Space Based QoS Estimation, Storage Methodology and Meta Model of a Web Service

Ravi Shankar Pandey ； Preeti Singh

《International Journal of Information and Computer Science》 5卷 (2016/03) Pp. 49-60

https://doi.org/10.14355/ijics.2016.05.006

摘要

Internet services have changed performance of the business organization to facilitate automation of business processes in the same organization or different organizations. These business processes are implemented by a web service or set of web services. These services are going to be composed automatically. These automatic compositions may suffer from failures. These failures may be occurred due to functional dissimilarity or non-functional requirement which are also called as Quality of service attributes (QoS). These QoS attributes contribute in service selection to minimize the service failures. These QoS attributes are either time based/space based. The space based Qos attribute is location affinity and is also responsible in failure of service composition. These QoS attributes (time based/space based) can be estimated from the history of the usage of web services. In this paper, we have proposed a model which computes the space based QoS attribute (location affinity) from the usage history of the web service which contains information about location. We have proposed a probabilistic model which is based on Hidden Markov Model (HMM) which includes the role of the hidden states of the web service in computation of a location affinity (Space based QoS attribute). We have also argued that this location affinity is not a single value but it is a list. These set of values of a location affinity are stored in a proposed data structure which we have called as Affinity Tree. We have presented an algorithm for computing affinity of the cluster. A Meta model of a web service which includes space based affinity is also proposed. We are describing all our proposals using an example.

關鍵字

Probabilistic Model ； QoS Factors ； Location Affinity ； Space Base Meta-model