| Summary: | 碩士 === 國立臺灣科技大學 === 資訊工程系 === 95 === Sensor Network for the past few years has become a very popular research topic due to vigorous development of communication system. Estimation about position of sensors is a fundamental and important link in related researches. Owing to this fact, many different position algorithms were created one after another. However, most algorithm methods need nodes of known coordinates, and most of the information of coordinates needs to be provided by GPS. For this reason, the indoor target environment, interference of environmental materials, and the climate, would all cause possibilities of errors in position. Therefore, the position algorithm without sensors of known nodal coordinates gradually gets researchers’ attention.
In most cases, Sensor Network only needs to know the distance between each other to operate relative applications in network. As a result of this, if we have an algorithm to calculate the relative distance of sensors, we can apply it to Sensor Network in order to avoid interference of uncertainties and achieve practical value. This is also the reason why the position algorithm without sensors of known nodal coordinates gradually wins its own spotlight.
Different from ordinary position algorithm with nodes of known coordinates, the position method without sensors of known nodal coordinates is just in its beginning status. So far as we know, the existed related algorithms all focus on measuring distance to estimate relative distance of sensors. In this thesis, we propose a technique that combines angle and distance measuring. The integration of these two methods is to estimate all coordinates and directions of sensors in local coordinate system. We try to get a more ideal position result. All details are explained on following pages.
|
|---|
