International Association
for Cryptologic Research

Wireless sensors attracts increasingly attention from both academia and industry owing to emerging applications built upon them such as Internet of Things, smart home, E-Health, and etc. It becomes a concern that the sensed value should be trusted in such applications scenarios. The security of the sensor used to resort to traditional cryptographic techniques, which, however, only provides limited protection by virtue of its vulnerability to physical attacks and may not economically viable. In this paper, we propose a new sensing methodology making remote sensing highly secure. In particular, we facilitate the susceptibility of physical unclonable function (PUF) to ambient environment variations, acting as a PUF sensor, to grantee the veracity of the sensing value even the PUF sensor is located in an untrusted remote location and the communication channel is also insecure without implementing relatively expensive crypto module. The PUF sensor is cost-efficient and most importantly anti-counterfeiting, while offers high level security. We demonstrate the practicability of the PUF sensor based on experimental implementations. In addition, we show that the extended sensing functionality of a PUF is actually improving PUF's resistance against modeling attacks.