Distributed Evolutionary Computation Using SOAP and REST Web Services

Abstract

In this chapter, a high-level comparison of both SOAP (Simple Object Access Protocol) and REST (Representational State Transfer) is made. These are the two main approaches for interfacing to the web with web services. Both approaches are different and present some advantages and disadvantages for interfacing to web services: SOAP is conceptually more difficult (has a steeper learning curve) and more “heavy-weight” than REST, although it lacks of standards support for security. In order to test their efficiency (in time), three experiments have been performed using both technologies: first a basic client-server model implementation to test communications has been implemented; then, a master-slave based genetic algorithm (GA) to solve an optimization problem has been used; and finally, as third experiment, an approach to evolutionary distributed optimization of multilayer perceptrons (MLP) using REST and language Perl has been done. In these experiments, a master-slave based evolutionary algorithm (EA) has been implemented, where slave processes evaluate the costly fitness function (training a MLP to solve a classification problem). As expected, the parallel version of the developed programs obtains similar or better results using much less time than the sequential version, obtaining a good speedup. The results obtained have shown that both SOAP and REST can be used as communication protocol for distributed evolutionary computation, obtaining a good speedup. Results obtained are comparable, and only for large amounts of data (big messages), REST communications take longer than SOAP communications.KeywordsEvolutionary ComputationSimple Object Access ProtocolMaster ProcessOperator PriorityGood SpeedupThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.