Performance enhanced wireless device authentication using multiple weighted device-specific characteristics

Although device-specific characteristics are often used to verify wireless devices, the performance of such authentication techniques are still not satisfactory in terms of accuracy and reliability. This paper develops a new approach to enhance the wireless device authentication through using more than one unique characteristics. To be specific, a weighted combination of multiple transmitter-specific characteristics is utilized as a new device fingerprint at the receiver side. This fingerprint is then applied in a hypothesis testing to authenticate wireless transmitters. Moreover, the optimal weights for the characteristics combination are derived to maximize the detection probability. Our simulation results validate the proposed authentication technique and show significantly enhanced authentication accuracy.