The Application Based on WiFi Signal Strength Using SVM for QR code Access Control System

• Main Authors: Chun-Ming Liao, 廖俊銘
• Other Authors: When-En Chen
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/49puqf

碩士 === 國立宜蘭大學 === 多媒體網路通訊數位學習碩士在職專班 === 107 === In this study, the WiFi module of the mobile device is used to detect the Received Signal Strength (RSS) of the Wireless access point (WAP) in the surrounding environment, and obtain the Media Access Control Address (MAC Address) of the WAP, using RSS and MAC as feature values, uses the Support Vector Machine (SVM) to classify the area where the QR code (Quick Response Code) is located. In the WiFi wireless environment formed by many AP signals, the mobile device receives the difference of the WAP signal strength in different areas, and constructs the fingerprint of different regions, thereby surmises whether the area of the QR code is correct. The QR code anti-counterfeiting verification of the access control system is the feature of this paper. The data of this study experiment were collected at the front of the E306 computer classroom on the third floor of the Ge-Zhi Building in the National Ilan University at 14 Oct 2018 and 28 Oct 2018 respectively. The test was divided into 26 test points within 2.5 meters. 2600 data were collected in single day, total 5200 data. Each data is divided into a training set and a test set by a ratio of 60:40, and the model is generated through LibSVM training and then predicted. This study proposes an improvement method for LibSVM normalization to improve the prediction accuracy.