Lightweight and secure encryption using channel measurements

Wireless sensor networks are the building blocks of the Internet-of-Things architectures, conveying critical and often sensitive and private information. As these networks often consist of severe resource constrained devices, lightweight encryption mechanisms are of paramount importance for achieving energy efficiency. However, the majority of the proposed algorithms do not fully fulfill the requirements for energy efficiency. Furthermore, key distribution schemes are necessary for their proper operation, making the network vulnerable to adversaries that manage to capture the keys during key exchange. In this work, we propose a scheme where key extraction is performed using channel measurements, thus there is no need for any key distribution mechanism. The derived keys are used for encryption/decryption using the primitives of the compressed sensing theory, which allows encryption and compression simultaneously. The evaluation results show that legitimate nodes experience a very low reconstruction (decryption) error. At the same time, adversaries located at a distance greater than half of the carrier frequency's wavelength, experience a higher error, thus being unable to capture sensitive information.