Using Binding Bigraphs to Model Context-Aware Adaptive Applications

Abstract

One of the key characteristics of context-aware adaptive applications (CAAAs) is to change their behaviors according to the user-configured adaptation rules at runtime. Some of these rules, however, often result in inconsistence with the specifications of the CAAAs because of the complexity of the adaptive applications. A formal modeling method based on binding bigraphs is proposed to solve this inconsistence in this paper. The main idea is to use fully the binding characteristics in bigraphs to constrain both the context variables of the CAAAs and the conditions that the bigraphical reactions occur. The adaptation rules, hence, can be optimized to reduce some conflicts in the design phase of CAAAs and in consequence some faults in CAAAs can be detected earlier. Taking the PhoneAdapter as an example, we compare it with the traditional formal model called A-FSM to show the effectiveness of the approach presented in this paper.