| Summary: | 碩士 === 國立陽明大學 === 醫學工程研究所 === 97 === In this paper we present COMPASS (Community-Optimized Measuring of
Positions Associated with Sensing Signals), a method to improve indoor location
sensing accuracy using Radio Frequency Identification (RFID), based on the design
using reference tags. The major advantage of COMPASS is that it provides more
accuracy location, compared with LANDMARC, and keeps the beauty of true
simplicity, giving it the ability to track down moving object in real-time environment.
Yet, based on our result, we do see that while improving the average accuracy,
some measurements turn out having a larger error, making a bigger maximum error of
the algorithm. We find out ways to improve and predict large error observations,
forming two types of hybrid systems, one with a dynamic community size and the
other with LANDMARC. Both systems succeed a better average positioning accuracy
and less maximum error compared with the original COMPASS system.
Finally, we point out suggestions for further research and future work.
|
|---|
