Research On Service Selection Model With QoS Mechanism In MANETs Environment

As the information technology develops, the concept of Service has been proposed to meet more and more complex business demands, which gets a good development. Service, which is loose-coupled, united interface, brings a good opportunity for network resource, in distributed heterogeneous environment, to achieve dynamic integrated so as to meet complex business demands. As a special wireless network, MANETs (Mobile Ad hoc Networks) has an outstanding application prospects in military and commercial domain. But in the other hand, because of inherent defects, MANETs itself bings a series of challenge to deploy services.With QoS we can directly reflect important factors that restrict the development of service, such as service availability, service security and service performance etc. For the defect of MANETs, there are some difficulities in carrying out QoS in MANETs environments.1. it's difficult to achieve united calculation, because When carrying on calculating QoS, the QoS attributions are divided into two parts:the exact ones and the fuzzy ones;2. In different service application scenes, the different service callers differ in QoS demand3. the recent research works for service are almost in the fixed network, and QoS refers to a set of technology for service, we need to solve the technical issues for service deployed in MANETsIn order to solve the difficulty of putting QoS methods in MANETs. We propose a service selection model with QoS in MANETs.1. We design the mathematic calculation framework to realize unified calculation for two kinds of QoS metric attributions, moreover, we add user preference method in requirement engineer so as to meet the vary QoS needs in service application scenes;2. We propose a set of technology for service in MANETs environment, to accomplish service selection model on the theory side;3. On the basis of service selection model, we achieve the model in the technical side so as to prove the feasibility of service selection model;4. With the two groups of contrasted experiment, we test and verify the better performance in improving service execution failure ratio for the model we proposed, so as to verify the effective for our service selection model.