A novel framework of self-adaptive fault-tolerant for pervasive computing

The increasing complexity of software and hardware resources and frequentative interaction among function components make fault-tolerant very challenging in pervasive computing system. In our paper, we propose a novel framework of self-adaptive fault-tolerant mechanism for pervasive computing environments. In our approach, the self-adaptive fault-tolerant mechanism is dynamically built according to various types of detected fault based on continuous monitoring, analysis of the component state. We put forward the architecture of fault-tolerant system and the policy-based fault-tolerant scheme, which adopt three-dimensional array of core features to capture spatial and temporal variability and the Event-Condition-Action rules. The mentioned mechanism has been designed and implemented on a prototype of office pervasive computing application systems, called POPCAS System. We have performed the experiments to evaluate the efficiency of the fault-tolerant mechanism. The results of the experiments show that the proposed mechanism can obviously improve reliability of the POPCAS System.