Exploring Parallelism to Improve the Performance of FrodoKEM in Hardware

Abstract FrodoKEM is a lattice-based key encapsulation mechanism, currently semi-finalist in NIST’s post-quantum standardisation effort. A condition for these candidates to use NIST standards sources of randomness (i.e. seed-expanding), and as such most utilise SHAKE, an XOF defined the SHA-3 standard. However, many candidates, this module significant implementation bottleneck. Trivium lightweight, ISO standard stream cipher which performs well hardware has been used previous designs cryptography. This research proposes optimised FrodoKEM, concentrating on high throughput by parallelising matrix multiplication operations within cryptographic scheme. process eased due its higher lower area consumption. The parallelisations proposed also complement addition first-order masking decapsulation module. Overall, we significantly increase FrodoKEM; see $$16\times $$ 16 × speed-up, achieving 825 per second, $$14\times 14 763 compared state art, whilst maintaining similar FPGA footprint less than 2000 slices.