MASc Seminar: Multi-Purpose Designs in Lightweight Cryptography

Monday, November 26, 2018 11:00 am - 11:00 am EST (GMT -05:00)

Candidate: Morgan Yue He

Title: Multi-Purpose Designs in Lightweight Cryptography

Date: November 26, 2018

Time: 11:00 AM

Place: EIT 3142

Supervisor(s): Gong, Guang

Abstract:

The purpose of this thesis is to explore a number of techniques used in lightweight cryptography design and their applications in the hardware designs of two lightweight permutations called sLiSCP and sLiSCP-light. Most of current methods in lightweight cryptography are optimized around one functionality and is very useful for applications that require their speci c design. Therefore, cryptographic circuits pose signicant challenges in the area of formal veri cation. Formal veri cation use mathematics to formulate correctness criteria of designs, to develop mathematical models of designs, and to verify designs against their correctness criteria.

In this thesis, we implemented two lightweight permutations designs of sLiSCP and sLiSCP-light. We veri ed the implementations of sLiSCP with completion functions and equivalence checking. During the veri cation of sLiSCP cipher, we discovered additional area that could be saved if we tweaked the design slightly. This would lead us to consider the design of sLiSCP-light which helps dramatically reduce area.

In this design, we apply the optimization techniques of pipelining and hardware re-use to create an optimized implementation of sLiSCP-light. We veri ed the optimized implementation of sLiSCP-light with completion functions and equivalence checking. During the veri cation of sLiSCP-light, we developed the methodology of parallel implementation that can decrease the number of veri cation obligations required to verify the correctness of a circuit.Our results were very positive in that our designs of sLiSCP and sLiSCP-light satis- ed the lightweight requirements, including hardware area, power, and throughput, for applications such as passive RFID tags.

Lastly, we did tests on the randomness of Simeck-like Fiestel structures. We wanted to observe the pseudorandom nature of structures similar to Simeck and Simon so we performed exhuastive tests on small instances of these structures to trace any treands in their behaviour. We con rmed that Simon and Simeck were very consistent and provided acceptable pseudorandom results; results which should carry over to their large case counterparts.