Computing the Square Roots of Unity to break RSA using Quantum Algorithms
1 like1,201 views
The document discusses breaking RSA encryption by computing square roots of unity using classical computing methods and Peter Shor’s quantum algorithm. It introduces various algorithms for finding square roots of unity in finite groups, their mathematical background, and potential applications in cryptography. Additionally, it includes a demonstration of these algorithms and their effectiveness in factoring composite numbers.
![39
import java.math.BigInteger;
import java.security.SecureRandom;
public class SqrtOneNaive implements IBigIntConstants
{
public static void main(String[] args)
{
/* Error handling is omitted */
BigInteger n = new BigInteger(args[0]);
BigInteger x = zero;
while(x.compareTo(n) <= 0) {
// is square(x) = 1 (mod n)?
if(x.multiply(x).mod(n).equals(one)) {
System.out.println(x + " is the square root of one ");
}
x = x.add(one);
}
}
}
Implementation of naïve Sqrt(1) algorithm](https://image.slidesharecdn.com/computingthesquarerootsofunitytobreakrsa-190330042902/85/Computing-the-Square-Roots-of-Unity-to-break-RSA-using-Quantum-Algorithms-39-320.jpg)
![40
public static void main(String[] args)
{
BigInteger n = new BigInteger(args[0]); // Error handling code is removed
BigInteger x;
while(true) {
// Pick a random element from Zn
x = BigIntUtil.getRandomElt(n, new SecureRandom());
// We don't want non-trivial roots '1' or '-1'.
// If x = -1 (mod n) or x = 1 (mod n), then we have to restart.
if(x.add(one).mod(n).equals(zero) ||
x.subtract(one).mod(n).equals(zero)) { continue; }
if(x.multiply(x).mod(n).equals(one)) {
System.out.println(x + " is the square root of one ");
System.out.println("Number of random tries = " + counter);
break;
}
}
}
Implementation of random Sqrt(1) algorithm](https://image.slidesharecdn.com/computingthesquarerootsofunitytobreakrsa-190330042902/85/Computing-the-Square-Roots-of-Unity-to-break-RSA-using-Quantum-Algorithms-40-320.jpg)
Computing the Square Roots of Unity to break RSA using Quantum Algorithms
- 1. Computing the Square Roots of Unity to break RSA using Quantum Algo Dr. Dharma Ganesan, Ph.D.,
- 2. Disclaimer ● The opinions expressed here are my own ○ But not the views of my employer ● The source code fragments and exploits shown here can be reused ○ But without any warranty nor accept any responsibility for failures ● Do not apply the exploit discussed here on other systems ○ Without obtaining authorization from owners 2
- 3. Goal ● Break RSA using Peter Shor’s quantum algorithm ○ But, we will not look into Quantum Computing ○ We will find the square roots of unity on a classical computer ● Investigate the square root of unity (i.e., sqrt(1)) in a Finite Group ○ Algorithm 1: Naïve/Bruteforce ○ Algorithm 2: Random Search ○ Algorithm 3: Peter Shor’s period calculation algorithm ○ Algorithm 4: Our little tweaks of Peter Shor’s period function (bounded search) 3
- 4. Prerequisite Some familiarity with the following topics will help to follow the rest of the slides ● Group Theory ● Number Theory ● Algorithms and Complexity Theory ● If not, it should still be possible to obtain a high-level overview 4
- 5. Agenda ● Mathematical foundation/facts ● Brief overview of RSA ● Different algorithms to find the square roots of unity (in a finite group) ● Demos of finding the square roots of unity to factor composite numbers ● Conclusion 5
- 6. 6 Notations and Facts GCD(x, y): The greatest common divisor that divides integers x and y Co-prime: If gcd(x, y) = 1, then x and y are co-primes Zn = { 0, 1, 2, …, n-1 }, n > 0; we may imagine Zn as a circular wall clock Z* n = { x ∈ Zn | gcd(x, n) = 1 }; (additional info: Z* n is a multiplicative group) φ(n): Euler’s Totient function denotes the number of elements in Z* n φ(nm) = φ(n).φ(m) (This property is called multiplicative) φ(p) = p-1, if p is a prime number
- 7. Notations and Facts ... ● x ≡ y (mod n) denotes that n divides x-y; x is congruent to y mod n ● Euler’s Theorem: aφ(n) ≡ 1 (mod n), if gcd(a, n) = 1 ● Fermat’s Little Theorem: ap ≡ a (mod p) ● Gauss’s Fundamental Theorem of Arithmetic: Any integer greater than 1 is either a prime or can be written as a unique product of primes ○ Euclid’s work is the foundation for this theorem, see The Elements ● Euclid’s Lemma: if a prime p divides the product of two natural numbers a and b, then p divides a or p divides b ● Euclid’s Infinitude of Primes (c. 300 BC): There are infinitely many primes 7
- 8. How can Bob send a message to Alice securely? 8 Public Key PuA ● Alice and Bob never met each other ● Bob will encrypt using Alice’s public key ○ Assume that public keys are known to the world ● Alice will decrypt using her private key ○ Private keys are secrets (never sent out) ● Bob can sign messages using his private key ○ Alice verifies message integrity using Bob’s public key ● Note: Alice and Bob need other evidence (e.g., passwords, certificates) to prove their identity to each other ● Who are Alice, Bob, and Eve? Private Key PrA Public Key PuB Private Key PrB
- 9. RSA Public Key Cryptography System ● Published in 1977 by Ron Rivest, Adi Shamir and Leonard Adleman ● Rooted in elegant mathematics - Group Theory and Number Theory ● Core idea: Anyone can encrypt a message using recipient's public key but ○ (as far as we know) no one can efficiently decrypt unless they got the matching private key ● Encryption and Decryption are inverse operations (math details later) ○ Work of Euclid, Euler, and Fermat provide the mathematical foundation of RSA ● Eavesdropper Eve cannot easily derive the secret (math details later) ○ Unless she solves “hard” number theory problems that are computationally intractable 9
- 10. RSA - Key Generation Algo 1. Select an appropriate bitlength of the RSA modulus n (e.g., 2048 bits) ○ Value of the parameter n is not chosen until step 3; small n is dangerous (details later) 2. Pick two independent, large random primes, p and q, of half of n’s bitlength ○ In practice, q < p < 2q to avoid attacks (e.g., Fermat’s factorization) 3. Compute n = p.q (n is also called the RSA modulus) 4. Compute Euler’s Totient (phi) Function φ(n) = φ(p.q) = φ(p)φ(q) = (p-1)(q-1) 5. Select numbers e and d from Zn such that e.d ≡ 1(mod φ(n)) ○ e must be relatively prime to φ(n) otherwise d cannot exist (i.e., we cannot decrypt) ○ d is the multiplicative inverse of e in Zn 6. Public key is the pair <n, e> and private key is 4-tuple <φ(n), d, p, q> 10
- 11. RSA Trapdoor ● RSA: Zn → Zn ● Let x and y ∈ Zn ● y = RSA(x) = xe mod n ○ We may view x as a plaintext, and y as the corresponding ciphertext ● x = RSA-1 (y) = yd mod n ● e and d are also called encryption and decryption exponents, respectively ● In practice, some random padding is used (not relevant for this presentation) 11
- 12. RSA Trapdoor variables’ dependency graph 12 Private variable Public variable Note: If we can factor n and find p and q, then private keys are open to the world
- 13. What are the square roots of unity in Z? ● 1 and -1 are the two square roots of one because (±1)2 = 1 ● This is true only if we work in the infinite set of integers Z ● In Cryptography, we often work with finite groups ● There are other non-trivial roots of unity other than ±1 in Z* n 13
- 14. Square roots of unity in a finite group Z* n Consider this tiny multiplicative group Z*15 = {1, 2, 4, 7, 8, 11, 13, 14} The square roots of unity are 1, 4, 11, and 14 Note that 14 ≡ -1 mod 15 (basically -1 is the same as 14) But, we have two other roots 4 and 11 of unity (e.g., 4*4 = 1 mod 15) 1 and 14 are trivial roots of unity, whereas 4 and 11 are non-trivial roots of unity Fact: Let n = pq, where p and q are odd primes, then there are four roots of unity 14
- 15. Algorithm: Factor n using the square roots of unity Problem: Given a composite number n and sqrt(1), find the factors p and q of n Let x be a square root of 1, then x2 = 1 (mod n) Thus, x2 - 1 = 0 (mod n) ⇒ (x+1)(x-1) = 0 (mod n) This means that n divides (x+1)(x-1) If (x+1) and (x-1) are not multiples of n, then the factors of n are the following: p = gcd(x+1, n) or gcd(x-1, n); q = n/p 15
- 16. Toy example: Let’s factor n = 15 using Z*15 Z*15 = {1, 2, 4, 7, 8, 11, 13, 14} Since 4 and 11 are square roots of unity, we can use one of them to factor n Let x = 4 x2 = 1 (mod 15) (x+1)(x-1) = 0 (mod 15) Factor p = gcd(x+1, 15) = gcd(5, 15) = 5; Factor q = n/p = 15/5 = 3 16
- 17. Factoring n using the square roots of unity ... If we can find the non-trivial square roots on n, then we can factor n Let’s investigate how easy it is to find the sqrt(1) in a finite multiplicative group Four different algorithms to find the sqrt(1): 1. Bruteforce/Naive 2. Random Search 3. Peter Shor’s Quantum Algorithm 4. Our little tweaks of Shor’s algorithm 17
- 18. Algorithm 1: Naïve sqrt(1) in Zn * Step 1: x = 0 Step 2: if(x < n) goto step 3; otherwise goto step 5 Step 3: if(x2 = 1 mod n) then x is a square root of one. Step 4: x = x + 1. goto step 2. Step 5: End. 18
- 19. 19 ~/crypto/RSA$ java SqrtOneNaive 2086784737 Input bit length = 31 Sqrt(n) = 45681 sqrt(1) = 1 bit length = 1 sqrt(1) = 530886943 bit length = 29 sqrt(1) = 1555897794 bit length = 31 sqrt(1) = 2086784736 bit length = 31 Note: sqrt(n) is only 45681, but sqrt(1) can be the same bit length as n (See the Appendix for the source code of Sqrt(1) naïve method) Algorithm 1: Find the sqrt(1) in the group Z*2086784737
- 20. Algorithm 1: Analysis of the naïve algorithm ● The naïve algorithm is simple to understand and implement ● However, it is very slow because we enumerate all numbers of the group Zn ● If n is a 2048-bit composite number, this algorithm will perform 22048 iterations ● In other words, it will take several billion years to calculate sqrt(1) 20
- 21. Algorithm 2: Random search to find sqrt(1) in Zn * Step 1: Pick a random x from Zn Step 2: if(x = ± 1 mod n) then goto step 1 Step 3: if(x2 = 1 mod n) then x is a sqrt(1) goto step 4; else step 1 Step 4: End. 21
- 22. 22 dharmalingam_ganesan11@instance-1:~/crypto/RSA$ java RandomSqrtOne 45649 sqrt(1) = 43738 Number of random tries = 236 ddharmalingam_ganesan11@instance-1:~/crypto/RSA$ java RandomSqrtOne 45649 sqrt(1) = 43738 Number of random tries = 82005 dharmalingam_ganesan11@instance-1:~/crypto/RSA$ java RandomSqrtOne 45649 sqrt(1) = 43738 Number of random tries = 167572 Algorithm 2: Find the sqrt(1) in the group Z*45649
- 23. 23 Algorithm 2: Analysis of Random Search Algo ● The random algorithm can find the square roots of one (for “small” n) ● But, it just needs a lots of tries to find a needle in the haystack ● In general, the number of tries is the size of the input composite number ● Random search is not able to find sqrt(1) for very large input n
- 24. 24
- 25. 25 Algorithm 3: Peter Shor’s Algorithm (core idea) ● To find an x such that x2 = 1 (mod n), we compute f(x) = ax mod n, random a ● Fact: Every element of a finite group generates a cyclic subgroup ● In other words, the function f(x) is the same as the cyclic subgroup <a> ● Thus, f(r) = 1 mod n for some r; equivalently, ar = 1 mod n for some r ● Once we find such an r, (ar/2 )2 = 1 mod n; (if r is even, the hacker is lucky) ● Thus, ar/2 is the sqrt(1); we know how to factor n once the sqrt(1) is exposed!
- 26. Algorithm 3: Finding the period r of the function f(x) 26 x f(x) = ax mod n 0 a0 mod n 1 a1 mod n 2 a2 mod n r ar mod n = 1 ● Peter Shor uses quantum computing to find the period r of the function f(x) ● We implemented this function on my classical computer (my laptop) ● Just to better understand the inner beauty of this algorithm
- 27. 27 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 1330 sqrt(1) = 1911 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 4522 sqrt(1) = 43738 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 22610 sqrt(1) = 1911 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 22610 sqrt(1) = 1911 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 4522 sqrt(1) = 1911 ~/crypto/RSA$ java -ea SqrtOneShor 45649 period = 22610 sqrt(1) = 43738
- 28. 28 ~/crypto/RSA$ time java -ea SqrtOneShor 1421727721 period = 236941974 Sqrt(1) = 521638680 real 0m47.429s user 0m47.100s sys 0m0.288s Period is very large
- 29. 29 ~/crypto/RSA$ java FactorShor 45649 period = 22610 p = 191 q = 239 ~/crypto/RSA$ java FactorShor 45649 period = 170 p = 239 q = 191 ~/crypto/RSA$ java FactorShor 45649 period = 3230 p = 239 q = 191 Algorithm 3: Factor a small number n = 45649
- 30. 30 Algorithm 3: Analysis of the period of f(x) = ax mod n ● The above set of runs shows that the period of f(x) can be very large ○ Thus, we cannot apply this algorithm to find sqrt(1) for a large group Zn used in RSA ● Remember, we are running the algorithm on a classical computer ● It is good to play with this algorithm on a classical computer to understand it ● Algorithm 4 talks about some little tweaks of Algorithm 3
- 31. Algorithm 4: Let’s tweak Peter Shor’s algorithm What if we put a bound b on the period r? Step 1: Initialize bound b (e.g., b = n1/2 , n1/4 , etc.), r = 0, success = false Step 2: Pick a random a from Zn * Step 3: if 0 ≤ r ≤ b goto step 4; otherwise goto step 7 Step 4: Compute f(r) = ar mod n Step 5: If f(r) != 1 then r = r+1 and goto step 3 Step 6: If r is even, then success = true; otherwise goto step 2 Step 7: If success = true, we found the period r; otherwise goto step 2 31
- 32. 32 ~/crypto/RSA$ time java -ea SqrtOneShor 1421727721 period = 33738 sqrt(1) = 521638680 real 0m45.786s user 0m45.510s sys 0m0.244s Algorithm 4: bound = n1/2
- 33. 33 ~/crypto/RSA$ time java -ea SqrtOneShor 1421727721 period = 18 Square root of one = 521638680 real 6m14.713s user 6m12.706s sys 0m1.759s Algorithm 4: bound = n1/4
- 34. 34 ~/crypto/RSA$ time java -ea SqrtOneShor 1421727721 period = 18 Square root of one = 521638680 real 2m29.435s user 2m28.400s sys 0m0.936s Algorithm 4: bound = n1/6
- 35. 35 ~/crypto/RSA$ time java -ea SqrtOneShor 1421727721 period = 6 Square root of one = 900089041 real 2m31.106s user 2m29.703s sys 0m1.304s Algorithm 4: bound = n1/8 Amazing, we found a cyclic subgroup of order 6
- 36. Algorithm 4: Analysis of our tweaks ● Z* n has even number of elements; there must be cyclic subgroups for each divisor of |Zn *|. (This is a theorem in Group Theory) ● The open question is how easy it is to find small cyclic subgroups of Z*n ● If we quickly find one cyclic subgroup, then sqrt(1) are easily revealed ● If we put a bound b on the period r, then sqrt(1) runs quick for a small n ● Our bounded random search variant of Shor’s algorithm is slow for a large n 36
- 37. Conclusion ● We studied the problem of finding the square roots of unity in a finite group ● We tried to factor composite numbers using sqrt(1) ● Experimental results showed that finding sqrt(1) in Z*n where n = pq is “hard” ● We implemented Shor’s function f(x) = ax mod n on a classical computer ● We learned that in general the period of f(x) is the same size as n ● We made minor tweaks by defining a bound on the period of f(x) = ax mod n ● More experimentation of our tweaks is part of future work 37
- 38. Appendix 38
- 39. 39 import java.math.BigInteger; import java.security.SecureRandom; public class SqrtOneNaive implements IBigIntConstants { public static void main(String[] args) { /* Error handling is omitted */ BigInteger n = new BigInteger(args[0]); BigInteger x = zero; while(x.compareTo(n) <= 0) { // is square(x) = 1 (mod n)? if(x.multiply(x).mod(n).equals(one)) { System.out.println(x + " is the square root of one "); } x = x.add(one); } } } Implementation of naïve Sqrt(1) algorithm
- 40. 40 public static void main(String[] args) { BigInteger n = new BigInteger(args[0]); // Error handling code is removed BigInteger x; while(true) { // Pick a random element from Zn x = BigIntUtil.getRandomElt(n, new SecureRandom()); // We don't want non-trivial roots '1' or '-1'. // If x = -1 (mod n) or x = 1 (mod n), then we have to restart. if(x.add(one).mod(n).equals(zero) || x.subtract(one).mod(n).equals(zero)) { continue; } if(x.multiply(x).mod(n).equals(one)) { System.out.println(x + " is the square root of one "); System.out.println("Number of random tries = " + counter); break; } } } Implementation of random Sqrt(1) algorithm
- 41. 41 Finding the Sqrt(1) using Peter Shor’s algorithm (simplied)
- 42. References ● W. Diffie and M. E. Hellman, “New Directions in Cryptography,” IEEE Transactions on Information Theory, vol. IT-22, no. 6, November, 1976 ● R. L. Rivest, A. Shamir, and L. Adleman, “A method for obtaining digital signatures and public-key cryptosystems,” CACM 21, 2, February, 1978 ● A. Menezes, P. van Oorschot, and S. Vanstone, “Handbook of Applied Cryptography,” CRC Press, 1996 ● P. Shor, “Polynomial-Time Algorithms for Prime Factorization and Discrete Logarithms on a Quantum Computer,” SIAM J.Sci.Statist.Comput. 26 (1997). 42