Subspace differential uniformity
Sondre Rønjom, Arne Sandrib, Joakim Sunde · Cryptography and Communications · 2026
Abstract The main contribution of this paper is to introduce Subspace Differential Uniformity (SDU) for S-boxes and block ciphers. The SDU is essentially a measure of how well any function spreads input differences clustered in affine subspaces away from affine clusters in output differences. We provide some lower bounds for the SDU and describe an efficient algorithm for computing the SDU. Moreover, we provide results for some popular classes of S-boxes up to $$n=8$$ .