Enhancing compressive sensing encryption in constrained devices using chaotic sequences
Abstract
The rapid proliferation of IoT (Internet-of-Things) applications has raised major concerns regarding privacy and security issues. Typical IoT applications, like ehealth, smart transportation, etc., are often used to convey highly sensitive and personal information. The vast majority of the IoT applications heavily depend on miniature sensors that are usually battery operated; hence, energy efficiency is also a crucial aspect. In this respect, we use the Compressive Sensing (CS) principles to enable simultaneous data encryption and compression. We study CS from the security point of view, showing that it is highly vulnerable to Chosen Plaintext Attacks. We propose a scheme based on chaotic sequences that makes CS immune to these attacks. We also show that attackers are not able to correctly decrypt captured ciphertext. Moreover, we describe the implementation of this crypto-scheme in a real mote, presenting evaluation results, in terms of energy consumption and execution time.