Lecture slides by Lawrie Brown for Cryptography and Network Security, 5/e, by William Stallings, Chapter Chapter 4 Basic

c Concepts in Number Theory and Finite Fields.

Intro quote.

Finite fields have become increasingly important in cryptography. A number of cryptographic algorithms rely heavily on properties of finite fields, notably the Advanced Encryption Standard (AES) and elliptic curve cryptography. The main purpose of this chapter is to provide the reader with sufficient background on the concepts of finite fields to be able to understand the design of AES and other cryptographic algorithms that use finite fields. We begin, in the first three sections, with some basic concepts from number theory that are needed in the remainder of the chapter; these include divisibility, the Euclidian algorithm, and modular arithmetic.

Define concept of divisors. We say that a nonzero b divides a if a=m.b for some m, where a, b, and m are integers. That is, b divides a if there is no remainder on division. Can denote this as b|a, and say that b is a divisor of a. For example, the positive divisors of 24 are 1,2,3,4,6,8,12, and 24. And have 13 | 182; 5 | 30; 17 | 289; 3 | 33; 17 | 0.

Subsequently, we will need some simple properties of divisibility for integers, which are as follows: If a|1, then a = s1. If a|b and b|a, then a = sb. Any b ! 0 divides 0. If a | b and b | c, then a | c If b|g and b|h, then b|(mg + nh) for arbitrary integers m and n. e.g. b = 7; g = 14; h = 63; m = 3; n = 2. 7|14 and 7|63. To show: 7|(3 x 14 + 2 x 63) We have (3 x 14 + 2 x 63) = 7(3 x 2 + 2 x 9) And it is obvious that 7|(7(3 x 2 + 2 x 9))

Given any positive integer n and any nonnegative integer a, if we divide a by n, we get an integer quotient q and an integer remainder r that obey the following relationship: a = qn + r, where 0 <= r < n; q = floor(a/n) which is referred to as the division algorithm. The remainder r is often referred to as a residue. Figure 4.1a demonstrates that, given a and positive n, it is always possible to find q and r that satisfy the preceding relationship. Represent the integers on the number line; a will fall somewhere on that line (positive a is shown, a similar demonstration can be made for negative a). Starting at 0, proceed to n, 2n, up to qn such that qn <= a and (q + 1)n > a. The distance from qn to a is r, and we have found the unique values of q and r. For example: a = 11; n = 7; a = 11; n = 7; 11 = 1 x 7 + 4; 11 = (2) x 7 + 3; r=4q=1 r = 3 q = 2

Figure 4.1b provides another example.

One of the basic techniques of number theory is the Euclidean algorithm, which is a simple procedure for determining the greatest common divisor of two positive integers. Use the notation gcd(a,b) to mean the greatest common divisor of a and b. The positive integer c is said to be the greatest common divisor of a and b if c is a divisor of a and of b; and any divisor of a and b is a divisor of c. We also define gcd(0, 0) = 0. State that two integers a and b are relatively prime if their only common positive integer factor is 1, ie GCD(a,b)=1.

Illustrate how we can compute successive instances of GCD(a,b) = GCD(b,a mod b). Note this MUST always terminate since will eventually get a mod b = 0 (ie no remainder left). Answer is then the last non-zero value. In this case GCD(1970,1066)=2.

This example shows hpw to find d = gcd(a, b) = gcd(1160718174, 316258250), shown in tabular form. In this example, we begin by dividing 1160718174 by 316258250, which gives 3 with a remainder of 211943424. Next we take 316258250 and divide it by 211943424. The process continues until we get a remainder of 0, yielding a result of 1078

Given any positive integer n and any nonnegative integer a, if we divide a by n, we get an integer quotient q and an integer remainder r. In modular arithmetic we are only interested in the remainder (or residue) after division by some modulus, and results with the same remainder are regarded as equivalent. Two integers a and b are said to be congruent modulo n, if (a mod n) =(b mod n).

10

Note that the (mod n) operator maps all integers into the set of integers {0, 1, . . . (n 1)}, denoted Zn. This is referred to as the set of residues, or residue classes (mod n). We can perform arithmetic operations within the confines of this set, and this technique is known as modular arithmetic. Finding the smallest nonnegative integer to which k is congruent modulo n is called reducing k modulo n. Then note some important properties of modular arithmetic which mean you can modulo reduce at any point and obtain an equivalent answer.

11

Modular arithmetic exhibits the properties shown, see text for details & proofs. Here are examples of the three properties: Given 11 mod 8 = 3; 15 mod 8 = 7 [(11 mod 8) + (15 mod 8)] mod 8 = 10 mod 8 = 2 (11 + 15) mod 8 = 26 mod 8 = 2 [(11 mod 8) (15 mod 8)] mod 8 = 4 mod 8 = 4 (11 15) mod 8 = 4 mod 8 = 4 [(11 mod 8) x (15 mod 8)] mod 8 = 21 mod 8 = 5 (11 x 15) mod 8 = 165 mod 8 = 5

12

Example showing addition in GF(8), from Stallings Table 4.2a. Table 4.2 provides an illustration of modular addition and multiplication modulo 8. Looking at addition, the results are straightforward and there is a regular pattern to the matrix. Both matrices are symmetric about the main diagonal, in conformance to the commutative property of addition and multiplication. As in ordinary addition, there is an additive inverse, or negative, to each integer in modular arithmetic. In this case, the negative of an integer x is the integer y such that (x + y) mod 8 = 0. To find the additive inverse of an integer in the lefthand column, scan across the corresponding row of the matrix to find the value 0; the integer at the top of that column is the additive inverse; thus (2 + 6) mod 8 = 0.

13

Continuing the example showing multiplication in GF(8), from Stallings Table 4.2b. Both matrices are symmetric about the main diagonal, in conformance to the commutative property of addition and multiplication. Similarly, the entries in the multiplication table are straightforward. In ordinary arithmetic, there is a multiplicative inverse, or reciprocal, to each integer. In modular arithmetic mod 8, the multiplicative inverse of x is the integer y such that (x x y) mod 8 = 1 mod 8. Now, to find the multiplicative inverse of an integer from the multiplication table, scan across the matrix in the row for that integer to find the value 1; the integer at the top of that column is the multiplicative inverse; thus (3 x 3) mod 8 = 1. Note that not all integers mod 8 have a multiplicative inverse; more about that later.

14

If we perform modular arithmetic within Zn, the properties shown in Table 4.3 hold for integers in Zn We show in the next section that this implies that Zn is a commutative ring with a multiplicative identity element. Note that unlike ordinary arithmetic, the following statement is true only with the attached condition: if (a x b) = (a x c) (mod n) then b = c (mod n) if a is relatively prime to n In general, an integer has a multiplicative inverse in Zn if that integer is relatively prime to n. Table 4.2 cin the text shows that the integers 1, 3, 5, and 7 have a multiplicative inverse in Z 8, but 2, 4, and 6 do not.

15

We now describe an algorithm credited to Euclid for easily finding the greatest common divisor of two integers. This algorithm has significance subsequently in this chapter. The Euclidean algorithm is an efficient way to find the GCD(a,b), and is derived from the observation that if a & b have a common factor d (ie. a=m.d & b=n.d) then d is also a factor in any difference between them, vis: a-p.b = (m.d)-p.(n.d) = d.(m-p.n). See text for more detailed proof. Euclid's Algorithm keeps computing successive differences until it vanishes, at which point the greatest common divisor has been reached. Some pseudocode from the text for this algorithm is shown.

16

We now proceed to look at an extension to the Euclidean algorithm that will be important for later computations in the area of finite fields and in encryption algorithms such as RSA. For given integers a and b, the extended Euclidean algorithm not only calculate the greatest common divisor d but also two additional integers x and y that satisfy the following equation: ax + by = d = gcd(a, b). It should be clear that x and y will have opposite signs. Can extend the Euclidean algorithm to determine x, y, d, given a and b. We again go through the sequence of divisions indicated in Equation Set (4.3) and we assume that at each step i, we can find integers x and y that satisfy r = ax + by. In each row, we calculate a new remainder r , based on the remainders of the previous two rows. We know from the original Euclidean algorithm that the process ends with a remainder of zero and that the greatest common divisor of a and b is d = gcd(a, b) = r n. But we also have determined that d = r n = axn + byn.

17

An important problem is to find multiplicative inverses in such finite fields. Can show that such inverses always exist, & can extend the Euclidean algorithm to find them as shown. See text for discussion as to why this works.

18

Example showing how to find the inverse of 550 in GF(1759), adapted from Stallings Table 4.4. In this example, let us use a = 1759 and b = 550 and solve for 1759x + 550y = gcd(1759, 550). The results are shown in Table 4.4. Thus, we have 1759 x (111) + 550 x 355 = 195249 + 195250 = 1.

19

Groups, rings, and fields are the fundamental elements of a branch of mathematics known as abstract algebra, or modern algebra. In abstract algebra, we are concerned with sets on whose elements we can operate algebraically; that is, we can combine two elements of the set, perhaps in several ways, to obtain a third element of the set. These operations are subject to specific rules, which define the nature of the set. By convention, the notation for the two principal classes of operations on set elements is usually the same as the notation for addition and multiplication on ordinary numbers. However, it is important to note that, in abstract algebra, we are not limited to ordinary arithmetical operations. A group G, sometimes denoted by {G, }, is a set of elements with a binary operation, denoted by , that associates to each ordered pair (a, b) of elements in G an element (a b) in G, such that the following axioms are obeyed: Closure, Associative, Identity element, Inverse element. Note - we have used . as operator: could be addition +, multiplication x or any other mathematical operator. A group can have a finite (fixed) number of elements, or it may be infinite. Note that integers (+ve, -ve and 0) using addition form an infinite abelian group. So do real numbers using multiplication.

20

Define exponentiation in a group as the repeated use of the group operator. Note that we are most familiar with it being applied to multiplication, but it is more general than that. If the repeated use of the operator on some value a in the group results in every possible value being created, then the group is said to be cyclic, and a is a generator of (or generates) the group G.

21

Next describe a ring. In essence, a ring is a set in which we can do addition, subtraction [a b = a + (b)], and multiplication without leaving the set, and which obeys the associative and distributive laws. We denote a Ring as {R,+,.} With respect to addition and multiplication, the set of all n-square matrices over the real numbers form a ring. The set of integers with addition & multiplication form an integral domain.

22

Lastly define a field. In essence, a field is a set in which we can do addition, subtraction, multiplication, and division without leaving the set. Division is defined with the following rule: a/b = a (b1). We denote a Field as {F,+,.} Examples of fields are: rational numbers, real numbers, complex numbers. Note that integers are NOT a field since there are no multiplicative inverses (except for 1).

23

These are terms we use for different sorts of "number systems", ones obeying different sets of laws. From group to ring to field we get more and more laws being obeyed, as shown here in Stallings Figure 4.2. As a memory aid, can use the acronym for groups: CAIN (Closure Associative Identity iNverse) & ABEL. Mostly we need to compute with Rings, if not Fields. When we do arithmetic modulo a prime, we have a field.

24

Infinite fields are not of particular interest in the context of cryptography. However, finite fields play a crucial role in many cryptographic algorithms. It can be shown that the order of a finite field (number of elements in the field) must be a positive power of a prime, & these are known as Galois fields, in honor of the mathematician who first studied finite fields, & are denoted GF(p^n). We are most interested in the cases where either n=1 - GF(p), or p=2 - GF(2^n).

25

Start by considering GF(p) over the set of integers {0p-1} with addition & multiplication modulo p. This forms a well-behaved finite field. Can find an inverse using the Extended Euclidean algorithm.

26

Table 4.5 shows arithmetic operations in GF(7). This is a field of order 7 using modular arithmetic modulo 7. As can be seen, it satisfies all of the properties required of a field (Figure 4.2). Compare this table with Table 4.2. In the latter case, we see that using modular arithmetic modulo 8, is not a field.

27

Next introduce the interesting subject of polynomial arithmetic, using polynomials in a single variable x, with several variants as listed above. Note we are usually not interested in evaluating a polynomial for any particular value of x, which is thus referred to as the indeterminate.

28

Polynomial arithmetic includes the operations of addition, subtraction, and multiplication, defined in the usual way, ie add or subtract corresponding coefficients, or multiply all terms by each other. The examples are from the text.

29

Consider variant where now when computing value of each coefficient do the calculation modulo some value, usually a prime. If the coefficients are computed in a field (eg GF(p)), then division on the polynomials is possible, and we have a polynomial ring. Are most interested in using GF(2) - ie all coefficients are 0 or 1, and any addition/subtraction of coefficients is done mod 2 (ie 2x is the same as 0x!), which is just the common XOR function.

30

Note that we can write any polynomial in the form of f(x) = q(x) g(x) + r(x), where division of f(x) by g(x) results in a quotient q(x) and remainder r(x). Can then extend the concept of divisors from the integer case, and show that the Euclidean algorithm can be extended to find the greatest common divisor of two polynomials whose coefficients are elements of a field. Define an irreducible (or prime) polynomial as one with no divisors other than itself & 1. If compute polynomial arithmetic modulo an irreducible polynomial, this forms a finite field, and the GCD & Inverse algorithms can be adapted for it.

31

We can extend the analogy between polynomial arithmetic over a field and integer arithmetic by defining the greatest common divisor as shown. We began this section with a discussion of arithmetic with ordinary polynomials. Arithmetic operations are performed on polynomials (addition, subtraction, multiplication, division) using the ordinary rules of algebra. Polynomial division is not allowed unless the coefficients are elements of a field. Next, we discussed polynomial arithmetic in which the coefficients are elements of GF(p). In this case, polynomial addition, subtraction, multiplication, and division are allowed. However, division is not exact; that is, in general division results in a quotient and a remainder. Finally, we showed that the Euclidean algorithm can be extended to find the greatest common divisor of two polynomials whose coefficients are elements of a field. All of the material in this section provides a foundation for the following section, in which polynomials are used to define finite fields of order pn.

32

Consider now the case of polynomial arithmetic with coordinates mod 2 and polynomials mod an irreducible polynomial m(x). That is Modular Polynomial Arithmetic uses the set S of all polynomials of degree n-1 or less over the field Zp. With the appropriate definition of arithmetic operations, each such set S is a finite field. The definition consists of the following elements: 1. Arithmetic follows the ordinary rules of polynomial arithmetic using the basic rules of algebra, with the following two refinements. 2. Arithmetic on the coefficients is performed modulo p. 3. If multiplication results in a polynomial of degree greater than n-1, then the polynomial is reduced modulo some irreducible polynomial m(x) of degree n. That is, we divide by m(x) and keep the remainder. This forms a finite field. And just as the Euclidean algorithm can be adapted to find the greatest common divisor of two polynomials, the extended Euclidean algorithm can be adapted to find the multiplicative inverse of a polynomial.

33

Example shows addition & multiplication in GF(23) modulo (x3+x+1), from Stallings Table 476.

34

A key motivation for using polynomial arithmetic in GF(2n) is that the polynomials can be represented as a bit string, using all possible bit values, and the calculations only use simple common machine instructions - addition is just XOR, and multiplication is shifts & XORs. See text for additional discussion. The shortcut for polynomial reduction comes from the observation that if in GF(2n) then irreducible poly g(x) has highest term xn , and if compute xn mod g(x) answer is g(x)- xn

35

Show here a few simple examples of addition, multiplication & modulo reduction in GF(23). Note the long form modulo reduction finds p(x)=q(x).m(x)+r(x) with r(x) being the desired remainder.

36

There is an equivalent technique for defining a finite field of the form GF(2n) using the same irreducible polynomial, based on powers of a generator of the group, which gives a nice implementation of multiplication. The generator can be found from the root of the irreducible polynomial, as discussed in the text.

37

Chapter 4 summary.

38