Deep-Fading Hole Avoidance for Secure Region Detection using Channel State InformationDeep-Fading Hole Avoidance for Secure Region Detection using Channel State Information
- Authors
- Suh, Jihwan; Yoo, Yongjae; Paek, Jeongyeup; Bahk, Saewoong
- Issue Date
- Dec-2022
- Publisher
- 한국통신학회
- Keywords
- Channel state information; deep fading hole; finger-printing localization; Internet of things; secure region detection.
- Citation
- Journal of Communications and Networks, v.24, no.6, pp 645 - 654
- Pages
- 10
- Journal Title
- Journal of Communications and Networks
- Volume
- 24
- Number
- 6
- Start Page
- 645
- End Page
- 654
- URI
- https://scholarworks.bwise.kr/cau/handle/2019.sw.cau/66252
- DOI
- 10.23919/JCN.2022.000046
- ISSN
- 1229-2370
1976-5541
- Abstract
- One of the critical challenges in many wirelesssystems is the deep fading hole problem where signals interferedestructively to create an abrupt change in signal amplitude dueto multipath fading. In this work, we tackle this challenge in thecontext of secure region detection (SRD) problem. Specifically, wepropose SHARD, a novel hole avoidance technique that analyzeschannel state information (CSI) to significantly improve theaccuracy of CSI-based SRD. SHARD identifies potential fadingholes in CSI amplitude, and removes an unreliable portion of datawhile utilizing the remaining unaffected part. To compensate forthe loss of information, we define phase-distance for reliable useof time-varying CSI phase, and neighboring reference points areutilized for accurate matching. Our real-world experiments showthat SHARD can achieve a near-perfect 99.96% true-negativeratio (successfully rejecting devices not in the secure region)and an excellent true-positive of 98.01% for practical usage,significantly better than state-of-the-art prior work. We believeour ideas can be generalized to many RF-based localizationsystems to mitigate the deep fading hole problem and improvetheir accuracy.
- Files in This Item
- There are no files associated with this item.
- Appears in
Collections - College of Software > School of Computer Science and Engineering > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.