What do I need to know to design a cryptosystem?
in [Cryptography]
|
Prev: An Important Distinction being stymied by Ignorance.
Next: Using a kind of running accumulation of ciphertext as chainingvalue of encryption
From: cplxphil on 12 Mar 2010 23:07 Hi all, I recently became interested in trying, in my spare time, to develop a cryptosystem. I have some math/CS training, and a degree in mathematics. What do I need to know to design a cryptosystem? What steps are involved? I would assume designing the protocol and then modifying it based on how easy it is to attack would be some of the steps. Is there a book that details what I need to know/do? Thanks, Phil
From: J.D. on 12 Mar 2010 23:30 On Mar 12, 11:07 pm, cplxphil <cplxp...(a)gmail.com> wrote: > Hi all, > > I recently became interested in trying, in my spare time, to develop a > cryptosystem. I have some math/CS training, and a degree in > mathematics. What do I need to know to design a cryptosystem? What > steps are involved? I would assume designing the protocol and then > modifying it based on how easy it is to attack would be some of the > steps. > > Is there a book that details what I need to know/do? > > Thanks, > Phil This advice is a little dated now, but most of it still seems valid: http://www.schneier.com/crypto-gram-9810.html#cipherdesign If there was one thing that would best prepare you for designing a cipher that is worth a damn, it would probably be getting good at breaking ciphers. Which doesn't just mean reading about how other people have broken ciphers -- it means figuring out on your own how to break ciphers without 'looking at the answers' (so to speak). Once you have developed that skill, only then can you approach your own designs from the perspective of a competent attacker, and thus see flaws and weaknesses and the best ways to correct them. One place to start learning how to break ciphers is here: http://www.schneier.com/paper-self-study.html
From: David Eather on 13 Mar 2010 00:25 On 13/03/2010 2:07 PM, cplxphil wrote: > > Hi all, > > I recently became interested in trying, in my spare time, to develop a > cryptosystem. I have some math/CS training, and a degree in > mathematics. What do I need to know to design a cryptosystem? What > steps are involved? I would assume designing the protocol and then > modifying it based on how easy it is to attack would be some of the > steps. > > Is there a book that details what I need to know/do? > > Thanks, > Phil You could also try "modern cryptanlysis" by C Swenson. It is not complete but will give you some ideas on what to avoid. The book could also do with a good copy editor - but that's a different problem.
From: adacrypt on 13 Mar 2010 04:00 On Mar 13, 5:25 am, David Eather <eat...(a)tpg.com.au> wrote: > On 13/03/2010 2:07 PM, cplxphil wrote: > > > > > Hi all, > > > I recently became interested in trying, in my spare time, to develop a > > cryptosystem. I have some math/CS training, and a degree in > > mathematics. What do I need to know to design a cryptosystem? What > > steps are involved? I would assume designing the protocol and then > > modifying it based on how easy it is to attack would be some of the > > steps. > > > Is there a book that details what I need to know/do? > > > Thanks, > > Phil > > You could also try "modern cryptanlysis" by C Swenson. It is not > complete but will give you some ideas on what to avoid. The book could > also do with a good copy editor - but that's a different problem. Hi, my advice is to stay well clear of all existing books. All of these are complexity-theoretic and are very heavy going and eventually take you nowhere. I am an engineer with a great application to mathematics - I suggest reading "The Code Book" by Simon Singh so as to get a feel for what you need and then simply carry on with your normal work and your mathematical intuition alone will alert to some pieces of mathematics that will suggest an algorithm for you to experiment with. Warning - the current state of the crypto scene is a disaster area in my view - it became number theoretic in or about 1960's but went down the wrong road completely and will now have to change - don't waste your time following it but try and understand the reasons why it has failed. I have two websites that will give you an idea of how I taught myself cryptography - http://www.adacrypt.com expounds on Vector Cryptpography and http://www.scalarcryptography uses modular arithmetic - these take a fresh radical approach to cryptography - they have up and running ciphers as claims to unbreakable systems with mathematical proof.- at the moment you are fortunate to be uncontaminated by the status quo which is headed for eventual crashing - take a fresh lead and do it your way - you'll soon see what I mean. Try - "A fresh Approach to Cryptography" on my website http://www.adacrypt.com just to see what I mean about going down the wrong road - welcome to crypto land and good luck with your research - my cryptography is attracting a lot of interest from experienced people which suggest it might be good for you also - Cheers - adacrypt
From: adacrypt on 13 Mar 2010 04:55
On Mar 13, 9:00 am, adacrypt <austin.oby...(a)hotmail.com> wrote: > On Mar 13, 5:25 am, David Eather <eat...(a)tpg.com.au> wrote: > > > > > > > On 13/03/2010 2:07 PM, cplxphil wrote: > > > > Hi all, > > > > I recently became interested in trying, in my spare time, to develop a > > > cryptosystem. I have some math/CS training, and a degree in > > > mathematics. What do I need to know to design a cryptosystem? What > > > steps are involved? I would assume designing the protocol and then > > > modifying it based on how easy it is to attack would be some of the > > > steps. > > > > Is there a book that details what I need to know/do? > > > > Thanks, > > > Phil > > > You could also try "modern cryptanlysis" by C Swenson. It is not > > complete but will give you some ideas on what to avoid. The book could > > also do with a good copy editor - but that's a different problem. > > Hi, my advice is to stay well clear of all existing books. All of > these are complexity-theoretic and are very heavy going and eventually > take you nowhere. I am an engineer with a great application to > mathematics - I suggest reading "The Code Book" by Simon Singh so as > to get a feel for what you need and then simply carry on with your > normal work and your mathematical intuition alone will alert to some > pieces of mathematics that will suggest an algorithm for you to > experiment with. Warning - the current state of the crypto scene is a > disaster area in my view - it became number theoretic in or about > 1960's but went down the wrong road completely and will now have to > change - don't waste your time following it but try and understand the > reasons why it has failed. > > I have two websites that will give you an idea of how I taught myself > cryptography -http://www.adacrypt.comexpounds on Vector > Cryptpography andhttp://www.scalarcryptographyuses modular > arithmetic - these take a fresh radical approach to cryptography - > they have up and running ciphers as claims to unbreakable systems with > mathematical proof.- at the moment you are fortunate to be > uncontaminated by the status quo which is headed for eventual crashing > - take a fresh lead and do it your way - you'll soon see what I mean. > > Try - "A fresh Approach to Cryptography" on my websitehttp://www.adacrypt..com > just to see what I mean about going down the wrong road - welcome to > crypto land and good luck with your research - my cryptography is > attracting a lot of interest from experienced people which suggest it > might be good for you also - Cheers - adacrypt- Hide quoted text - > > - Show quoted text - Correction: the second link should read - http://www.scalarcryptography.co.uk |