The evolution of software services
From communications and banking to automated manufacturing and logistics, software provides a plethora of important services that are increasingly interconnected. In service-oriented computing (SOC), the official definition of service is a collection of capabilities. SOC has become a successful approach to building and maintaining complex distributed systems operating in dynamic and open environments. Software services with the ability to adapt and evolve in a way similar to biological systems are called polymorphic services. In pioneering work, scientists developed a unified approach to evolution and adaptation fully compatible with SOC and of strategic importance to software engineering. The project 'Composition and management of polymorphic services' (COMPOSER) exploited qualitative context variables and fuzzy logic to achieve its goals. Fuzzy logic is used to model context uncertainty. Evolution takes place through evolution of groups of fuzzy rules that represent each contextual stimulus. The rules are modified with each incoming set of data points. Adaptation has been modelled as a reaction to the context stimuli. Thus, service evolution reflects context-based feedback from the service environment. The approach was applied to transport congestion management in urban environments with promising results. The polymorphic service approach developed within the context of COMPOSER falls within the broader context of autonomic service composition and management. It is already gaining acceptance within the SOC community. Continuation of the work is in progress with a national proposal to apply the methodology to improve workers' welfare in industrial environments.
Keywords
Software services, service-oriented computing, polymorphic services, software engineering, fuzzy logic
