Analysis of the data streams trend in sensor network based on sliding window

This paper presents an analysis of the data streams trend in sensor network based on sliding window. The method combine the sliding window technique with the nonlinear polynomial fitting of least square, which is used to analysis the trend of data flow that obtained from the sensors which is applied in the object being monitored. Effective trend analysis can provide the basis for early warning, maintenance and repair, management decisions while the object is being monitored in the special climate, traffic conditions and severe abnormal operating conditions.