Fitness metrics for QoS-Aware web service composition using metaheuristics

Abstract

Quality of Service (QoS)-aware composition ofweb services is significant field of research, moving towards heuristic solutions based on soft computing techniques. However, the existing solution are specific to one or two QoS factors mainly response time and reliability. According to the real-time practices, the assessment of services by one or two QoS factors is impractical. Moreover these soft computing approaches have the computational complexity as O(n2), due to the exponential increase in the number of web services that are available to select. In this context we propose two fitness metrics using metaheuristics that help in assessing web services and composition fitness, resulting in the computational complexity of web service composition as O(n ∗ log(n)). Our experiment results indicate the significance of the proposed approach towards scalable and robust QoS-aware service composition.