This PhD dissertation investigates how to overcome the negative effects of cross cutting concerns in the development of composite service applications. It proposes a combination of dynamic aspect oriented programming with a rules driven service composition mechanism. This combination allows very flexible utilization of aspects based on run-time data. The thesis contributes a join-point model and it integrates techniques for weaving and advice definition into the underlying composition language and execution engine. A particular focus of the thesis is telecommunication applications with their unique model for utilizing heterogeneous constituent services and their severe real-time requirements. Next to its primary use for modular implementation and flexible deployment of concerns in telecommunication applications, the dissertation discusses AOP as a feature for automated management and customization of service applications. The verification of the proposed solution contributes a detailed assessment of run-time performance, including a theoretical model of the AOP implementation. It allows predicting the performance of various alternative solutions. The proposed solution for combined AOP and service composition provides properties, which directly address challenges in pervasive computing and the Internet of things. Thus, this dissertation advances beyond the telecommunication domain with results applicable to various highly relevant technical developments.
|Qualification||Doctor of Philosophy|
|Award date||8 Apr 2016|
|Place of Publication||Tilburg|
|Print ISBNs||978 90 5668 468 6|
|Publication status||Published - 2016|