BROWSE
Detail
Conference Paper
Dynamic Collection based Smoothed Radiomap Generation System
Cited - time in 
Share 
Share
- Authors
- Jooyoung Kim, Myungin Ji, Youngsu Cho, Yangkoo Lee, Sangjoon Park
- Issue Date
- 2013-10
- Citation
- International Conference on Indoor Positioning and Indoor Navigation (IPIN) 2013, pp.1-2
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- The fingerprinting method has been considered as a promising technology of indoor localization. However, it has a serious problem to employ to applications targeting broad service areas due to the high cost for constructing DB. In the conventional DB generation system, it utilizes a statistic collection process. In the process, every collector gathers signal characteristics on every point which divides a service area into a grid. To achieve sufficient data set for calculating the characteristics, a collector usually stands on the points for a while, up to a couple of minutes. Therefore, it takes too much time and cost to construct a fingerprint DB for whole serving areas. To deal with this problem, a dynamic collection based fingerprint DB generation system is proposed. In the proposed system, collectors walk along predesigned routes and gather signal measurements in motion using smartphone, not staying on a point. Therefore, the proposed system remarkably reduces the cost for constructing DB by improving the time consuming collection process. However, the signal collecting in motion causes to decrease the reliability of the DB because of the signal variation which is significant in indoor areas. To mitigate the variation, we exploit a moving average filter to the smooth noisy measurements. As a result, the proposed system improves the efficiency of the fingerprint DB generation system with reasonable positioning performance. Experimental result proves the validity of the proposed dynamic collection based smoothed radiomap. The average positioning error using the proposed smoothed radiomap is about 7.41m, and the standard deviation of the error is 6.17m.
- KSP Keywords
- Data sets, Experimental Result, Fingerprinting method, Generation system, Moving average filter(MAF), Noisy measurements, Positioning Error, Signal characteristics, Signal collecting, Standard deviation(STD), indoor localization