Date Added: Nov 2012
High quality sensor data stream is crucial to wireless sensor networks applications. However raw data streams in wireless sensor networks tend to be not reliable. Therefore, improving sensor data quality is an important issue for all kinds of wireless sensor networks applications. In this paper, the authors proposed an integrated framework, MuMoCo, which is based on such a fact: the factors leading to outlier or data missing such as events, insufficient power, or malicious nodes have similar influence on each modality of data created by a node at the same time. Not considering the correlation among different modalities of data may probably lead to a contradictory: different conclusions of data verification according to different modalities of data.