非一致上下文的自动修复技术

在普适计算中,上下文持续快速变化,上下文感知的应用根据上下文变化自动调整自身的行为以适应这种变化。然而,由于不可预测的环境噪声和动态变化等诸多因素的影响,很容易造成上下文不一致。非一致上下文会导致应用异常,甚至失败。为了解决这个问题,提出了一种自动修复上下文的新技术。它分为两个阶段:生成抽象修复用例;执行用例,以验证上下文不一致性是否已经修复。从用户的实际需要出发,生成的修复用例可以分别满足完整性或正确性。实验结果表明,这种技术在修复成功率上远远超过其他技术。

Automated Fixing of Inconsistent Contexts

In pervasive computing, environmental contexts are subject to frequent changes. Context-aware applications need to adapt their behavior accordingly. However, contexts can be easily inconsistent due to many reasons including unpredicted environmental noises and dynamics. Context inconsistency can lead to anomaly or even failure to applications. To address such problems, this paper proposes a novel technique for automatically fixing inconsistent contexts. It consists of two stages: generating abstract repair cases; executing them concretely to validate whether context inconsistency has been fixed. The generated repair cases can be complete or sound, depending on actual requirements from users. The experimental results show that the technique achieves a higher repairing rate than existing techniques.