•  
  •  
 

Abstract

Bluetooth devices actively broadcast software when pairing to connect. Even during the connection process, the connection can be monitored to view information about the transmission. Using this information, anyone can hijack your existing connection and steal data. Bluetooth connections can be single or multiple. Thus, while connecting to a device, the same device could be simultaneously connected to another device. To avoid this problem, a new software is proposed to support the ID-based authentication process for paired devices by integrating an authentication method based on the biometric features of the device owner. The proposed two-factor security authentication system for pairing devices via Bluetooth is based on the principle of authentication for the possibility of pairing devices with different operating systems. The two-factor authentication is designed based on the device’s ID and the biometric features of the palm. Biometric features of the hand images used in this work are then extracted according to the LDA method for use in recognition steps by DT and bagging algorithms. The machine learning algorithms were validated in the second phase of the authentication, where the bagging accuracy using LDA reached 84.335%, while the accuracy of DT with LDA reached 81.601%.

Download

Share

COinS