New Strategy for Evaluating Differential Clustering Effect of uBlock

uBlock is a family of block ciphers designed by Wu et al. and was selected as one of the two first-prize winners in 2018 China’s National Cipher Designing Competition. Prior to the work described here, there have been no third-party differential cryptanalytic results on uBlock. Due to its fast diffusion and carefully designed linear layer, it’s hard to get the differential bounds of active S-boxes when the S-box and linear layer are modeled accurately. Therefore, in this paper, we mainly focus on evaluating the differential clustering effect of uBlock, and mount key recovery attacks based on them. We adopt a combined strategy to evaluate its clustering effect. As a result, we obtain many 7- and 8-round differentials for uBlock-128 and uBlock-256. Based on these differentials, we provide 10- and 11-round differential attacks on uBlock-128/128 and uBlock-128/256, respectively. 10-round differential and rectangle attacks on uBlock-256/256 are also proposed. All the provided attacks are the best ones to our knowledge. However, the full-round security of uBlock is far from being threatened.