A genetic algorithm for the adaptation of service compositions

Modeling and Controlling Dynamic Service Compositions

“…By introducing a special structure of a timed automaton, the semantics of classic automata were extended to support parallel execution of services beside their given ability to express interleaving, decisions and looping. The model was already successfully used to represent service compositions that are transformed through the application of genetic operators in order to vary the service compositions's functionalities [17]; further results focusing on the evaluation of those created service compositions will be published soon. Within future work, the expressiveness of guards will be upgraded from propositional logic to first order logic, enabling the additional consideration of effects' properties and interdependencies.…”

Self-evolving applications over opportunistic communication systems

“…This work is dedicated to genetic algo rithms. Therefore, we take the graphs of a composite service as genotype and modify them by the use of genetic operators [6]. For, instance me genetic operators modify output-input connection or change the execution order.…”

“…A service composition model consists of a workflow graph based on modified timed automata and a dataflow graph, which serves as the basis for evolutionary mechanisms [5]. By applying genetic operators to service compositions, we are able to evolve service compositions in order to substitute parts of the application that have dropped out and to provide an equivalent functionality with a different set of base services [6]. In this paper, we present the service model, and the corre sponding execution environment, for enabling dynamic mo bile service compositions.…”