科研之友
微信
新浪微博
Facebook
分享链接
摘要
Consisting of various sensing and computing devices deployed in a changing environment, a sensor network's raw sensed data have many uncertainties. A natural way to deal with them is generating belief messages. Sensing objects continuously change with time, so are their beliefs. Therefore, dynamic models are required to monitor distributed states in the system. This paper presents a CTBN based intelligent system for modeling dynamics and processing uncertainties in sensor networks. Algorithms for message passing and parameter updating for adapting the model to the changing environment are provided. The effectiveness of the system is shown in experiments.
