Backtracking based integer factorisation, primality testing and square root calculation

BACKTRACKING BASED INTEGER
FACTORISATION, PRIMALITY TESTING
AND SQUARE ROOT CALCULATION
Mohammed Golam Kaosar
School of Computing and Mathematics
Charles Sturt University Australia
mkaosar@csu.edu.au

ABSTRACT
Breaking a big integer into two factors is a famous problem in the field of Mathematics and
Cryptography for years. Many crypto-systems use such a big number as their key or part of a
key with the assumption - it is too big that the fastest factorisation algorithms running on the
fastest computers would take impractically long period of time to factorise. Hence, many efforts
have been provided to break those crypto-systems by finding two factors of an integer for
decades. In this paper, a new factorisation technique is proposed which is based on the concept
of backtracking. Binary bit by bit operations are performed to find two factors of a given
integer. This proposed solution can be applied in computing square root, primality test, finding
prime factors of integer numbers etc. If the proposed solution is proven to be efficient enough, it
may break the security of many crypto-systems. Implementation and performance comparison of
the technique is kept for future research.

KEYWORDS
Information Security, Crypto-system, Factorization, Primality test, Backtracking.

1. INTRODUCTION
Integer factorisation is known as the decomposition of a composite integer number into small
divisors. As for example, 91 is a composite integer which is a composition of 7 and 13, i.e. 91 = 7
x 13. Both 7 and 13 are known as the factors of 91. If the factor is a prime number, then it is
known as prime factor. In the above example, both of them are prime factors. As the size of the
number increases, it becomes very difficult to find its factors. Sometimes, a sophisticated
algorithm running in the fastest computer may take hundreds of years to find a factor of a large
number. As a matter of fact, many cryptographic algorithms, such as RSA [11], use a big number
(1024 or 2048 bits) in generating keys with the assumption that, fastest technique with the help of
many computers would not be able to factorise that number within a practically feasible time.
This paper proposes a new technique to decompose a composite integer into two factors using
backtracking technique. Repetitive application of the proposed technique will find all possible
David C. Wyld et al. (Eds) : CCSIT, SIPP, AISC, PDCTA, NLP - 2014
pp. 77–83, 2014. © CS & IT-CSCP 2014

DOI : 10.5121/csit.2014.4207

78

Computer Science & Information Technology (CS & IT)

factors of a given integer number. The proposed solution also can be used in testing the primality
and finding the square root of an integer (if there exist any).
The impact of the proposed solution can be tremendous, depending upon its performance. If it can
factorise a big number (as big as the number used in some crypto-systems) within a practically
feasible amount of time, it would make a big change in the field of Cryptography. Further impact,
implementation and performance analyses are kept for future research.
The rest of the paper is organised as follows: Section 2 presents existing solutions for factorising
integer numbers in brief. Section 3 discusses the proposed solution with an example and Section 4
concludes the paper with some future research directions.

2. EXISTING SOLUTIONS
There have been many efforts proposed to factorize an integer number. The use of prime factors
in crypto-systems increased much research interest to finding a practical solution to factorise a
big integer. Integer factorization intrinsically leads towards the solution of primality testing and
finding square root of an integer number. Some of the existing solutions are as follows:
•

Trial Division: Trial division algorithm finds whether a given integer N is divisible by
any positive number less than N. This is a simple and brute force approach which is very
time consuming to find a solution.

•

Wheel Factorisation: This is a graphical method of factorising an integer. In this method,
natural numbers are marked around the wheel to form spokes of primes and their
multiples.

•

ρ1 = 1 and
Rho Methods: Rho ( ρ ) method generates ρ1, ρ2, ρ3… where
parameter 10 is chosen by users. This method finds factors of N by computing gcd(N,
(ρ2- ρ1) (ρ4 – ρ2)….( ρ2n – ρn )). Parameter n is also chosen by user. This method initially
was proposed by Pollard [8]. Some variations, improvements and optimisations of Rho
methods are proposed in – [9], [2], [1] etc.

•

Fermat's and Euler's Factorisation: Fermat's Factorisation technique [7] represents odd
integer N as a2 – b2 where both a and b are positive integer. Thus, N is equal to a2 – b2 =
( a + b ) ( a – b ) and factors are ( a + b ) and ( a – b ). In Euler's Factorisation technique
[6], N is represented as a2 + b2 and c2 + d2, where a, b, c and d are positive integers.

•

Other Methods: Various other methods of factorisation algorithms are: group
factorisation technique of Pollard's ρ – 1 [8], William’s ρ + 1 factorisation [12], Lenstra’s
elliptic curve factorization [5]; Dixon’s factorization [3]; Quadratic sieve factorization
[10] etc.

3. PROPOSED SOLUTION
This section presents the proposed solution to factorize an integer with a simple example. The
solution is also extended for primality testing and square root calculation.


3.1 Factorising an Integer

Figure 1: Binary standard multiplication of A and B to produce N

Shifted Multiplicands Matrix (SMM), as shown in Figure 2, has the dimension of
with entry as follows:

Figure 2: SMM block diagram

79

80


Each bit in N is a binary summation of bits of SMM's corresponding positions, as displayed in
Equation 2. Where, Sk and Ck are the sum and carry bit of SMM entries at column k.

Further detail of binary multiplication method can be found in [4].
Now an algorithm is to be developed which would choose ai and bi in such a way that, nk
calculated in Equation 2 becomes equal to rk, for all k ≤ ( 2l – 1). Following example explains
how a number is factorised using the proposed technique.
3.1.1 An Example
Say an integer R = 12 and hence, m = 4. We have to find two factors, A and B, or R. Binary
representation of R, A, B and S M M, as discussed in Section 3.1, are now showed in step 1 of
Fig. 3, where fields of A and B are empty and entries of S M M are set to 0 initially.
We need to consider every bit of R from right to left. In Step 2 we try to choose values for a1, b1
and S M M [1] [1] such that n1 calculated using Equation 2 becomes equal to r1. Therefore,
S M M [1] [1] must be 0. Possible values of { a1, b1 } could be { 0,0 } or { 1,0} or { 0,1}, but
cannot be {1,1}. Let us say we choose { 0, 0}. Now immediately we can set S M M [2] [2] =
S M M [3] [3] = 0, since a1 would make these values to 0 regardless of the value of bi, i = 1 to l.
Similarly S M M [1] [2] = S M M [1] [3] = 0, since b1 would make these values to 0 regardless of
the value of ai, i = 1 to l.
Now, we move to the next bit of R in step 3. Values of SMM in column 2 are already consistent.
We have the option to choose any value for {a2, b2}. Say us choose {1, 1}. Therefore, S M M
[2][3] can be set to 1. Similarly we consider for r3 in Step 4. Only one position of S M M in this
column is left which has to be 0 to be consistent with r3. Choose any value for {a3, b3}. Let us
say we choose {0,0}. Hence S M M [2][4] and S M M [2][5] have to be 0. But in the next column
(column 4), all values become 0. Therefore, a conflict arises, since the summation could not
produce a value equal to r4. Hence, a backtracking will be necessary for r3. Now, let us try with
the values of {a3, b3} an {1,1} in Step 5. This time it creates conflict too. In next step, let us try
with {a3, b3} - {1,0} in step 6. Now the values of A, B and S M M becomes consistent with that
of R. Therefore, two factors of R would be A = 6 and B = 2.


Figure 3: Steps of factorisation for the example

3.1.2 The Algorithm
The proposed solution appears in the following algorithm in a concised manner:

81

82


3.2 Primality Test and Square Root
The proposed factorisation technique can easily be extended for the following solutions:
Square Root
If we set or require, A and B must be equal, then we get the square root of R (if it is a perfect
square) in A and B. In this case, the computation even would be quicker, since more values of
SMM would be pre-set.
Primality Test
If the proposed algorithm finds no solution after finishing all steps then, (R) is a prime number.

4. CONCLUSION AND FUTURE WORK
In spite of the existence of many factorisation techniques, cryptographic algorithms are still in
work with the assumption that, factorising a very large number would take too long time to be
practical for the crypto-system to be insecure. Therefore, a new solution in the factorisation
family should draw much attention to the crypto-community. In this paper, only the algorithm is
discussed in brief without considering its implementation and performance evaluation. If the
proposed solution performs such that, it becomes capable to factorise those big numbers, then it
may break the security of some crypto-systems, such as RSA.
In future, the algorithm can be implemented to measure its performance. It is also important to
see whether this solution can factorise very big numbers used in some crypto-systems. If fails,
still its performance can be compared with other existing factorisation solutions. A hybrid
approach also can be thought to engage this solution with others. The algorithm itself can be
improved by introducing some pruning and optimisation techniques too.


83

REFERENCES
[1]

Richard P. Brent. An improved monte carlo factorization algorithm. BIT Numerical Mathematics,
20:176-184, 1980. 10.1007/BF01933190.
[2] Richard P. Brent. Factorization of the tenth fermat number. MATH.COMP, 68:429-451, 1999.
[3] J. D. Dixon. Asymptotically fast factorization of integers. Math. Comp.,36 (153):255260, 1981.
[4] Harris and David. Digital Design and Computer Architecture : From Gates to Processors. Elsevier,
Burlington 2007, 2007.
[5] A.K. Lenstra. Fast and rigorous factorization under the generalized riemann hypothesis. Indagationes
Mathematicae (Proceedings), 91(4):443-454, 1988.
[6] James Mckee. Turning euler's factoring method into a factoring algorithm. Bulletin of the London
Mathematical Society, 28, 1996.
[7] James McKee. Speeding fermat's factoring method. Math. Comput.,68(228):1729-1737, October
1999.
[8] J. M. Pollard. Theorems on factorization and primality testing. Mathematical Proceedings of the
Cambridge Philosophical Society, 76:521-528,1974.
[9] J. M. Pollard. A monte carlo method for factorization. BIT Numerical Mathematics, 15:331-334,
1975. 10.1007/BF01933667.
[10] Carl Pomerance. The quadratic sieve factoring algorithm. In Thomas Beth, Norbert Cot, and Ingemar
Ingemarsson, editors, Advances in Cryptology, volume 209 of Lecture Notes in Computer Science,
pages 169-182. Springer Berlin / Heidelberg, 1985. 10.1007/3-540-39757-417.
[11] R. L. Rivest, A. Shamir, and L. Adleman. A method for obtaining digital signatures and public-key
cryptosystems. Commun. ACM, 21(2):120-126, 1978.
[12] H. C. Williams. A p + 1 method of factoring. Math. Comput., 39:225-234,1982.

AUTHOR
Dr. Mohammed Kaosar is one of the faculty members in the School of Computing
and Mathematics, Faculty of Business, Charles Sturt University, Australia. He has
wide experience of teaching various ICT courses in several universities in
Australia, KSA and Bangladesh. Previously, he used to work as a post-doc
research fellow after his completion of PhD from the School of Engineering and
Science, Victoria University Melbourne, Australia. Prior to that, Dr. Kaosar
finished his MS in Computer Engineering and BSc in Computer Science and
Engineering in the year of 2006 and 2001 from KSA and Bangladesh respectively.
He also has experience of working in many academic, research and commercial
projects. He has published good number of research papers in high quality
journals and conferences including, IEEE Transactions on Knowledge and Data
Engineering (TKDE), Data & Knowledge Engineering (DKE), Computer Communications, IEEE
International Conference on Data Engineering (ICDE- 2012) etc. He is an active member of various
professional associations including IEEE, EAI.

Backtracking based integer factorisation, primality testing and square root calculation

Recommended

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Backtracking based integer factorisation, primality testing and square root calculation (20)

Recently uploaded (20)

Backtracking based integer factorisation, primality testing and square root calculation