Improvement of the decentralized random decrement technique in wireless sensor networks

Random Decrement Technique (RDT) based decentralized computing approaches implemented in wireless sensor networks (WSNs) have shown advantages for modal parameters and data aggregation identification. However, previous studies of RDT-based approaches from ambient vibration data have usually assumed that the input excitation is a broad-band stochastic process modeled by stationary white or filtered white noise. In addition, the choice of the triggering condition in RDT is intimately related to data communication. In the present paper, a theoretical justification of the random decrement technique is presented for nonstationary white noise excitations. Local extremum triggering condition is chosen and implemented for the purpose of minimum data communication in RDT based distributed computing strategy. The performance of improved RDT was assessed in terms of (1) accuracy of the estimated modal properties and (2) efficiency in the wireless data communication. Numerical simulation confirms the validity of the proposed method for identification of modal parameters from nonstationary ambient response data and better efficiency in decentralized data aggregation.