Service selection in service oriented architecture using probabilistic approach and asynchronous queues with interceptor validation

Abstract

In service Oriented Architecture, many services are offered with similar functionality but with different service quality parameters. Thus the service selection using a deterministic approach causes conflicts and inefficient results. We use asynchronous queue to model the service inventory architecture avoiding unnecessary locking of resources and thus allowing a provision to consumers to get their required services without intervening and with temporally decoupled fashion. Actually this kind of service selection strategy is considered in regards with game theory to eliminate fluctuations of queue length. It offers a discrete random service which is equal to some request requested by consumers, it means service can be provided based on probability mass function as a substitute of deterministic decisions for selecting a proper service provider as of the consumers. Once the request is taken out from the queue, it is delivered to the interceptor that has validation and sanitization module. It thus reduces the peak queue length and reduces periodic fluctuations in the queue length.