Increase Software Dependability via AOP-Based Runtime Adaptation

Abstract

Abstract-Increasinglycomplex application systems operating in dynamic environments are forcing researchers to find ways of developing and evolving dependable software systems dynamically. Runtime Adaptation is a critical feature for dependable application systems to cope with changes in the operating conditions and to meet the required dependability. AOP can improve the modularity of self-adaptive software by encapsulating the crosscutting concerns into aspects cleanly. Therefore, in this paper we take the advantages of AOP and self-adaptive technique, and propose an AOP-based dynamic adaptation approach. We show how to increase system dependability by runtime adaptation from the perspective of system internal state and external environment. Meanwhile, we illustrate our approach by applying it to a simplified on-line train ticket reservation system.