Development of information systems and services over peer-to-peer overlay networks

Abstract

In the current generation of the Internet, an overlay network is a virtual network of nodes and logical links built over the Internet. Therefore, overlay networks can be viewed as a “middle layer” above the basic Internet protocols but below the application layer. To understand why overlay networks are important and the role they play, it is worth considering the different reasons which have motivated their emergence. Firstly, overlay networks may exist to support the special requirements of a particular sort of applications or a specific community of users. Second, and related to the above, overlays may play an important role in the dynamic evolution of Internet technology. In this respect, the traditional architecture to develop distributed applications was based on the client-server model. This model was most appropriate when there were no PCs, only work-stations, whose power was far from large and expensive mainframe computers. Today, the PCs used at home and business have a similar performance to the servers at prices substantially lower. Therefore, peer-to-peer (P2P) emerge as an alternative to take advantage of the full processing capacity, storage and excess bandwidth of interconnected computers. Basically, a P2P network refers to a network that does not have fixed clients and servers, but a series of nodes that act both as clients and servers for other network nodes. This network model contrasts with the traditional client-server model used in Internet applications. Thus, all nodes behave in the same way and are able to perform the same type of operation, although they can differ in local configuration, processing speed, bandwidth of the connection to the Internet, and storage capacity. Regarding to the mechanisms and protocols in which P2P networks are based, we can clearly distinguish two necessary components: I) Mechanisms for locating resources (such as users, services or content) and II) Communication mechanisms to access resources (such as the establishment of a call or the download of a file). Typically, P2P networks are classified according to the mechanisms for locating resources in unstructured networks, superpeer-based networks and structured networks. The properties of P2P networks make them very suitable for the development of new information systems and services. In this work, the main characteristics of the previously presented P2P networks are studied in depth. In addition, it is also analyzed their suitability for the development of specific information systems and services over them. On this matter, one of the most important characteristics of unstructured P2P networks, which is that they present a flat topological architecture, where all the nodes belong to the same level, makes this kind of networks specially suitable for the implementation