# Differential attack on repeating pairs of zeros in blocks

Let's consider $$5$$ types of vulnerabilities in one round of encryption in 128-bit cipher.

1. We got 00 at the beginning of the block (the first two bits are two zeros).
2. We got 00 in the middle of the block.
3. We got 00 at the end of the block.
4. We got 00 randomly put in the block (it depends on the key).
5. We got 00 shifted in the block in consecutive rounds (let's say in first-round they are $$3,4$$ bits, in the second-round $$5,6$$, then on the $$7,8$$ position and so on).

And let's say they will occur only if the number is equal to $$0+4n$$ or $$2+4n$$ (it depends on the key). Let's say these are only vulnerabilities in that encryption. If we consider $$20$$ rounds which type is hardest to attack?

For sure in case 3., if we will have a bit of luck and in all-around such vulnerabilities will occur for $$0+4n$$ numbers we can find 00 at the end of every encrypted number of the form $$0+4n$$. Case 1. is also insecure for similar reasons. So I think it would be best if such two zeros will occur somewhere in the middle or will be shifted by some number of bits in consecutive rounds or maybe put randomly in the block.

• 1) 25% of messages contain 00. 2) Why do you use word "vulnerabilities"? We speak about vulnerability only if it is clear what attack is possible. Please explain why you us this word or remove it. 3) Why have you posted this text? What exactly is your question? – mentallurg Jan 1 at 20:53
• 1) Yes. And only odd numbers in this example. 2) I don't know better word (maybe disadvantages?). I thought if there are some such obvious schemes in encrytpion we can name them "vulnerabilities". 3) I want to know what is the hardest to attack. I got such a problem in my cipher and I have no idea how to compare option 2.,4. and 5. I suspect that this can be attacked with truncated differential cryptanalysis (if I understand it right) but I don't know how to make such attack (or how to estimate the complexity of its best version). – Tom Jan 1 at 21:54