| Summary: | 碩士 === 國立中正大學 === 電機工程所 === 95 === Target tracking sensor network is considered as one of the killer application for sensor network system. There are a lot of research issues in design and implement of the target tracking sensor network, such as data fusion, routing, and energy conservation, and etc. Energy conservation is one of the most important issues.
Unfortunately, the physical limitation on high inaccuracy of radio signal is a challenge for target tracking sensor network. Many researches neglect this critical factor and may underestimate the total energy consumption of the sensor network system. The possible improvements for the causes from radio signals quality are either triggering the sensor more times or activate multiple sensors.
However, the object will change its position and state anytime. It might not have more opportunity for the sensor nodes to detect twice at that time and position. For the consideration of energy conservation, the less active nodes, the more energy saved, it is a tradeoff between accuracy and energy.
Based on the physical limitation of radio signals and for real time monitoring the moving object, we develop an adaptive grey-fuzzy with weighted least square predictor to improve the accuracy of moving object trace.
|
|---|
