Multisensor fusion based localization method for mobile sensor network

Gas detection is the main concern in factory or home safety. Wireless sensor network, sensors combined with a microprocessor and wireless communication, can be applied in gas detection. There are a lot of researches discuss the fixed gas sensors in certain location, however, this system still requires that people carry gas detectors to patrol the entire environment due to sensors failure or sensing coverage issues. Disaster is still possible if a gas leak should occur, and mobile sensor nodes are useful to avoid this situation. Mobile sensor network system requires localization system to know its own position. In this paper, we use Kalman filter to process the radio frequency signal, and then use triangulation localization method to find its own coordinate through radio frequency (RF) signal and dead reckoning system. However, the data obtained from radio or dead reckoning is easily influenced from noise. The accuracy of localization system can be increased by using covariance intersection (CI) data fusion technique. In this paper, we propose and demonstrate a prototype for gas detection by mobile sensor network technology.