meta user
network profile
| bio | website | |
|---|---|---|
| location | ||
| age | ||
| visits | member for | 5 months |
| seen | May 17 at 0:37 | |
| stats | profile views | 0 |
|
Jan 28 |
awarded | Supporter |
|
Jan 19 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? @stevenh Eve (or, in the parlance of steganography, Willie or the warden) should not even be able to suspect a hidden message better than chance. Roughly, given two potential covertexts, one of which contains a message and the other which does not, the warden should not be able to determine which text contains the hidden message with probability > 0.5, without knowledge of the key, even when the method is known. It is often the case that determining which bits of the covertext carry the message is unknown unless the key is known, even when it is known that the covertext contains a message. |
|
Jan 18 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? @BramCohen I'm sorry. I was being unnecessarily obtuse. I was simply alluding to my feeling that there is not sufficient detail provided here for me to fully analyze your scheme. E.g., you mention a "specifically designed stream cipher" but do not detail an IV (or if there is one at all). I am still unclear on what the set of equations your row reduction is supposed to be solving. And what happens if this set of equations has no solution (if that's even possible)? What if there is more than one solution (if that's even possible)? |
|
Jan 17 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? @BramCohen Yes. But what is easy for some is not necessarily easy for all (eg: Mac OS X requires Xcode for a C compiler which is required for PyCrypto). Every minute installing software is a minute not analyzing. Given the level of detail you've provided here and in the text files in your repo, is it also your intent that your Python code should be the primary documentation for your method? I worry that this would limit cryptanalysis only to those conversant in Python. I do not think you have provided enough information in text for me to reproduce (and analyze) your algorithm on my own. |
|
Jan 17 |
awarded | Informed |
|
Jan 17 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? @Bram Cohen: They are simple command line tools which require Python3 and the PyCrypto and sha3 libraries. Is the intent to limit feedback only from those who already have or are willing to install these? |
|
Jan 17 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? @stevenh: I believe you are misunderstanding one of the primary goals of steganography. The goal is to embed messages in covertexts which are indistinguishable from covertexts without embedded messages, even when the embedding method is completely known to the attacker. This is the exact opposite of "security by obscurity". "Security by obscurity" refers to obscurity of the methods, not of the message. |
|
Jan 16 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? I'm sorry. I guess I'm not quite able to fill in all the details myself. I think "contiguous section of sixteen bytes" refers to covertext not message, right? Is it inclusive of alternates? Is there a separate xor per alternate cover? Is the result of "xors the outputs" a bit or a word? Can I assume that the "outputs" are the output blocks of the AES-encoding? What is the "specific value"? What are the rows that you're applying row reduction to? I think that you're using row reduction to solve some system of equations. If so, what is the set of equations? |
|
Jan 16 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? Your question asks for a review of the "cryptography" and "math" of your method but I'm not sure there is enough information here to provide that. Is the intent for us to infer this from the source code directly? I fear that you may not receive adequate response if that is the case. A typical presentation of cryptographic methods starts with a full description of the algorithm (often in pseudocode), an outline of the various design choices (choice of constants, why n rounds and not n+1 rounds, etc...), and your own cryptanalysis of the method (what attacks have you tried, etc...). |
|
Jan 16 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? Ok. You may want to review your Explanation.txt file where the terms plaintext and ciphertext are also used, possibly in a confusing way. Aside from the terminology, I feel like there are still quite a few details missing. You mention a "custom stream cipher" but only give a very broad description. You describe the three layers as "encode(pack(encrypt())))" , but give only the coarsest description of each (even in the Description.txt), e.g. the arguments to encrypt aren't specified. |
|
Jan 16 |
comment |
Is a steganographic technique which has a universal decoder novel/secure? Your terminology is a little confusing to me. When you mention "plaintext with possible alternates" are you referring to a covertext? Similarly, what does "flipping an alternate" mean? And what does it mean for "alternates have at least fifteen fixed bytes between them"? |
