions of dynamic nodes involved in the process of sharing and collaboration without relying in central authorities.