Chan P.YLai A.I.-CPEI-YUAN WURUEY-BEEI WU2022-04-252022-04-25202114248220https://www.scopus.com/inward/record.uri?eid=2-s2.0-85113716816&doi=10.3390%2fs21165665&partnerID=40&md5=f9cd60034abaf85c0960394a9aaa032fhttps://scholars.lib.ntu.edu.tw/handle/123456789/607270This paper proposes a practical physical tampering detection mechanism using inexpen-sive commercial off?the?shelf (COTS) Wi?Fi endpoint devices with a deep neural network (DNN) on channel state information (CSI) in the Wi?Fi signals. Attributed to the DNN that identifies physical tampering events due to the multi?subcarrier characteristics in CSI, our methodology takes effect using only one COTS Wi?Fi endpoint with a single embedded antenna to detect changes in the rel-ative orientation between the Wi?Fi infrastructure and the endpoint, in contrast to previous sophis-ticated, proprietary approaches. Preliminary results show that our detectors manage to achieve a 95.89% true positive rate (TPR) with no worse than a 4.12% false positive rate (FPR) in detecting physical tampering events. ? 2021 by the authors. Licensee MDPI, Basel, Switzerland.Channel state information (CSI)COTS Wi?Fi mobile deviceDeep neural network (DNN)Physical tampering detectionSingle embedded antennaAntennasCommercial off-the-shelfDeep neural networksEmbedded antennaFalse positive ratesSubcarrierTampering detectionTrue positive ratesChannel state informationarticledeep neural networkpositivity ratepreliminary dataNeural Networks, Computer[SDGs]SDG9[SDGs]SDG11journal article10.3390/s21165665344511072-s2.0-85113716816