- This topic has 14 replies, 10 voices, and was last updated 3 years, 11 months ago by
.
-
Posts
-
Harry any help? I have cracked 6A but 6B is confusing me a lot. Any hints please?
[Perhaps later, but not yet!! Harry]
We have 6A too but no clue about 6B, we think we know the kind of cipher but it needs a kind of random alphabet and need hints as to how to find it please!
The answer to 6A and a lots of research of the what is mentioned about 6B in 6A should really be all you need to crack it! I’m sure you guys are close!
Could I have some help too please? I think I know the keyword from [… edited by Harry] from 6B working backwards. But it is [….]? […..]? In that order? Hope I haven’t put in any spoilers! [I have taken out the spoilers, Harry]
[Can’t give hints today, but there is nothing here that is not as described by Pearl. Hope that helps, Harry]
@Harry could we get a hint for 6B now since it’s past the deadline?
I found it was really quite a step up compared to the previous problems. It required a *lot* of research, and I eventually came across an academic paper on this, which suggested using a genetic algorithm to solve it (a slightly more elegant alternative to brute force). I have an old copy of numerical recipes (it’s a great book about algorithms) and after quite a lot of Rust (a nice C++ alternative which avoids lots of memory bugs which usually cause security risks) code, I was able to solve the thing.
There’s definitely simpler ways to do it (I don’t recommend writing a genetic algorithm because I think it’s a case of me having only a hammer and therefore treating everything like it’s a nail). Note that Pearl gives some advice about what to do in 6A – specifically around details for the transposition key. It should be easier to use an abstraction here – consider both stages separately. To work out whether solutions are correct things like the index of coincidence are going to be your friend (because solving the transposition stage won’t give you plaintext).
I hope these aren’t too many hints, but I assume (hope) Harry moderates them before these are posted to the forum so I don’t give anything away that I’m not supposed to.@Teymour_aldridge, by “genetic algorithm”, do you mean a hill-climbing algorithm that varies the permutation
for the transposition stage, and keeping the new permutation if it results in a higher IoC?This works just fine for a transposition with less than about 6 columns. For more than than, you can end up
with several degenerate global maxima.If you mean something else, then I would be glad to hear about it.
Anyway, 6B has few enough columns that brute-forcing the transposition stage is feasible.
For anyone who is struggling (if Harry lets this through) a nice way to do this would be to look at how you could transform it into a cipher that you can solve, perhaps even one that has come up in the competition already… From there it should all be plain sailing, Good luck!
It’s similar to a hill-climbing algorithm, but instead of just trying to climb a hill the process for selecting a better solution is inspired by evolutionary mechanisms. This relies on the solutions being somehow related (which is true for this case – if you take the “correct” solution and mutate it, the further you) and one effectively discards families of solutions which are related based on this – so often more effective. Fundamentally if one is to treat this as a combinatorial optimisation problem there are a lot of algorithms which can be used. I just picked a genetic algorithm because I thought it would be cool to implement one.
Yes it only works for shorter key lengths (I had a look on the internet and one paper said this sort of approach is only effective for a key length of 7), but Pearl did give indication of the key length.@Teymour_aldridge, could you please post a reference to the key-length-7 paper (and any other papers)?
Thank you.6B Hint. OK with just over 3 hrs to go when I posted this…
Think INSIDE THE BOX what is the name of the cipher?
[Sorry, the Elves were sleeping when this came in. Nice hint though. Harry]
Which one eats message 52553? Is it DASHER, DANCER, PRANCER, VIXEN, COMET, CUPID, DONNER or BLITZEN?
Maybe Harry will think that my clock runs in advance and what follows is TNT.
A help to simplify substitution. We can’t use the frequency counter of the site because it needs only groups of 1 letter.
So I substitute 1 letter in place of 2 letters.
• Copy the large text divided in blocks of 2 into a text editor, in uppercases.
• Select the first group “AD” and replace all with “a” lowercase in red.
• Go to the next group of 2 “DF” and replace all with “b”, and so on.
You will end with a simple substitution cipher easier to solve. There is a big crib.
If you compare the plain text with the text in groups of 2 you can fill in the ADFGVX grid. Look at the key: EUREKA!
🙂
Et voila.Hi Madness,
The main paper is:
http://iieng.org/images/proceedings_pdf/U1215011.pdf
There is another one which was interesting as well.
https://people.cs.uct.ac.za/~jkenwood/JasonBrownbridge.pdf
All the best
- You must be logged in to reply to this topic.