Time Delay Estimation via Correlation in the Non-Synchronous Sensor Network by the Internet of Things

This paper is concerned with the elimination of time delay among data from different channels sampled by the sensor network from the Internet of things (IoT), especially the wireless sensor network (WSN) in Structural Health Monitoring (SHM). In this paper, a technique using correlation has been proposed and moving window plays an important role in this algorithm. Obtaining the correlation coefficient at each point, use interpolation function to fit a curve so that the sub-sampling-interval part of time delay can be calculated, which is enlightened by sub-pixel deformation algorithm from Digital Image Correlation (DIC). The length of each segmentation and the choice of interpolation function will affect the computational efficiency and the quality of result, so they should be carefully confirmed. Finally, a numerical experiment has been conducted to verify this method and analysis the error of this method, and gives out a satisfactory consequence.