Analysis of Lightweight Cryptographic Primitives for the IoT Devices
Shibam Ghosh# and Orr Dunkelman - Department of Computer Science
PhD Grant 2021
The Internet of Things (IoT) has immense prospects to change the world. We witness IoT devices on numerous everyday objects that are connected to the internet to exchange real-world data. However, IoT implementations present several security challenges and can become a target of cyberattacks (https://www.gartner.com/en/doc/iot-security-primer-challenges-and-emerging-practices). Therefore, we need to deploy cryptographic algorithms on such IoT devices to provide integrity and confidentiality of data. However, IoT devices work in highly constrained environments. Among various other constraints, these devices usually contain small micro-controllers with a small number of simple instructions and run in a short time with low energy. For these reasons, conventional cryptographic schemes are not well-suited for such devices.
On the other hand, lightweight cryptography enables secure encryption, even for devices with limited resources. Thus recent years have seen significant progress in developing lightweight symmetric crypto primitives. In addition, the ongoing Lightweight Cryptography (LWC) project by the US National Institute of Standards and Technology (NIST) (https://csrc.nist.gov/Projects/lightweight-cryptography) gave a new impetus to designing and analyzing lightweight schemes.
As the lightweight cryptographic primitives work in resource-constrained devices, these work with simpler components and as a result, lightweight schemes are vulnerable due to lower security margins. Thus ensuring enough security is essential. However, the security of the cryptographic scheme is not a binary statement. The only way to guarantee sufficient security is cryptanalysis. Therefore, our research focuses on the cryptanalysis of lightweight cryptographic primitives that are essential for the safety of IoT devices. We have contributed to the worldwide effort in the cryptanalysis of lightweight cryptographic schemes that are submitted to the NIST LwC project. We used a Constraint Satisfaction Problem (CSP) based approach to analyze the security of the KNOT permutation (https://www.springerprofessional.de/en/automatic-search-for-bit-based-division-property/19711962). The KNOT permutation is the main primitive used in the KNOT family of lightweight authenticated encryption algorithms and hash functions, submitted to the NIST lightweight crypto competition. We also implemented our model to NIST lightweight candidate Ascon (https://ascon.iaik.tugraz.at/), and obtained improved results.