Research On IOT Lightweight Web Services Composition Based On Hadoop

IOT Web service is the product of Internet of Things Technology and Web Service Technology. As an important part of the information age, it is widely used in various applications of cloud computing platforms in distributed environment. Although the data and operations are various in IOT, the IOT Web services can implemente the data sharing and interoperability. However, the function of a single IOT Web service’s function is limited, when an existing single IOT Web service cannot meet the needs of users, we need to combinate small size of IOT Web services into complicated services, dynamically generating the new large size IOT Web services to complete the user-specified task.Although many researchers have proposed a variety of combined methods of IOT Web services, several problems remain exist. First, the lack of a concrete description of lightweight services, having the description, we could create and publish IOT Web services to reduce the burden of server, avoid the waste of resources, improve concurrency of services composition. Second, the lack of method for demand optimization, the process of services composition pays more attention on personalized needs of users. Research on IOT lightweight Web services composition based on Hadoop must take the needs of users’preferences into account to adapt to the needs of users. Thirdly, the lack of a relatively efficient service search method in the mass of IOT Web services. Fourth, the lack of method to measure the matching degree of service requested by the users and large-grained service based on users’needs. In addition, it also lack of method to measure the complex relationship between the small size services. The paper’s main task is to solve the above problems.To solve the above problems, the paper includes the following contents:(1) The design and description standard of IOT RESTful web services.We find that the existing standards do not apply for certain services, in the contrary, add complexity. Therefore, on the basis of the current national standards, this paper designed and gave IOT REST- style services metadata standards.(2) Information reduction algorithm based on user needs.The algorithm realized the requirement optimization of the IOT RESTful web services under Hadoop.(3) Improve the IOT REST-style services search algorithm.This paper put forword single IOT RESTful service search algorithm and combinated RESTful IOT Web service search algorithm.(4) IOT RESTful service automatic aggregation algorithm based on association rule relations.In this paper, by studying the relationship between services, we picked out the best services with maximum correlation to combine together from mass services.The significance of this paper is as follows. First, for IOT Web services we designed metadata standards, reducing the description of the unnessary content and application. Second, from the users’individualized demand information, we unearthed the user’s preferences, and used the preference aggregation principle to achieve the optimizated services. Third, we introduced the concept of "MC" to quantify the service search and to provide users with flexible search mechanism. Fourth, with polymerization thought as the instruction, and the Internet of things REST services for support, this paper proposed automatic service aggregation algorithm. The design of dynamic polymerization has increased its scalability and provides solutions that interact with the flow of information.