Service Composition And Verification Based On Scene

SOA separates software into distinct services allow developer reuse them in different applications. With the popularity of SOA, the method of service composition has been extensively studied. However, with the increasing available services and enhancement of mutual cooperation between different organizations, we are still faced with the following challenges.1) For an enterprise(organization), when they encounter a similar development requirement, developers need to be repeated in the service discovery, service matching, service composition.2) When the service increases, it becomes very difficult to select a service and handle the services from different sources.Combined with the published research, we give a new service composition and verify method based on scene. We focus on the abstract template and service context which combine to express the requirement. On this basis, according to the service context, we reuse the legacy composition. For the other part we can not reuse, we give a method to construct it with linear logic proof approach. Meanwhile, we convert the linear logic and context to communicating sequential processes so that we can have a simulation for result and give a verification feedback.This article focuses on the following points:1. Abstract domain and service contextAbstract template is generalized from a set of software with similar requirements. Specifically, is pulled out of a practical application scenario might include which stages, and each stage at least include which functions.In our approach, we give a standard way to collect requirement based on domain template model(DTM) extended from the feature model. Based on defined DTM, we use the feature context and phase context to specify each software requirement, which include specify features and how the feature implement by compositing atomic services and so on.2.We solve the service composition problem based on linear logicAnd our approach of complement a service composition consists of two parts. Firstly, we present how to extract the legacy component to meet the common requirement. And for the requirement we can not achieve by reusing, we give the atomic service a formal description with classic linear logic and construct the corresponding composition service through a CLL proof. We validated our approach based on a case study and give a promising result.3.We solve the interface matching problem with service semanticWe mentioned the case the interface of service composition will appear heterogeneous as the different service source. We summarize and classify heterogeneous situation, and propose appropriate solutions for each type of heterogeneous. Then we point out the complexity of the situation can be broken down into multiple single heterogeneous.4. Verify the result with communicating sequential processesWe convert the linear logic and context to communicating sequential processes so that we can have a simulation for result and give a verification feedback.In this paper, we proposed a method to describe software requirement, and then reuse the exsiting software as much as possible to achieve rapid development purposes. On this basis, we describe the methods for solving interface matching and verify the results of the service composition. Finally, we give a case study in the field of unified communications to verify our method.