• RSS
  • Twitter
  • FaceBook

Security Forums

Log in

FAQ | Search | Usergroups | Profile | Register | RSS | Posting Guidelines | Recent Posts

Steganography.

Users browsing this topic:0 Security Fans, 0 Stealth Security Fans
Registered Security Fans: None
Goto page 1, 2  Next
Post new topic   Reply to topic   Printer-friendly version    Networking/Security Forums Index -> Cryptographic Software and Hardware

View previous topic :: View next topic  
Author Message
JustinT
Trusted SF Member
Trusted SF Member


Joined: 17 Apr 2003
Posts: 16777215
Location: Asheville, NC, US / Uberlāndia, MG, Brazil

Offline

PostPosted: Sun Aug 01, 2004 10:36 pm    Post subject: Steganography. Reply with quote

Setting aside convenience, or any consideration for additional overhead associated with other layers coupled with encryption, any component that adds to the workload of an attacker, and to the complexity of the attack he or she wishes to perform, is essentially an effective component, even if the degree is small and the effect is too insignificant to go to this extreme. To be politically correct, it isn't entire useless, depending on the methodology.

Steganography satisfies one particular goal - security through obscurity, by obfuscating the actual existence of data, whether it's encrypted or not. It's obvious that obfuscating the existence of encrypted could either frustrate an attacker more so, or just not do any good at all. At the most, it would cost additional overhead. Personally, I don't use steganographic techniques to secure any data; I feel cryptography is sufficient enough at securing data, be it stored personally, or disseminated to colleagues through transit. However, there are very ingenious concepts being developed amongst the community; my "favorite", thus far, is Hydan, since it actually demonstrates promising applications for steganography in conventional practice.

Perhaps the one solid reason that steganography hasn't gained such partnership with the cryptographic community is because as cryptographers, we design primitives to rely on structure soundness; we design them to not rely on security through obscurity. We strive for confidentiality. In other words, when an attacker applies traffic analysis, we expect to be comfortable with the fact that they know communication is being initiated, who is initiating and communicating, the amount of data is being communicated, and when the communication takes place, so long as they aren't capable of revealing the plaintext meaning of the data being handled during such communication. Steganography attempts to mask the "scent" of these issues that sparks the awareness of an attacker; thus, obscuring the existence of encrypted data that would prompt traffic analysis.

There is no de facto methodology, or specified standard, and via cryptography, we can achieve levels of security that are sufficient for most applications. It's actually good practice, since it teaches us to build cryptographic primitives and protocols that are resilient enough to allow open dispersal. This security is achieved when all cards are on the table, but are of no use to the attacker. Steganography attempts to fill a niche that doesn't practically need filling, in my opinion. It definitely has its merit, and I can't see a reason to believe that it's worthless, but at the same time, I can't see an imminent reason to require integration into a cryptographic policy or methodology. For those who wish to use it - go ahead. If you're using weak cryptography, the most it could do is lengthen the time it takes for an attacker to become aware of this. If you're using strong cryptography, you probably have little need for it.

Security through obscurity, in itself, isn't a good habit to fall into, and if you can design a system to rely on Kerckhoffs's Principle, and on this principle only, you're benefitting yourself. The fact is, if you use steganography, use it as an added, optional measure. If you must rely on it, you should probably take a second look at the cryptography you're trying to obscure. The increase of security in relying on security through obscurity is a magnitude smaller than the decrease of security in doing so. Designing a channel to be secure diminishes the need for it to be covert. Overall, steganography is an interesting branch of study, and carries beneficial properties. If overhead, and other cost-conscious aspects, is not of any concern, by all means, keep on playing with it. Most importantly, pay much more attention to the cryptography and the security of its deployment. This is a bigger issue.
Back to top
View user's profile Send private message Visit poster's website
M3DU54
Trusted SF Member
Trusted SF Member


Joined: 11 May 2002
Posts: 1
Location: Las Palmas de Gran Canaria

Offline

PostPosted: Sun Aug 01, 2004 11:20 pm    Post subject: Reply with quote

UziMonkey wrote:
My point was, why bother hiding it if evesdroppers have no chance of reading the message anyway? Who cares if they know you're sending data. I mean, sure, this would have uses in cases of espionage and all sorts of James Bondish stuff, but it doesn't make it any more secure for normal users. It's only practical use is if you're trying to be sneaky.


Check RIP act, encryption alone may not be enough to safeguard your freedom. Also note previously made points about suspicion - Sometimes the very act of sending a concealed message will attract unwanted attention.

Unconvinced? Imagine that you live in a hardline country where your particular ideology could get you killed and encryption is considered sure sign of subversion ... you will see that, in such circumstances, it is not enough to conceal the content of a message - you must also operate via covert and plausible channels.

Steganography aims to provide just such covert channels within credible traffic. It makes a great deal of sense if done correctly.

I guess the problem is that many people seem to confuse the two technologies... encryption and steganography address two completely different issues and are NOT interchangeable. Once this is understood and accepted then the unique value of each becomes apparent.


JustinT wrote:
Steganography attempts to fill a niche that doesn't practically need filling, in my opinion. It definitely has its merit, and I can't see a reason to believe that it's worthless, but at the same time, I can't see an imminent reason to require integration into a cryptographic policy or methodology. For those who wish to use it - go ahead. If you're using weak cryptography, the most it could do is lengthen the time it takes for an attacker to become aware of this. If you're using strong cryptography, you probably have little need for it.


As much as I respect Justins superior understanding of cryptographic method I feel we diverge slightly on this. Perhaps it is simply because I feel that steganography cannot be valued accurately against purely cryptographic values - there is a deeply political element that is not addressed by pure crypographic strength. Out-in-the-open cryptography may work well in countries which tolerate it, democracies protecting free thought and speech for example - but there are others living in ruthless states where the mere use of crypto could result in imprisonment, torture or capital punishment. As a member of Amnesty International I know just how common these injustices are and, whilst such conditions exist, there will remain a VERY real need for strong steganography.

M3Dz
Back to top
View user's profile Send private message
JustinT
Trusted SF Member
Trusted SF Member


Joined: 17 Apr 2003
Posts: 16777215
Location: Asheville, NC, US / Uberlāndia, MG, Brazil

Offline

PostPosted: Mon Aug 02, 2004 1:12 am    Post subject: Differing views. Reply with quote

M3DU54 wrote:

As much as I respect Justins superior understanding of cryptographic method I feel we diverge slightly on this. Perhaps it is simply because I feel that steganography cannot be valued accurately against purely cryptographic values - there is a deeply political element that is not addressed by pure crypographic strength. Out-in-the-open cryptography may work well in countries which tolerate it, democracies protecting free thought and speech for example - but there are others living in ruthless states where the mere use of crypto could result in imprisonment, torture or capital punishment. As a member of Amnesty International I know just how common these injustices are and, whilst such conditions exist, there will remain a VERY real need for strong steganography.


Point well taken and appreciated.

I can't disagree with the politics that surround the use of cryptography. Actually, this is right on the money, when considering other portions of the globe that look at cryptography as a resistance movement, rather than harboring of one's natural right to privacy. In many of these areas, one's natural right to privacy often isn't honored. My thoughts on the matter focus on the design of cryptographic systems, in the sense that I've noticed the integration of steganography as a "relied upon" component for the overall security of the system. It seems my focus is directed towards the system of security, rather than the system of politics. However, I can't argue with the realistic need for such a covert and plausible channel of communication, within the confinement of a governing area that abhors the mere use of confidential conversion.

It is vital that one understand the purposes of encryption, as opposed to what steganography aims to provide. Encryption provides confidentiality by concealing the meaning of information; steganography conceals the existence of that information. It's more so a secondary layer, when approached from that standpoint, but they are quite the opposite, in purpose. My opinion isn't so much related to the realistic political need for this technology, as a standalone measure, but rather, what folks realize about what it can and can not provide, as a component in a cryptographic system.

Practically, where one can assume that cryptography is a non-legally-stifled activity, steganography's method of existence-obfuscation isn't as much of a necessity and can sometimes pose false hopes to those who misunderstand its definition. I design with raw semantics in mind, which naively disregard politics, but you're absolutely right. When you zoom out on the wider application of privacy, steganography does provide a relatively secure manner of being confidential. This is one of the naive trademarks of a cryptographic designer - everything is kept on a mathematical and structural basis. As such, our conditional and environmental concerns revolve around the inner-workings, rather than the outer-applications. When you disregard the politics, instead, the scale tips the opposite way. Necessity fades into optionality.

Perhaps this calls for another branch of political science - politicryptography [/play_on_words]. I never said I was a comedian. ;) But, in all honesty, your opinion does raise a very imminent threat to the future of privacy. Sometimes, privacy isn't enough; sometimes, you have to ensure that your privacy is private. It goes to show that totalitarianism is more than a 1984-Orwell-scented term; it's a reality.

(How ironic, that my existence began in the year of that literature's setting. Spooky. ;))
Back to top
View user's profile Send private message Visit poster's website
AnonViper
Just Arrived
Just Arrived


Joined: 13 Mar 2004
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 7:02 pm    Post subject: Reply with quote

*deleted*

Last edited by AnonViper on Sat Dec 04, 2004 2:14 am; edited 1 time in total
Back to top
View user's profile Send private message Visit poster's website
necro1234
Just Arrived
Just Arrived


Joined: 13 Oct 2003
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 7:31 pm    Post subject: Reply with quote

Hi AnonViper

Just curious, I understand full well that the way alot of Stego apps work its quite easy to see that there is "garbage" data in the image/sound file.
In the case of something that operates alont the lines of Hide In Pic though where the data is spread out in the image, and it is spread according to the password you gave.
As long as the original image was not available (you created your own for example and then destroyed the original) how or what kind of method would/could be used to see that there is infact something else hidden in there, ( and to prove it).

Any info would be appreiated, Ive read up on apps like OutGuess and these stego detection applications, but Ive been looking since I found HIP and have found no app as yet that admits to detecting its method.

Thanx alot

Sheldon
Back to top
View user's profile Send private message
necro1234
Just Arrived
Just Arrived


Joined: 13 Oct 2003
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 7:56 pm    Post subject: Reply with quote

Hi guys

Since there is this topic here lemm throw this one in.
Rubberhose for linux.
Ive read the papers on how this works, you for example encrypt your 20Gb HDD with random data.
You then make a partition1 of 10Gb, it then chooses which sectors randomly it is gonna use, you then make a 4Gb partition2, it does the same.

Partition1 and partition2 dont know about each other.
So you need to mount both so that P1 knows about P2 at that time, then you can write your data to whichever partition.

Is this not similar to the way which the BestCrypt Hidden Containers work?
In that If you make a 20Gb volume with a 4Gb hidden volume, the 20Gb volume does not know the 4Gb one exists.
It just sees the scrambled data, so it still shows it has 20Gb free space.

Any thoughts on this?
Also does anyone know whats become of the group making Rubberhose, Ive used it a little on Debain, I was waiting and hoping for the NT/2K version to be released.

Thanx guys

Sheldon
Back to top
View user's profile Send private message
AnonViper
Just Arrived
Just Arrived


Joined: 13 Mar 2004
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 8:19 pm    Post subject: Reply with quote

*deleted*

Last edited by AnonViper on Sat Dec 04, 2004 2:15 am; edited 1 time in total
Back to top
View user's profile Send private message Visit poster's website
M3DU54
Trusted SF Member
Trusted SF Member


Joined: 11 May 2002
Posts: 1
Location: Las Palmas de Gran Canaria

Offline

PostPosted: Mon Aug 02, 2004 9:13 pm    Post subject: Reply with quote

necro1234 wrote:
Hi AnonViper

Just curious, I understand full well that the way alot of Stego apps work its quite easy to see that there is "garbage" data in the image/sound file.
In the case of something that operates alont the lines of Hide In Pic though where the data is spread out in the image, and it is spread according to the password you gave.
As long as the original image was not available (you created your own for example and then destroyed the original) how or what kind of method would/could be used to see that there is infact something else hidden in there, ( and to prove it).

Well, since many (most) steganography programs have been written by the clueless detection often isn't difficult. Some nameless cretins actually SELL a product that simply appends scrambled data to a legitimate image file.

Conversely, strong steganography aims to blend the data in such a way that statistical analysis is difficult or impossible. Correctly done this is very difficult to detect - if the encryption used is statistically neutral and without identifiable characteristics it can be virtually impossible to prove that steganography has been used at all, even where the stegging method is known.

I'd disagree with the supposed '1% rule' as it seems altogether too arbitrary. How much you can safely steg into a given file surely depends very much on the file format itself, also some (particularly artificial) images are wholly unsuitable for stegging at all whereas some images and formats lend themselves well to heavy stegging (Chaotic natural images with an R6G5B6 or 24bit RGB for example, are far less constrained than, say, palletised images or JFIFs) Any decent steganography tool should therefore vary its stegging depth accordingly and de-steg blindly (due to lack of depth or positional markers)

Unfortunately most commercial steganography efforts use either markers or steg uniformly across each file which pretty much defeats the purpose if you're using a medium such as the internet where authorities can easily 'fish' en-mass for such images.


But, despite the recent hype, steganography really isn't new. I believe that historically the russians, true innovators of the clandestine arts, used similar techniques to encode messages into alternating blobs that made up newsprint images. Each alternate 'blob' was an average of the legitimate blobs to its left and its right. To 'encode' on these redundant blobs you would increase or decrease the blobs weight slightly against this calculated average and the overall impact on the final image was negligible. Despite this leaving a statistical anomaly the method was never discovered during its lifetime, due in part to the huge resources wasted on bogus numbers stations which churned out broadcast garbage with tantalising statistical biases. There are probably older examples that I am not aware of.


M3Dz
Back to top
View user's profile Send private message
necro1234
Just Arrived
Just Arrived


Joined: 13 Oct 2003
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 9:49 pm    Post subject: Reply with quote

I see, interesting
Makes me stand by my statement that when there is something created to protect, not long after there will be something to defeat it.

Im curious though now after reading your post, if for example I have a 4Mb BMP image and use say HIP (cause I still think tis neat) and fill the max 50% of the image a few times over using different passwords to scatter these subtle changes all over.
Then add my say 350Kb file to the image with my final password if this would help at all as far as tracking down what has changed.
It would be kinda like scrambeling data firstly before encrypting.
I picture from a camera ofcause can be seen to have changed, and ofcause JPG's compression I know all to well show up easily.
But something randomly generated (like throwing paint on canvas), if this was used in the way Ive just said, to the human eye I doubt there will be anything to make a person think twice about the image, and if someone had to physically open up the file and check each pixel, with all the randomly changed data I would only guess this would help a person to keep the attacker off of the file in question.

Thanx for the imput AnonViper

Cheers

Sheldon
Back to top
View user's profile Send private message
AnonViper
Just Arrived
Just Arrived


Joined: 13 Mar 2004
Posts: 0


Offline

PostPosted: Mon Aug 02, 2004 11:29 pm    Post subject: Reply with quote

*deleted*
Back to top
View user's profile Send private message Visit poster's website
Display posts from previous:   

Post new topic   Reply to topic   Printer-friendly version    Networking/Security Forums Index -> Cryptographic Software and Hardware All times are GMT + 2 Hours
Goto page 1, 2  Next
Page 1 of 2


 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

Community Area

Log in | Register