Temporal Data-Driven Sleep Scheduling and Spatial Data-Driven Anomaly Detection for Clustered Wireless Sensor Networks

Author： Li Gang He Bin Huang Hongwei Tang Limin

Publisher： MDPI

E-ISSN： 1424-8220|16|10|1601-1601

ISSN： 1424-8220

Source： Sensors, Vol.16, Iss.10, 2016-09, pp. : 1601-1601

Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.

Previous Menu Next

Abstract

The spatial–temporal correlation is an important feature of sensor data in wireless sensor networks (WSNs). Most of the existing works based on the spatial–temporal correlation can be divided into two parts: redundancy reduction and anomaly detection. These two parts are pursued separately in existing works. In this work, the combination of temporal data-driven sleep scheduling (TDSS) and spatial data-driven anomaly detection is proposed, where TDSS can reduce data redundancy. The TDSS model is inspired by transmission control protocol (TCP) congestion control. Based on long and linear cluster structure in the tunnel monitoring system, cooperative TDSS and spatial data-driven anomaly detection are then proposed. To realize synchronous acquisition in the same ring for analyzing the situation of every ring, TDSS is implemented in a cooperative way in the cluster. To keep the precision of sensor data, spatial data-driven anomaly detection based on the spatial correlation and Kriging method is realized to generate an anomaly indicator. The experiment results show that cooperative TDSS can realize non-uniform sensing effectively to reduce the energy consumption. In addition, spatial data-driven anomaly detection is quite significant for maintaining and improving the precision of sensor data.