Mobile sensor node localization and path planning for ubiquitous service

This paper is based on sensor network theory aim to localize the mobile sensor node. Our mobile sensor node is used to apply ubiquitous service in indoor environment. Therefore the localization and path planning for mobile sensor node is an essential factor. Path planning technology plays an important role in mobile sensor node. The mobile sensor node must successfully find a collision free path to reach from start point to goal point. And this path is the minimum distance. We propose a method to solve path planning problem which the model of non-vertexes (circle) obstacles is constructed to two-dimension map. Indoor localization technology which we developed is RSS (Received Signal Strength) in this paper. We use wireless LAN medium (Zigbee) to implement. RSS data can be transformed to distance by path lost function. Due to the radio frequency is significantly affected by diffraction, reflections, scattering and multi-path. In order to reduce the measurement error of RSS value we utilize Kalman filter theory. Besides, scale invariant feature transform (SIFT) has been implement on the robot to identify object with camera.