OK. Een nieuw projekt, maar waar ben je mee bezig?

Pagina: 1
Acties:
  • 109 views sinds 30-01-2008
  • Reageer

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
OK. Er is een nieuw projekt, iedereen heeft het erover.
Maar ik zou graag weten wat er nou wordt uitgerekend. Ik wil wel eens weten waar mijn pc voor rammelt B-)

  • Sequence
  • Registratie: Maart 2000
  • Laatst online: 27-05-2024

Sequence

Online marketing

zo staat het op de site:
http://www.nd.edu/~cmonico/eccp109/
<HR>What are we doing?
<HR>

We are coordinating a distributed effort to solve Certicom's ECCp-109 challenge. The challenge is to solve a particular elliptic curve discrete logarithm problem.
The Elliptic Curve Discrete Logarithm Problem (ECDLP) is the basis for a powerful cryptosystem. The very rudimentary idea of our particular problem is the following:
We have a curve, C, of the form:

y2 = x3 +ax +b

For some constants 'a' and 'b'. We're also given some fixed large prime 'p'. A "point on the curve" is a pair of integers (u,v) that satisfy this equation modulo 'p'. This means, simply, that 'p' divides v2-u3-au-b. Now the leap of faith: There is a method by which we can "add" two points on the curve to get another point on the curve. We call it addition, but it looks nothing like what we normally think of as addition. Just think of it as a rule that tells us how to obtain a third point on the curve from two given points. If you know a little algebra, I'll tell you that these points together with this operation form an abelian group.
Here's the point of the challenge: Certicom has chosen a point, 'P', on this curve and a very big integer 'k'. They then computed Q := kP. This means that they added 'P' to itself 'k' times and called the result 'Q'. But there is a clever way to do this with only log(k) operations. So they can choose a REALLY big 'k', and still compute kP. Now, we know 'P' and 'Q' and our mission is to find 'k'. But 'k' is far too big to simply start trying k=1,k=2,k=3,... so we need to do it in a smarter way. Here are the exact challenge parameters for ECCp-109:
p =564538252084441556247016902735257
y2 = x3 +321094768129147601892514872825668x +430782315140218274262276694323197
P = (97339010987059066523156133908935, 149670372846169285760682371978898)
Q = (44646769697405861057630861884284, 522968098895785888047540374779097)

What does this have to do with cryptography? One of the major problems of cryptography is to achieve secure communication over an insecure channel. Think of it in the following way: You're about to buy something on the internet, from company X. To do this, you need to give them your credit card number. But the internet is an inherintly insecure channel. That is, there are eavesdroppers everywhere just watching all your communications with company X. Let's call one of the eavesdroppers Eve. So how will you give them your credit card without Eve getting it as well? Of course, the answer is by encrypting it. But a little thought quickly convinces you that it's not that simple. You and company X have to agree on some encryption which both of you can 'do' and 'undo'. But Eve is listening while you try to agree on this. So here's my point: you and company X need to agree on some secret information that you will both know but Eve will not. And elliptic curves provide one way of doing this. Here's how it works (this is Diffie-Hellman key exchange with elliptic curves):
You and X agree on some elliptic curve and some point 'P' on the curve. X secretly chooses some large integer 's' and computes the point A := sP. That is, A is 'P' added to itself 's' times. X then tells you the point 'A' but not the integer 's'. You do something similar, and secretly pick a large integer 't'. You then add 'P' to itself 't' times to get the point B := tP. You tell X what the point 'B' is, but not the integer 't'.
Now you know the point A and your secret integer 't'. You add A to itself 't' times, and obtain Z := tA. Company X does something similar and adds your point B to itself 's' times to get Z := sB. I have not accidentally used the same 'Z' twice - these two calculations actually give the same point, since

Z = tA = t(sP)=(ts)P = (st)P = s(tP) = sB
This is your secret information that the eavesdropper will not know. It's a small technical detail of how you use this secret information to make an encryption scheme. But if you think about it, it's not that far of a leap. Once you two know something that nobody else knows, you can find a way to do encryption with it.
Why doesn't the eavesdropper know it? Let's look carefully at what Eve sees you and X exchange. She sees the elliptic curve you agree on and the point 'P'. She sees the points that you exchange: 'A' and 'B'. But this is all that she knows: She knows neither 's' nor 't'. How could she try to figure out the secret piece of information 'Z'? Certainly she can add A and B, but this doesn't help since:

A+B = sP + tP = (s+t)P
and this is not the same point as Z=(st)P. If she could figure out one of the integers 's' or 't', she could get Z. But how does she figure out one of these integers? This is exactly the elliptic curve discrete log problem: She knows, for example, P and A and must find an integer 's' so that P = sA. So, she must solve the same problem that we're trying to solve in this challenge. This is also part of the reason for this series of challenges that Certicom has proposed: To see exactly what Eve can actually solve. Since we're going to solve this challenge, you would actually use larger parameters so that Eve could not solve it. Specifically, the problem gets harder as the prime 'p' gets larger. Our prime 'p' is a 109 bit integer for this challenge, so you and X would use something larger than 109 bits (maybe 200 bits). I should mention that we are going to _barely_ solve this problem. That is, if the prime were 131 bits, our attack would not work. It would take about 65536 times longer (or need 65536 times as many machines to do it in the same amount of time).

Verwijderd

ps: waarom heb jij eigenlijk je eigen team?

  • Goingin
  • Registratie: Januari 2000
  • Laatst online: 08-02-2023
Foutje denk ik

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Mijn eigen team is zekers geen foutje :9
Maar om ff op mijn vraag terug te komen, van dat engelse gebrabbel met al die moeilijke codes enzo snap niemand er toch geen hol van? |:(

Verwijderd

Op maandag 11 juni 2001 20:43 schreef My het volgende:
Mijn eigen team is zekers geen foutje :9
Maar om ff op mijn vraag terug te komen, van dat engelse gebrabbel met al die moeilijke codes enzo snap niemand er toch geen hol van? |:(
vaagggggggggg :? Maar wat jij wil.... Puzzelze!

Denk dat dat wel meevalt.

  • Marcj
  • Registratie: November 2000
  • Laatst online: 22-05 16:38
My, ik heb dit ook al ergens anders gepost, maar ben je nou lid van PNpX of niet :? (en zo ja, waarom je eigen team starten?)

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Marcj, ik wil de weg bij PNpX, door al dat gerotzooi |:(, tja nu is het weer goed maar ik blijf nog maar eventjes weg >:)

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Ow... en btw ik hep nog steeds geen antwoordt op mijn vraag (behalve dat moeilijk wiskundig gebrabbel in het engels) PLZ HELP ME :9

Verwijderd

Je lost een geheimtaal op.

Sim-pel ...

  • DeX
  • Registratie: Juni 2001
  • Laatst online: 13-01 09:52

DeX

Sorry hoor maar ik snap nog steeds het nut niet van RC-5. Weten jullie ook wat je zit te ontcijferen? Of zitten jullie alleen maar als een stel schapen achter de Koe aan? ena als het hele ding oncijfert is... wat staat er dan???

  • NightHawk4
  • Registratie: September 2000
  • Laatst online: 22-05 19:11

NightHawk4

Lotte is Liev :>

Op dinsdag 12 juni 2001 15:31 schreef DeX het volgende:
Sorry hoor maar ik snap nog steeds het nut niet van RC-5. Weten jullie ook wat je zit te ontcijferen? Of zitten jullie alleen maar als een stel schapen achter de Koe aan? ena als het hele ding oncijfert is... wat staat er dan???
Dan staat er dat we met z'n allen voor een tonnetje of 2 aan stroom hebben weggegooit :?

  • DeX
  • Registratie: Juni 2001
  • Laatst online: 13-01 09:52

DeX

Op dinsdag 12 juni 2001 15:33 schreef NightHawk het volgende:

[..]

Dan staat er dat we met z'n allen voor een tonnetje of 2 aan stroom hebben weggegooit :?
Inderdaad...

Mijn idee:

Jullie werken allemaal zonder het te weten voor de CIA. jullie zitten pakketjes uit communistische landen te ontcijferen. zodat de CIA erachter kan komen waar kinderporno en warez vandaan komen. :) . Daarnaast is het ook leuk om te weten hoeveel Nukes die lui hebben. RC-5 kraken duurt veelste lang voor hun dus laten ze het jullie schaapjes maar doen. Hoe zit het met DES? wie gaat dat ontciujferen?

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
LOL :7

Ik zit dus geheimtaal te ontcijferen (zoals wordt gezegt) maar wat hebben die wiskundige formules daar mee te doen???

Verwijderd

Die wiskunde is de manier om ervoor te zorgen dat de verzender en de ontvanger elk een deel van een sleutel hebben, en aan de hand van dat deel de rest van de oplossing ter ontcijfering (complete sleutel) kunnen berekenen. Iemand die de transmissie tussen zender en ontvanger onderschept mist beide delen, en een vergelijking met 2 onbekenden is in principe niet oplosbaar. De enige methode om dit te doen is (theoretisch) voor een onbekende -oneindig tot +oneindig in te vullen, min of meer wat je met dit project en RC5-64 doet.
De wiskunde in het verhaal legt uit hoe het mogelijk is dat je jouw deel van de sleutel completeert zonder het andere deel te kennen.

Verwijderd

Op dinsdag 12 juni 2001 15:37 schreef DeX het volgende:

[..]

Inderdaad...

Mijn idee:

Jullie werken allemaal zonder het te weten voor de CIA. jullie zitten pakketjes uit communistische landen te ontcijferen. zodat de CIA erachter kan komen waar kinderporno en warez vandaan komen. :) . Daarnaast is het ook leuk om te weten hoeveel Nukes die lui hebben. RC-5 kraken duurt veelste lang voor hun dus laten ze het jullie schaapjes maar doen. Hoe zit het met DES? wie gaat dat ontcijferen?
Da's niet alleen jouw idee, UD (naar ik meen) heeft iets dergelijks in de gebruikersovereenkomst op moeten nemen dat er geen hidden feats draaien. Je kunt weer rustig :z :Z ;) >:)

DES-III is allang opgelost ?!?

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Ok thnx Nemesis zo weet ik wat ik doe, en waar m'n pc voor staat te rammelen 8-)

  • Marcj
  • Registratie: November 2000
  • Laatst online: 22-05 16:38
Je weet dat alles wat je nu voor jezelf doet, later niet naar PNpX kan geven :)

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Ja dat weet ik... MAAR...
Ik heb toegezegt volgend project van eccp WEL bij dpc te komen, dus via het subteam pnpx. 8-)

  • Marcj
  • Registratie: November 2000
  • Laatst online: 22-05 16:38
Op woensdag 13 juni 2001 13:28 schreef My het volgende:
Ja dat weet ik... MAAR...
Ik heb toegezegt volgend project van eccp WEL bij dpc te komen, dus via het subteam pnpx. 8-)
Je zit nu ook niet bij dpc??? Wat doe je hier dan :? ;)

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Volgens mij hoef je nie perce lid van dpc te zijn om hier berichies te posten :?

  • Crack
  • Registratie: Februari 2000
  • Laatst online: 30-04 09:17

Crack

...ehh.......Mooh?!?

Op dinsdag 12 juni 2001 19:36 schreef My het volgende:
LOL :7

Ik zit dus geheimtaal te ontcijferen (zoals wordt gezegt) maar wat hebben die wiskundige formules daar mee te doen???
What are the advantages of ECC?
The big win with ECC, as compared to other public-key algorithms, is key size.
[knip][/knip]
ECC also has less computational overhead than RSA, primarily because it does not have to analyze prime numbers, a fairly expensive operation.

Together, these facts allow for a more efficient cryptosystem. ECC devices will require less storage, less power, less memory, and less bandwidth. This allows you to implement cryptography in platforms that are constrained, such as wireless devices, handheld computers, smart cards, and thin-clients. It also provides a big win in situations where efficiency is extremely important, such as on a bottlenecked web server supporting e-commerce.
(Volledige FAQ)

Wereldrecord voor ChicaneLinacB90 Badges


  • Goingin
  • Registratie: Januari 2000
  • Laatst online: 08-02-2023
Op woensdag 13 juni 2001 14:10 schreef My het volgende:
Volgens mij hoef je nie perce lid van dpc te zijn om hier berichies te posten :?
Waarom wil je zowieso je eigen internationale team??

Als jij met je team nou ff DPC joint, dan hebben we ook nog wat aan je, en zelf heb je ook wat eer aan je werk, want met je eigen team hoog komen is moeilijk :)

  • My
  • Registratie: Mei 2000
  • Laatst online: 06-01 21:17
Ik vond het eens leuk om met mezelf en gasten een leuk teampje te maken...
Ik was eruit gekicked bij pnpx dus heb ik gasten geronseld van een andere server :7
Pagina: 1