Abstract:In order to establish an integrated estimation system for quality of information in WSN (wireless sensor network), a flexible tailorable framework was proposed to dynamically estimate information accuracy and information timeliness in different WSN applications. The framework aggregates all information of the sensor nodes at the sink node on the basis of information fusion theory and timescale marking technique. Through approximating the fusion result to ground truth, the accuracy of measurement data and decision information was quantified by utilizing the numerical/nominal difference between the observed information and the fusion result, respectively. According to application deadline and delay sensitivity of information acquisition, timeliness was classified as three categories for characterization and representation, and quantified through the lightweight measurement of the information acquisition time. The estimation results along the time are dynamically updated by adopting a sliding window mechanism and an incremental calculation algorithm. The effectiveness of the framework was validated by simulations, including timeliness estimation in three target surveillance application scenarios and accuracy estimation in two environmental monitoring application scenarios. Simulation results match with the reference criteria, which demonstrates that the framework can be flexibly tailored to estimate information accuracy and information timeliness in different WSN applications.