Paper 2017/534

Functional Graph Revisited: Updates on (Second) Preimage Attacks on Hash Combiners

Zhenzhen Bao and Lei Wang and Jian Guo and Dawu Gu

Abstract

This paper studies functional-graph-based (second) preimage attacks against hash combiners. By exploiting more properties of cyclic nodes of functional graph, we find an improved preimage attack against the XOR combiner with a complexity of $2^{5n/8}$, while the previous best-known complexity is $2^{2n/3}$. Moreover, we find the first generic second-preimage attack on Zipper hash with an optimal complexity of $2^{3n/5}$.

Note: Full version, revise the content regarding the expected number of trials reaching deep-iterates simultaneously