The imminent arrival of quantum computing has accelerated the need for cryptographic systems resistant to quantum attacks. Such attacks exploit the vulnerability in private and public key encryption systems, where the public key is derived from the private key, which could be refactored from the public key. To address this issue, the National Institute of Standards and Technology (NIST) launched a global competition in 2016 to create quantum-resistant algorithms. CRYSTALS-Kyber, a lattice-based algorithm focused on the learning with errors (LWE) problem, was selected for standardization. This work introduces RKyber, a variant that instead targets the learning with rounding (LWR) problem, simplifying computations by using deterministic errors rather than random noise. Both algorithms were executed 1000 times, showing that RKyber offers improved speed at the cost of some security.

The imminent arrival of quantum computing has accelerated the need for cryptographic systems resistant to quantum attacks. Such attacks exploit the vulnerability in private and public key encryption systems, where the public key is derived from the private key, which could be refactored from the public key. To address this issue, the National Institute of Standards and Technology (NIST) launched a global competition in 2016 to create quantum-resistant algorithms. CRYSTALS-Kyber, a lattice-based algorithm focused on the learning with errors (LWE) problem, was selected for standardization. This work introduces RKyber, a variant that instead targets the learning with rounding (LWR) problem, simplifying computations by using deterministic errors rather than random noise. Both algorithms were executed 1000 times, showing that RKyber offers improved speed at the cost of some security.