An accurate fault location method for wireless sensor network based on random matrix theory

Article type: Research Article

Authors: Wang, Qi^{; *}

Affiliations: Yuncheng University, Yuncheng 044000, Shanxi, P.R. China

Correspondence: [*] Corresponding author. E-mail: yiwang@36haojie.com.

Abstract: In order to improve the effect of fault location, this paper proposes an accurate fault location method for wireless sensor networks based on random matrix theory. The standard non Hermite matrix is used to extract accurate fault location data. Considering the volatility of the original data, the original random matrix is preprocessed. Based on the real-time sliding time window method, the space-time characteristic data of network faults are determined, and the precise fault location of wireless sensor networks based on random matrix theory is realized.Experimental results show that the false positive rate of the proposed method is only 2%. The average fault location accuracy is as high as 96.4% and the fault location time is only 15.1 s, which shows that the proposed method has a good location effect.

Keywords: Random matrix theory, covariance matrix, Hermite matrix, wireless sensor network, network fault location

DOI: 10.3233/WEB-220026

Journal: Web Intelligence, vol. 21, no. 1, pp. 93-102, 2023