(A) two-step approach to improve floor-level accuracy of large scale indoor localization based on WLAN fingerprints무선랜 핑거프린트 방식 대규모 실내 측위의 층간 정확도를 향상시키는 2단계 접근법

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 732
  • Download : 0
DC FieldValueLanguage
dc.contributor.advisorHan, Dong-Soo-
dc.contributor.advisor한동수-
dc.contributor.authorYang, Hyun-Il-
dc.contributor.author양현일-
dc.date.accessioned2011-12-13T06:09:54Z-
dc.date.available2011-12-13T06:09:54Z-
dc.date.issued2011-
dc.identifier.urihttp://library.kaist.ac.kr/search/detail/view.do?bibCtrlNo=467938&flag=dissertation-
dc.identifier.urihttp://hdl.handle.net/10203/34979-
dc.description학위논문(석사) - 한국과학기술원 : 전산학과, 2011.2, [ v, 36 p. ]-
dc.description.abstractWith proliferation of smart phones in recent years, location based services become main stream of IT service market. To support location-based services, many localization techniques have been proposed, such as GPS, RFID, GSM, and WLAN-based. Because of cost advantages, WLAN based localization is most popular in indoor situation, where GPS, the dominant localization technique, is hardly working inside the building. There are many applications using WLAN localization such as indoor navigation. To determine user’s current location using WLAN signals, most of researches use Received Signal Strength Indication (RSSI) from wireless access points. Since WLAN fingerprint is easily gathered by smart phone APIs, and can be used in localization material without information about the position of access points, many localization systems adopt fingerprinting method. In WLAN localization system, floor-level errors are occurred relatively less often than distance-level so that they occupy small portion of indoor localization errors usually. Though, however, the effect of floor-level errors cannot be ignorable. Especially, floor-level error incurs much serious problem in 2D indoor navigation as the application shows only single floor map of current location. If the system estimates wrong floor location, users will view screen of different floor information that are entirely unrelated. In this study, we introduce our floor-level error correcting method working on WFCA. Simply, the correction method is based on access point set similarity of fingerprints using Jaccard’s coefficient. WFCA is also the new concept to predict floor-level errors in off-line phase using collected learning data. Using the combination of off- and on-line phase methodologies, we devise a new localization method with less floor-level errors than conventional system. We applied our proposed method to COEX, World Trade Center in Samseong-dong, Gangnam-gu, Seoul, Korea. As the result, with reasonable time cos...eng
dc.languageeng-
dc.publisher한국과학기술원-
dc.subjectWLAN fingerprints based positioning-
dc.subjectindoor localization-
dc.subjectlocalization error correction-
dc.subject실내 측위 오차 보정-
dc.subject무선랜 핑거프린트 기반 측위-
dc.subject실내 측위-
dc.title(A) two-step approach to improve floor-level accuracy of large scale indoor localization based on WLAN fingerprints-
dc.title.alternative무선랜 핑거프린트 방식 대규모 실내 측위의 층간 정확도를 향상시키는 2단계 접근법-
dc.typeThesis(Master)-
dc.identifier.CNRN467938/325007 -
dc.description.department한국과학기술원 : 전산학과, -
dc.identifier.uid020094345-
dc.contributor.localauthorHan, Dong-Soo-
dc.contributor.localauthor한동수-
Appears in Collection
CS-Theses_Master(석사논문)
Files in This Item
There are no files associated with this item.

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0