The Research On Compression-fusion Based Event Detection Scheme Of Wireless Sensor Networks

Event detection is one of the most important applications of wireless sensor networks (WSN). In the real-world applications, sensor nodes’ monitoring data may be incorrect because of the impact of various environmental factors, resulting in misjudgments on events in monitoring region when event detection is performed. Therefore, it is important to research the event detection algorithm with fault-tolerant mechanism of WSN to improve the reliability of event detection. At the same time, sensor nodes will run out of its limited energy and be dead quickly if event detection is performed on the vast amount of monitoring data directly, and as a result, the lifetime of WSN will be shorten. So it is absolutely significant to research the event detection algorithm with data compression mechanism while ensuring the event detection reliability to reduce the energy consumption and improve the energy efficiency of event detection as well as prolong the lifetime of WSN.The research of this paper focus on the event detection scheme of WSN. Two event detection schemes are proposed in this paper for the shortcomings of the current event detection schemes while combining the characteristics of WSN. Its main work includes:1. The brief analysis of the limitations for event detection applications and requirements of event detection schemes caused by the characteristics of WSN is introduced.2. The related work on event detection schemes in WSN is summarized and then, the shortcomings of the current event detection schemes are analyzed.3. A fault-tolerant and fast event detection scheme for WSN is proposed for event misjudgments caused by erroneous sensor data and the event detection delay. The scheme utilizes fault-tolerant mechanism to eliminate the interference of erroneous sensor data and fault nodes, and employs event prediction mechanism to predict whether an event occurred to improve the reliability and real-time of event detection. At last, some experiments are conducted to compare the proposed scheme with another two event detection methods for wireless sensor networks.4. A SAX compression based event detection scheme for WSN is proposed for the limited energy and reliability of event detection when event detection is conducted in WSN. The scheme uses SAX to compress and replace the monitoring data with symbols, and then utilizes Markov model to establish the abnormal transition probability model of SAX symbol to measure the degree of abnormality of SAX symbol sequence to improve the energy efficiency and reliability of event detection. At last, some experiments are conducted to test the proposed scheme and prove its effectiveness.The proposed two event detection schemes for WSN solve the compatible of the reliability, real-time and energy efficiency of event detection for WSN and have a good event detection performance.