Bp Neural Network-Based Web Service Selection Algorithm

Web service is a service-oriented architecture technology, it provides services through standard web protocols to ensure application services can be interoperated on different platforms. With the improvement in computational complexity and the people’s demand for application services, the number of web services providers began to surge, web services with the same or similar functions may be provided by a number of different web service providers, these services has different quality of service. So, how to choose the most appropriate service from all of the same or similar web services to provide to users had become a key issue in the web services environment, this issue need to be resolved soon.To solve the problem of web service selection, this paper established a web service selection model, the model consists of three key components:service registration centers, service requestor and service provider. Services provided by service provider is described with WSDL language, the XML form language can facilitate the procedure of search, registration and call, service requestor use the SOAP protocol to call services. Web registration centers involved the quality of web service selection, evaluation controller, the core of the controller is based on the error back propagation (BP) neural network algorithm. Neural network controller can choose the most appropriate service from web services with similar quality of service.BP neural network is a multi-front network of one-way propagation, commonly formed by three or more layers, the neural network used in this paper consists of input layer, hidden layer and output layer. The BP neural network is one of neural networks that most widely used currently. The BP neural network can learn and store large amounts of input-output mapping functions model without prior knowledge of the equation of this mapping. Neural network can remember these mappings, and use those to fit in new data. Slow convergence and the oscillation in learning process exist in the traditional BP neural network algorithm. But the novel BP neural network algorithm proposed in this paper has overcome these deficiencies in training process largely.Use the web service selection model proposed in this paper to establish selection model for the simulation of the power supply room scene, and apply the novel BP neural network algorithm to select UPS with the best service quality. Compare the novel algorithm with the traditional BP algorithm in network training and UPS services selecting, the result verified that the novel BP algorithm can converges faster and can select the better quality of UPS service.