Title: QoS-driven optimisation of composite web services: an approach based on GRASP and analytical models
Authors: Rubens S. Matos; Paulo R.M. Maciel; Ricardo M.A. Silva
Addresses: Center of Informatics, Federal University of Pernambuco, Recife, Brazil ' Center of Informatics, Federal University of Pernambuco, Recife, Brazil ' Center of Informatics, Federal University of Pernambuco, Recife, Brazil
Abstract: Development of applications based on the composition of web services is growing for a large range of domains. Due to the concurrency and synchronisation characteristics of some applications, analytical modelling is helpful for planning and predicting the quality-of-service (QoS) measures of the composite services. This study proposes an optimisation method adapted from the greedy randomised adaptive search procedure (GRASP) and integrated with analytical model solving to find service providers, which will leverage the performance and reliability of a composite web service. The proposed approach finds solutions close to the best-known configurations and is computationally efficient, enabling the fast discovery of a high-quality assignment of providers even for large scenarios with thousands of possible combinations.
Keywords: composite web services; combinatorial optimisation; Markov chains; analytical modelling; performance evaluation; dependability evaluation; QoS prediction; quality of service; reliability; GRASP.
International Journal of Web and Grid Services, 2013 Vol.9 No.3, pp.304 - 321
Received: 06 May 2013
Accepted: 20 Jun 2013
Published online: 02 Aug 2013 *