Authors: (1) Sheng Liu, School of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, Hunan, China; (2) Chengqing Li (Corresponding author), School of Computer Science, Xiangtan University, Xiangtan 411105, Hunan, China; (3) Qiao Hu, School of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, Hunan, China. Authors: Authors: (1) Sheng Liu, School of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, Hunan, China; (2) Chengqing Li (Corresponding author), School of Computer Science, Xiangtan University, Xiangtan 411105, Hunan, China; (3) Qiao Hu, School of Computer Science and Electronic Engineering, Hunan University, Changsha 410082, Hunan, China. Table of Links Abstract and 1. Introduction Abstract and 1. Introduction 2. Description of two analyzed image encryption algorithms 2. Description of two analyzed image encryption algorithms 3. Cryptanalysis of IEATD 3. Cryptanalysis of IEATD 4. Cryptanalysis of IEACD 4. Cryptanalysis of IEACD 5. Conclusion, Acknowledgements, and References 5. Conclusion, Acknowledgements, and References 4. Cryptanalysis of IEACD To cope with the insecurity problems of IEATD reported in [13], multiple confusion and diffusion operations are appended, making the algorithm become another algorithm IEACD. In fact, the original keystream generation mechanism indeed exists a serious pitfall, which leads to that the patched algorithm IEACD still cannot withstand chosen-plaintext attack. In this section, three weaknesses of IEACD are first analyzed to facilitate description of the following chosen-plaintext attack. 4.1. Three weaknesses of IEACD 4.1. Three weaknesses of IEACD • The real size of key space is much smaller than the expected one Due to the limitation of finite-precision presentation, dynamics of any chaotic system is definitely degraded when it is implemented in a digital device. As investigated in [14, 15, 16], the structure of the statemapping network (SMN) of a digitized chaotic system implemented with fixed-point precision e+1 is largely dominated by that with precision e. The short period problems of PRNG based on Logistic map (5) implemented in a digital device (with fixed-point arithmetic or floating-point arithmetic) were comprehensively discussed in [15]. As shown in Fig. 3, discretized Ikeda system obeys this rule also. No matter what the precision is, the SMN of discretized Ikeda system follows the following rules: 1) an SMN is composed of some weakly connected components; 2) there are some self-loops (an edge connecting a node to itself); 3) As for each connected component, there is one and only one cycle (including special cycle, selfloop), and every node evolves to it via a transient process; 4) Many nodes have two and only two parent nodes. Generating a pseudo-random number sequence by the orbits determined by a chaotic map is actually walking along a path of an SMN. Now, one can see that the period of a sequence by solving the discretized Ikeda system may be very short (even only one). So, there are a number of equivalent secret keys and invalid secret keys as for the function of IEATD and IEACD. Note that such pitfall always exists no matter how large the precision e gets. • Insensibility of keystream generation mechanics • Improper configuration of keystream The two-round crossover diffusion is performed to resist chosen-plaintext attack and differential attack, but the keystream Z used in permutation is wrongly reused in the diffusion part, which makes the algorithm more insecure. From Eq. (6), (8) and (10), one has 4.2. Chosen-plaintext attack on IEACD* 4.2. Chosen-plaintext attack on IEACD* 4.2.1. Determining Z 4.2.1. Determining Z Z Referring to Eq. (15) and XORing the two cipher-images, one has In the process of determining Z 0 , as for a known or given element z 0 (i), one attempts to find its neighbor by verifying whether the corresponding equation holds. Therefore, Z 0 is reconstructed by seeking the relative positions of elements. As mentioned before, after constructing Z 0 , the permutation vector Z can be restored via z(u(i)) = z 0 (i) for i = 0 ∼ (HW − 1) Now, the attack in case of b ≤ 1 is discussed. In fact, the special cases of b can be identified through Eq. (18a) and (18b). If b = 1, since z 0 (b − 1) is the first element in Z 0 , no element can be found via Eq. (18b). Hence, one should determine z 0 (0) via Eq. (18a) and then find the remainder of Z 0 through Eq. (18c). If b = 0, the attack is failed. Since Eq. (18a) and (18b) both no longer hold, one would directly attempt to determine z 0 (1) through 5. Conclusion This paper analyzed security performance of an image encryption algorithm based on a first-order time-delay system IEATD and the enhanced version IEACD. Although another research group proposed a chosen-plaintext attack on IEATD, we presented an enhanced attack using the correlation between adjacent vectors of one plain-image and the corresponding cipher-image. Although IEACD is designed by the attacking group with intention to fix the security defects of IEATD, there still exist some security pitfalls, such as invalid secret keys, insensibility of keystream generation mechanics, and improper configuration of keystream. Based on these, we designed an efficient chosen-plaintext attack and verified it with extensive experiments. The serious insecurity of the two algorithms cannot be improved by simple modifications. They can work as typical counterexamples to remind us to recast scenario-oriented image encryption algorithms following the guidelines and lessons summarized in [9, 7, 3]. Acknowledgement This work was supported by the National Natural Science Foundation of China (no. 61772447), Scientific Research Fund of Hunan Provincial Education Department (no. 20C1759), and Science and Technology Program of Changsha (no. kq2004021). References [1] A. L. Abu Dalhoum, B. A. Mahafzah, A. A. Awwad, I. Aldamari, A. Ortega, and M. Alfonseca, “Digital image scrambling using 2D cellular automata,” IEEE Multimedia, vol. 19, no. 4, pp. 28–36, 2012. [2] G. Ye and X. Huang, “An image encryption algorithm based on autoblocking and electrocardiography,” IEEE Multimedia, vol. 23, no. 2, pp. 64–71, 2016. [3] C. Li, Y. Zhang, and E. Y. Xie, “When an attacker meets a cipherimage in 2018: a year in review,” Journal of Information Security and Applications, vol. 48, p. art. no. 102361, 2019. [4] O. Mannai, R. Bechikh, H. Hermassi, R. Rhouma, and S. Belghith, “A new image encryption scheme based on a simple first-order timedelay system with appropriate nonlinearity,” Nonlinear Dynamics, vol. 82, pp. 107–117, 2015. [5] A. Jolfaei, X.-W. Wu, and V. Muthukkumarasamy, “On the security of permutation-only image encryption schemes,” IEEE Transactions on Information Forensics and Security, vol. 11, no. 2, pp. 235–246, 2016. [6] C. Li, D. Lin, J. Lu, and F. Hao, “Cryptanalyzing an image en- ¨ cryption algorithm based on autoblocking and electrocardiography,” IEEE Multimedia, vol. 25, no. 4, pp. 46–56, 2018. [7] M. Preishuber, T. Huetter, S. Katzenbeisser, and A. Uhl, “Depreciating motivation and empirical security analysis of chaos-based image and video encryption,” IEEE Transactions on Information Forensics and Security, vol. 13, no. 9, pp. 2137–2150, 2018. [8] J. Chen, L. Chen, and Y. Zhou, “Cryptanalysis of image ciphers with permutation-substitution network and chaos,” IEEE Transactions on Circuits and Systems for Video Technology, vol. 31, no. 6, pp. 2494– 2508, 2021. [9] C. E. Shannon, “Communication theory of secrecy systems,” Bell System Technical Journal, vol. 28, no. 4, pp. 656–715, 1949. [10] X. Chai, X. Fu, Z. Gan, Y. Lu, and Y. Chen, “A color image cryptosystem based on dynamic dna encryption and chaos,” Signal Processing, vol. 155, pp. 44–62, 2019. [11] Z. Hua, Z. Zhu, Y. Chen, and Y. Li, “Color image encryption using orthogonal latin squares and a new 2D chaotic system,” Nonlinear Dynamics, vol. 104, p. 4505–4522, 2021. [12] K. Ikeda, H. Daido, and O. Akimoto, “Optical turbulence: chaotic behavior of transmitted light from a ring cavity,” Physical Review Letters, vol. 45, no. 9, p. 709, 1980. [13] M. Li, H. Fan, Y. Xiang, Y. Li, and Y. Zhang, “Cryptanalysis and improvement of a chaotic image encryption by first-order time-delay system,” IEEE Multimedia, vol. 25, no. 3, pp. 92–101, 2018. [14] C. Fan and Q. Ding, “Analysing the dynamics of digital chaotic maps via a new period search algorithm,” Nonlinear Dynamics, vol. 97, no. 1, pp. 831–841, 2019. [15] C. Li, B. Feng, S. Li, J. Kurths, and G. Chen, “Dynamic analysis of digital chaotic maps via state-mapping networks,” IEEE Transactions on Circuits and Systems I: Regular Papers, vol. 66, no. 6, pp. 2322–2335, 2019. [16] C. Li, K. Tan, B. Feng, and J. Lu, “The graph structure of the gen- ¨ eralized discrete arnold cat map,” IEEE Transactions on Computers, 2021. [17] C. Li and K.-T. Lo, “Optimal quantitative cryptanalysis of permutation-only multimedia ciphers against plaintext attacks,” Signal Processing, vol. 91, no. 4, pp. 949–954, 2011. [18] C. Li, D. Lin, and J. Lu, “Cryptanalyzing an image-scrambling encryption algorithm of pixel bits,” IEEE Multimedia, vol. 24, no. 3, pp. 64–71, 2017. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. This paper is available on arxiv under CC by 4.0 Deed (Attribution 4.0 International) license. available on arxiv
