Performance and Signature-Size Trade-Offs in Lattice-Based Post-Quantum Threshold Ring Signatures
DOI:
https://doi.org/10.1922/0hb9px27Keywords:
post-quantum cryptography, lattice-based cryptography, threshold ring signatures, signature size, computational performanceAbstract
Using the LoTRS benchmark dataset, this study has investigated the computational performance and signature-size trade-offs of lattice-based post-quantum threshold ring signatures (LoTRS). The analysis was performed on 14 benchmark configurations, the main evaluation being based upon eight threshold ring-signature settings obtained by setting (N=32) and (N=100) and by varying the threshold. The performance was evaluated by the key generation, key aggregation, signing and verification times and communication and storage overheads by the signature, public-key, secret-key and ring public-key sizes. The figure indicates that the number of threshold and ring size have a significant impact on signing and verification costs. The signing time was 97.81 to 923.60 ms and verification time was 33.95 to 311.00 ms for the main configurations. In comparison, the size of the signature grew comparatively slowly from 42.91 to 54.82 KB. The most significant gains with respect to configuration were in key-aggregation time and ring public-key size. Additional analyses using correlation and regression also showed a high correlation between threshold level and computational cost and a high correlation between ring size and signature size. It is seen that LoTRS provides a relatively moderate growth in the signature size while the computational overhead has increased significantly, thus pointing to a key scalability trade-off for practical implementation of post-quantum cryptography.