| Summary: | 碩士 === 元智大學 === 電機工程學系 === 102 === Many indoor positioning systems are using database comparison method. This method require a previous site survey by collecting a set of received signal strength from access points (APs) in the vicinity, then build a database called radio map. When a client at the time of positioning, the set of signal strengths are compared with the database components associated with all reference points, and the result specifying the reference point or points of best match is used to indicate the estimate of target location. This method provides high accuracy performance and does not need to know the location of WiFi APs. However, the transmitted power of these APs may dynamically change due to the communication reasons. When the power adjusted, the previously created database will not reflect the current state. The mismatch phenomenon will cause large errors. An intuitive idea is to collect all of received signal strength from APs in the vicinity at each transmitted power, build a huge database for comparison. But it requires a time-consuming site survey and a great amount of manual efforts.
This approach focus on database comparison method of indoor positioning system. Investigates how a figerprint database should be dynamically updated when the transmitted power of the APs change. We also devise a new method for adapt radio map without to spend a lot of manual efforts. The adaptive radio map can accurately reflect the current state of APs, maintain high accuracy.
|
|---|
