Adaptation solutions for Service Compositions

Funded by:CNPq (Brazilian government agency)
Duration:March 2013 until February 2017
Contact:Dr. L. Ferreira Pires (Luís)


Service-Oriented Computing (SOC) has been used in business environments in order to integrate heterogeneous systems. The dynamic nature of these environments causes changes in the application requirements. As a result, service composition must be flexible, dynamic and adaptive. These features motivate the need to ensure the runtime system correct behaviour without stopping its operation. In order to achieve this goal, service composition must be monitored at runtime, and its execution trace must be analysed according to the functional requirements described at design-time. Once some undesirable behaviour is detected during the execution of a service composition, an adaptation strategy must be triggered. These strategies can range from simply trying to re-invoke the service to replacing the service by another equivalent service. The development of an adaptive system is still an opportunity due to the complexity of dealing with adaptation issues, for example, when, where, why and how to provide an adaptation. Existing solutions have been implemented focus on scenarios that adopt specific criteria for the design, monitoring and adaptation of the service composition. For example, a specific adaptation solution can be able to meet scenarios where the service composition has been developed in WS-BPEL and its functional requirements described in temporal logic. Considering this context, we are identifying in the literature which adaptation scenarios have not been implemented. Our contribution consists of providing a solution that meets different adaptation scenarios for service composition. 


  • Glaucia Melissa Medeiros Campos (PhD student)
  • Luís Ferreira Pires (supervisor, UT)
  • Nelson Rosa (supervisor, UFPE)