From gokaysa@gmail.com Mon Oct 13 10:24:28 2008 Date: Mon, 13 Oct 2008 20:24:28 +0300 From: Gokay Saldamli To: Megan L. Maguire , Cetin Kaya Koc Cc: Alex Troesch , Cody Walker , Jacob Topper , Carl Minden Subject: Fwd: re-intro X-IMAPbase: 1279770016 1 Status: O X-Status: X-Keywords: X-UID: 1   opppsss... gone to Megan only, here are some answers. gokay    On Mon, Oct 13, 2008 at 8:20 PM, Gokay Saldamli wrote:   Let me try to answer your questions 1. f(t) = t^4+t^3+t^2+t+1 2. I havent explicitly wrote them yet but you may extract it from the maple code or give me a day to extract it for you. 3. same as (2)   for the rest of your questions, short answer is I dont have written documents but I can give some basic references for 1 and 6 (mostly attached papers).    For question 4, you have to go to the NIST website, but I will all look for it.   I will give some brief hardware requirement later but I dont have anything written at the moment for software we will see all together  once a working code is finished   we dont have any constant other than alpha and gamma, I think I gave the reasoning for those two already.   Moreover, I include Mr. Cevahir to this conversation, he attacked the algorithm and test its strength (mostly theoretically) against linear and diff. attacks. I believe he will have a few words on questions 3 and 6.      a few words on attachments - FFT hash is proposed by Schorr and broken a few times even after corrections paper 35.pdf and 45.pdf explains the algorithm and its attacks - Lash is another hash algorithm, the paper saarinen_etal_lash.pdf can be very helpful showing how you should write the discussions on security. - diffuni.pdf is also the paper AES s-box refers, describes the bend functions - enhancing merkle damgard.pdf describes enhancement of Merkle-Damgard scheme - check also S. Lucks. Design principles for iterated hash functions. Cryptology ePrint Archive, 2004/253, 2004.   thanks... gokay         On Mon, Oct 13, 2008 at 1:28 AM, Megan L. Maguire wrote: Gokay, As we proceeded to delve into this project, Alex and I realized that we needed some information from you. If you could send us any information on the following ASAP, we would be very appreciative. 1. The explicit form of the irreducible polynomial in GF(2^4) 2. Look up tables for the DFTs or a library or method for calculating them, if available 3. Look up tables for the multiplication and inversion processes if you have them. Also, other information that we will need soon is: 1. Design rational. 2. Information about efficiency and memory requirements on hardware and software. 3. Information regarding effectiveness against all cryptographic attacks. 4. What exactly constitutes the Monte Carlo and Known Answer tests. 5. Justification for any constants used. 6. Any publications relating to the security of the algorithm. Also, are we correct in our conclusion that there are no tunable parameters in the algorithm? Thank you so much for your cooperation. We will contact you as we encounter more questions. Regards, Megan Maguire         Quoting Gokay Saldamli : Hi there, I am glad to see that we have so many enthusiastic students around. As you might witness Spectral hashing involves FFTs, polynomials, finite fields (my familiarity comes from my PhD years) and AES design. All these concepts are extremely interesting to learn and I strongly encourage you to get involve with these abstract subjects at some point of your studies whenever time permits. Surely coming 10-20 days is not the time for this so be hungry and be stupid (as Steve says) and ask as many foolish question as you can so we can have fast pace to finish the necessary tasks before the deadline. In hardware s-hash rules!, its core takes just 8 cycles (where SHA's 80) but when it comes to software I am sorry for the software guys. I think it will noticably be slower that the current SHA but still might open new frontiers. I will send you a readable maple code as soon as possible, and working on some notes describing the mod17 and mod (t^4+t^3+t^2+t+1) arithmetic needed for fields GF(17) and GF(2^4) respectively. cheers... gokay On Sun, Oct 12, 2008 at 7:24 AM, Cetin Kaya Koc wrote: Gokay, Megan, Alex, Jacob, Cody & Carl: Sorry about the emails (too many unnecessary ones) and the address mix up. Everyone's email in this one is correct. I was writing the emails from my iPhone which sometimes is not easy. Megan & Alex are the math group. Jacob, Cody & Carl are the cs group. Gokay will answer our questions for a quick resolution of the issues before submissions. -- ________________________________ Cetin Kaya Koc http://cs.ucsb.edu/~koc Cell: +1 805 403 4191 -- Megan L. Maguire mmaguire@umail.ucsb.edu [ Part 2, Application/OCTET-STREAM (Name: "sHash.mws") 107KB. ] [ Unable to print this part. ] [ Part 3, Application/PDF 404KB. ] [ Unable to print this part. ] [ Part 4, Application/PDF 527KB. ] [ Unable to print this part. ] [ Part 5, Application/PDF 463KB. ] [ Unable to print this part. ] [ Part 6, Application/PDF 356KB. ] [ Unable to print this part. ] [ Part 7, Application/PDF 620KB. ] [ Unable to print this part. ]