面向情境感知的不确定性数据融合策略

情境感知计算是对传感器感知的数据进行融合从而得到用户情境信息的过程。文中针对传感器数据存在的不确定性问题,在Context Toolkit模型的基础上提出了一种面向情境感知的数据融合策略,把数据融合分为底层融合和高层融合,在这两层上分别采用模糊逻辑和动态贝叶斯网络的数据融合方法。仿真实验结果表明,这种将模糊逻辑和动态贝叶斯网络相结合的策略能有效地处理情境感知过程中的不确定数据和信息,从而提高情境感知计算的容错性和准确度。

Strategy of Uncertainty Data Fusion for Context-Awareness

Context-aware computing is a process of fusing sensor data to perceive the user＇s context information.Considering the uncertainty property of sensor data,it proposes a context-aware data fusion strategy based on the Context Toolkit model,classifying the data fusion into two layers：the lower layer fusion and the high layer fusion.In these two layers,the fuzzy logic theory and DBN data fusion algorithm are used respectively.Simulation results show that this combined approach can not only effectively tackle with the uncertain data and information during the context awareness process,but also improve the fault-tolerance and accuracy of context awareness.