An Adaptive Genetic Programming Approach to QoS-aware Web Services Composition.
Yang Yu,Hui Ma,Mengjie Zhang
DOI: https://doi.org/10.1109/cec.2013.6557771
2013-01-01
Abstract:Web services are software entities that can be deployed, discovered and invoked in the distributed environment of the Internet through a set of standards such as Simple Object Access Protocol (SOAP), Web Services Description Language (WSDL) and Universal Description, Discovery and Integration (UDDI). However, atomic web service can only provide simple functionality. A range of web services are required to be incorporated into one composite service in order to offer value-added and complicated functionality when no existing web service can be found to satisfy users' request. In service-oriented architecture (SOA), web services composition has become an efficient solution to support business-to-business and enterprise application integration (EAI). In addition to functional properties (i.e., inputs and outputs), web services have non-functional properties called quality of service (QoS) that encompasses a number of parameters such as execution cost, response time and availability. Nowadays with the rapid increase in the number of available web services, a great number of services provide overlapping or identical functionality but vary in QoS attribute values. Due to the huge search space of the composition problem, a genetic programming (GP) approach is proposed in this paper, which aims to produce the desired outputs based on available inputs, as well as ensure that the composite service has the optimal QoS value. Furthermore, an adaptive method is applied to the standard form of GP in order to avoid low rate of convergence and premature convergence. A series of experiments have been conducted to evaluate the proposed approach, and the results show that the adaptive genetic programming approach (AGP) has a good performance in finding a valid solution within low search time and is superior to the traditional approaches