Improvement Of Service Composition Method Based On CANEs For Active Network

With the evolvement of the Internet, the number of users and applications using Internet increases very quickly. The current Internet architecture lacks flexibility. The new network services are always postponed by tedious standardization process. In a traditional network, it is difficult to deploy new network protocols and services, and this problem could not be overcome by itself. The active network makes change through adding computing to network node. Active network, however, is a feasible way. Traditional networks can only transport data passively according to the predefined protocols. Active network provides a novel way to solve the problems of traditional networks. Active network is programmable network and the user can insert customized programs into the switcher of the network to perform customized computations on messages flowing through them. Programs can be inserted into the nodes of active networks to perform customized computation. In this paper, on the basis of describing the background and the motivation of designing active network, the new architecture is presented in detail, its effect on network architecture is discussed.1. Bowman is an extensible platform for active network. It layers active network functionality in user-space software over Unix or Linux. There are several elements of Bowman that are of interest; namely, channels, a-flows, the state store, and so on.2. The CANEs EE has two goals: to support the development of active applications which require reasonable forwarding performance, and to provide a framework for the modular construction of services.Therefore the core functionality of active networks is to identify transfer load and execute active applications to perform the computation required by the end users through the understanding of the platform of Bowman/CANEs. More specifically, the programming model used in the CANEs EE comprises two parts: a fixed part (the underlying program) that represents the uniform processing applied on every packet, and a variable part (the injected program) that represents customized functionality on the packets. The specific points in the underlying program where an injected program may be executed are called slots. Composition of services is achieved by selecting and available underlying program and then specifying a set of injected programs for customization. Injected programs may be node-resident or loaded from a remote site. But this is the fault we considered. The service deployment lacks flexibility. We have a new idea of wiping of the underlying program.This paper presents an improved service composition method based on CANEs EE and verifies its true on the platform of Bowman and CANEs through an active reliable multicast based on AER/NCA protocol. So we conclude that it provides an effective and flexible service deployment mechanism.