当前位置:网站首页>Mathematical basis of information security Chapter 4 -- quadratic residual and square root

Mathematical basis of information security Chapter 4 -- quadratic residual and square root

2022-06-11 17:28:00 L3H_ CoLin

Definition 4.1 set up m As a positive integer , If the congruence x 2 ≡ a ( m o d    m ) , ( a , m ) = 1 x^2\equiv a(\mod m),(a,m)=1 x2a(modm),(a,m)=1, Have solution , be a It's called a model m Quadratic residue of , Otherwise it is called a module m Quadratic non residual

Definition 4.2 Legendre symbol : ( a p ) (\frac{a}{p}) (pa), When a For the mould p The second remaining time of , The value is 1; Not 2 Time remaining value is -1, if p ∣ a p|a pa Then the value is 0

Theorem 4.1 p As a prime number , if a ≡ b ( m o d    p ) a\equiv b(\mod p) ab(modp), be ( a p ) = ( b p ) (\frac{a}{p})=(\frac{b}{p}) (pa)=(pb)
Congruence x 2 ≡ a ( m o d    p ) x^2\equiv a(\mod p) x2a(modp) The solution of can be seen as in a finite field Z p \mathbb Z_p Zp Find the polynomial in x 2 − a x^2-a x2a The root of the .

Theorem 4.2 Euler's criterion :p Is an odd prime number , For any integer a, ( a p ) ≡ a p − 1 2 ( m o d    p ) (\frac{a}{p})\equiv a^{\frac{p-1}{2}}(\mod p) (pa)a2p1(modp)
prove :
set up g g g yes Z p \mathbb Z_p Zp The original elements of , be Z p ∗ = { g 0 , g 1 , . . . , g p − 2 } \mathbb Z_p^*=\{g^0,g^1,...,g^{p-2}\} Zp={ g0,g1,...,gp2}
For all 0 ≤ i ≤ p − 1 2 0\le i\le \frac{p-1}{2} 0i2p1, ( g 2 i ) p − 1 2 = ( g p − 1 ) i = 1 , ( g 2 i + 1 ) p − 1 2 = ( g p − 1 ) i g p − 1 2 , g p − 1 2 = − 1 (g^{2i})^{\frac{p-1}{2}}=(g^{p-1})^i=1,(g^{2i+1})^{\frac{p-1}{2}}=(g^{p-1})^ig^{\frac{p-1}{2}},g^{\frac{p-1}{2}}=-1 (g2i)2p1=(gp1)i=1,(g2i+1)2p1=(gp1)ig2p1,g2p1=1( from g We know from the original g p − 1 2 ≠ 1 g^{\frac{p-1}{2}}\ne 1 g2p1=1, but ( g p − 1 2 ) 2 = 1 (g^{\frac{p-1}{2}})^2= 1 (g2p1)2=1, So it can only be -1), so ( g p − 1 2 ) 2 i + 1 = − 1 (g^{\frac{p-1}{2}})^{2i+1}=-1 (g2p1)2i+1=1
Due to the consideration of module p The quadratic congruence of , therefore a You can view it as Z p \mathbb Z_p Zp The element equivalent to its congruence in
When a ≡ g 2 i ( m o d    p ) , 0 ≤ i < p − 1 2 a\equiv g^{2i}(\mod p),0\le i<\frac{p-1}{2} ag2i(modp),0i<2p1, polynomial x 2 − a = x 2 − g 2 i x^2-a=x^2-g^{2i} x2a=x2g2i Rooted ±gi, so ( a p ) = 1 ≡ a p − 1 2 ( m o d    p ) (\frac{a}{p})=1\equiv a^{\frac{p-1}{2}}(\mod p) (pa)=1a2p1(modp)
When a ≡ g 2 i + 1 ( m o d    p ) , 0 ≤ i < p − 1 2 a\equiv g^{2i+1}(\mod p),0\le i<\frac{p-1}{2} ag2i+1(modp),0i<2p1, polynomial x 2 − a = x 2 − g 2 i + 1 x^2-a=x^2-g^{2i+1} x2a=x2g2i+1 There must be no root . Otherwise, if x 0 2 = g 2 i + 1 x_0^2=g^{2i+1} x02=g2i+1, that 1 = ( x 0 2 ) p − 1 2 = ( g 2 i + 1 ) p − 1 2 = − 1 1=(x_0^2)^{\frac{p-1}{2}}=(g^{2i+1})^{\frac{p-1}{2}}=-1 1=(x02)2p1=(g2i+1)2p1=1 contradiction . so ( a p ) = − 1 ≡ a p − 1 2 ( m o d    p ) (\frac{a}{p})=-1\equiv a^{\frac{p-1}{2}}(\mod p) (pa)=1a2p1(modp)

According to the above Theorem , model p The quadratic remainder of is p − 1 2 \frac{p-1}{2} 2p1 individual ( All even powers of primitive )

inference set up p Is an odd prime number , be
( 1 p ) = 1 (\frac{1}{p})=1 (p1)=1
( − 1 p ) = ( − 1 ) p − 1 2 (\frac{-1}{p})=(-1)^\frac{p-1}{2} (p1)=(1)2p1
( a b p ) = ( a p ) ( b p ) (\frac{ab}{p})=(\frac{a}{p})(\frac{b}{p}) (pab)=(pa)(pb)
( a n p ) = ( a p ) n , n > 0 (\frac{a^n}{p})=(\frac{a}{p})^n,n>0 (pan)=(pa)n,n>0

Theorem 4.3 set up p Is an odd prime number , be ( 2 p ) = ( − 1 ) p 2 − 1 8 (\frac{2}{p})=(-1)^{\frac{p^2-1}{8}} (p2)=(1)8p21
prove : Proof of construction
( p − 1 ) ! ≡ 1 ⋅ 3 ⋅ 5 ⋅ . . . ⋅ ( p − 2 ) ⋅ 2 ⋅ 4 ⋅ . . . ⋅ ( p − 1 ) (p-1)!\equiv 1\cdot 3\cdot 5\cdot ... \cdot (p-2)\cdot 2\cdot 4\cdot ... \cdot (p-1) (p1)!135...(p2)24...(p1)
about 4k+1 Type p Yes
≡ 1 ⋅ ( p − 2 ) ⋅ 3 ⋅ ( p − 4 ) ⋅ 5 ⋅ . . . ⋅ p − 3 2 ⋅ ( p − p − 1 2 ) ⋅ 2 p − 1 2 ⋅ ( p − 1 2 ! ) ( m o d    p ) \equiv 1\cdot (p-2)\cdot 3\cdot (p-4)\cdot 5\cdot ...\cdot \frac{p-3}{2}\cdot (p-\frac{p-1}{2})\cdot 2^{\frac{p-1}{2}}\cdot (\frac{p-1}{2}!)(\mod p) 1(p2)3(p4)5...2p3(p2p1)22p1(2p1!)(modp)( All the even numbers in the last half are mentioned 2 come out , The first half can be alternately mentioned -1)
≡ ( − 1 ) p − 1 4 ⋅ 2 p − 1 2 ⋅ ( p − 1 2 ! ) 2 ( m o d    p ) \equiv (-1)^{\frac{p-1}{4}}\cdot 2^{\frac{p-1}{2}}\cdot(\frac{p-1}{2}!)^2(\mod p) (1)4p122p1(2p1!)2(modp)
about 4k+3 Type p Yes
≡ 1 ⋅ ( p − 2 ) ⋅ 3 ⋅ ( p − 4 ) ⋅ 5 ⋅ . . . ⋅ ( p − p − 3 2 ) ⋅ p − 1 2 ⋅ 2 p − 1 2 ⋅ ( p − 1 2 ! ) ( m o d    p ) \equiv 1\cdot (p-2)\cdot 3\cdot (p-4)\cdot 5\cdot ...\cdot(p-\frac{p-3}{2})\cdot\frac{p-1}{2}\cdot 2^{\frac{p-1}{2}}\cdot (\frac{p-1}{2}!)(\mod p) 1(p2)3(p4)5...(p2p3)2p122p1(2p1!)(modp)
≡ ( − 1 ) p − 3 4 ⋅ 2 p − 1 2 ⋅ ( p − 1 2 ! ) 2 ( m o d    p ) \equiv (-1)^{\frac{p-3}{4}}\cdot 2^{\frac{p-1}{2}}\cdot(\frac{p-1}{2}!)^2(\mod p) (1)4p322p1(2p1!)2(modp)
Wilson Theorem know ( p − 1 ) ! ≡ − 1 ( m o d    p ) (p-1)!\equiv -1(\mod p) (p1)!1(modp), And ( p − 1 2 ! ) 2 ≡ ( − 1 ) p + 1 2 ( m o d    p ) (\frac{p-1}{2}!)^2\equiv(-1)^{\frac{p+1}{2}}(\mod p) (2p1!)2(1)2p+1(modp)( Turn into ( − 1 ) p − 1 2 ( p − 1 ) ! (-1)^{\frac{p-1}{2}}(p-1)! (1)2p1(p1)! Know when p ≡ ± 1 ( m o d    8 ) p\equiv ±1(\mod 8) p±1(mod8) when , 2 p − 1 2 ≡ 1 ( m o d    p ) 2^{\frac{p-1}{2}}\equiv 1(\mod p) 22p11(modp), When p ≡ ± 3 ( m o d    8 ) p\equiv ±3(\mod 8) p±3(mod8) when , 2 p − 1 2 ≡ − 1 ( m o d    p ) 2^{\frac{p-1}{2}}\equiv -1(\mod p) 22p11(modp), Comprehensive verification shows that 2 p − 1 2 ≡ ( − 1 ) p 2 − 1 8 ( m o d    p ) 2^{\frac{p-1}{2}}\equiv(-1)^{\frac{p^2-1}{8}}(\mod p) 22p1(1)8p21(modp), According to Euler's rule ( 2 p ) = ( − 1 ) p 2 − 1 8 (\frac{2}{p})=(-1)^{\frac{p^2-1}{8}} (p2)=(1)8p21

Theorem 4.4 The law of quadratic reciprocity :p,q Is a mutually prime odd prime number , be ( q p ) = ( − 1 ) p − 1 2 q − 1 2 ( p q ) (\frac{q}{p})=(-1)^{\frac{p-1}{2}\frac{q-1}{2}}(\frac{p}{q}) (pq)=(1)2p12q1(qp)
prove : It's too complicated , No need to master

Definition 4.3 Jacobi symbols : m = ∏ i = 1 n p i , p i m=\prod_{i=1}^np_i,p_i m=i=1npi,pi It's an odd prime , For any integer a Definition a model m The Jacobian symbol of is ( a m ) = ∏ i = 1 n ( a p i ) (\frac{a}{m})=\prod_{i=1}^n(\frac{a}{p_i}) (ma)=i=1n(pia),m When it is an odd prime number , Its Jacobian symbol is Legendre symbol .

Theorem 4.5 set up m Positive odd number , a ≡ b ( m o d    m ) ⇒ ( a m ) = ( b m ) a\equiv b(\mod m)\Rightarrow (\frac{a}{m})=(\frac{b}{m}) ab(modm)(ma)=(mb)

Theorem 4.6 set up m Positive odd number , be
(1) ( 1 m ) = 1 (\frac{1}{m})=1 (m1)=1
(2) ( a b m ) = ( a m ) ( b m ) (\frac{ab}{ m})=(\frac{a}{m})(\frac{b}{m}) (mab)=(ma)(mb)
(3) ( a n m ) = ( a m ) n (\frac{a^n}{m})=(\frac{a}{m})^n (man)=(ma)n
(4) ( − 1 m ) = ( − 1 ) m − 1 2 (\frac{-1}{m})=(-1)^{\frac{m-1}{2}} (m1)=(1)2m1
(5) ( 2 m ) = ( − 1 ) m 2 − 1 8 (\frac{2}{m})=(-1)^{\frac{m^2-1}{8}} (m2)=(1)8m21

Theorem 4.7 set up m,n Positive odd number , be ( n m ) = ( − 1 ) m − 1 2 n − 1 2 ( m n ) (\frac{n}{m})=(-1)^{\frac{m-1}{2}\frac{n-1}{2}}(\frac{m}{n}) (mn)=(1)2m12n1(nm)

Definition 4.4 Quadratic residual problem : Unknown n In the case of the decomposition of , Generally judge an integer a Whether it is a module n The quadratic residue of is a difficult problem , It is called quadratic residual problem .

encryption algorithm 1——Rabin encryption algorithm
Alice Select two 4k+3 Prime of type ( be called Blum prime number )p,q, Calculation n=pq, take p,q Public as private key n.
encryption : Plaintext is an integer m, Ciphertext c=m2(mod n)
Decrypt : Solve the congruence equation c=x2(mod n) You can get 4 A solution , Choose the meaningful solution as the plaintext m.

computing method ——a=x2(mod p),p=4k+3 Solution method
If the above formula has a solution , It's in [0,p-1] There must be a solution , Therefore, when the number is not large, it can be used to a Keep adding p Until you find a perfect square ( This method is right p nothing 4k+3 The limitation of , however p It is inconvenient when it is very big )
from ( a p ) = 1 (\frac{a}{p})=1 (pa)=1 According to Euler's rule a p − 1 2 ≡ 1 ( m o d    p ) a^{\frac{p-1}{2}}\equiv 1(\mod p) a2p11(modp), There are ( a p − 1 4 ) 2 ≡ a ( m o d    p ) (a^{\frac{p-1}{4}})^2\equiv a(\mod p) (a4p1)2a(modp), Therefore, the solution is x ≡ ± a p − 1 4 ( m o d    p ) x\equiv ±a^{\frac{p-1}{4}}(\mod p) x±a4p1(modp)

encryption algorithm 2——Goldwasser-Micali encryption algorithm
Alice Choose two different prime numbers p,q, And integer y Satisfy ( y p ) = ( y q ) = − 1 (\frac{y}{p})=(\frac{y}{q})=-1 (py)=(qy)=1. Calculation n=pq,p and q The seat private key is public n,y
encryption : Convert binary integers m As clear text , The first i The place is marked as bi, For everyone , Random selection 0<xi<n, If the bit is 0 Calculation ci=xi2(mod n), Otherwise, calculate ci=yxi2(mod n), Ciphertext for all c
Decrypt : if ci For the mould n Quadratic residue of , Then judge bi=0, otherwise bi=1

Definition 4.5 set up <g> Is an element that consists of g The generated one n Metacyclic group , Then for any a∈<g>, There is 0≤i<n,a=gi, call i For g Bottom a Indicators of , Write it down as indga. The problem of finding indicators , In cryptography, it is usually called discrete logarithm problem .n It is a difficult problem to solve the discrete logarithm problem when the integer is large enough .

Theorem 4.8 set up <g> It's a n Metacyclic group ,a∈<g>, If for a positive integer m Yes :
(1) am=e
(2) For any prime factor p|m, a m p ≠ e a^{\frac{m}{p}}\ne e apm=e, be ord(a)=m, And m|n

Definition 4.8 Primordial root : set up m As a positive integer , Integers a Satisfy (a,m)=1,a model m The order of ordm(a) Refer to a(mod m) stay Z m ∗ \mathbb Z_m^* Zm Order in ; If Z m ∗ \mathbb Z_m^* Zm Is a cyclic group , Integers a It's called a model m The primordial root of means a(mod m) by Z m ∗ \mathbb Z_m^* Zm The generator of

According to the above definition ,a In the mold m Modulo of all integers in the remaining classes m All orders are ordm(a)

According to the original root definition : When m=2,4 when , model m The original roots are 1,3
In a general way , If and only if m=2,4,pa,2pa(p Is an odd prime number ,a≥1), model m It has roots

Theorem 4.9 set up <g> It's a n Metacyclic group ,a,b∈<g>, be indgab ≡ \equiv indga+indgb(mod n)
prove :indga=x,indgb=y, be gx+y=ab= g i n d g a b g^{ind_gab} gindgab
namely g x + y − i n d g a b = e g^{x+y-ind_gab}=e gx+yindgab=e, also ord(g)=n, so n|x+y-indgab, Therefore, the conclusion is tenable

Definition 4.9 set up m It is greater than 1 The positive integer , If n Second congruence xn=a(mod m), (a,m)=1 Have solution , be a It is called a module m Of n The second surplus , Otherwise, it is a module m Of n Secondary non residual .

Theorem 4.14 ( High order residue ) set up m For more than 1 The positive integer ,g For the mould m One of the original roots ,(a,m)=1,d=(n, φ \varphi φ(m)), that xn=a(mod m) The necessary and sufficient condition for a solution is a φ ( m ) d ≡ 1 ( m o d    m ) a^{\frac{\varphi(m)}{d}}\equiv 1(\mod m) adφ(m)1(modm)
prove :g For the mould m One of the original roots , therefore Z m ∗ = < g > , x n ≡ a ( m o d    m ) \mathbb Z_m^*=<g>,x^n\equiv a(\mod m) Zm=<g>,xna(modm) The necessary and sufficient condition for a solution is i n d g x n = i n d g a ⇒ n i n d g x ≡ i n d g a ( m o d    φ ( m ) ) ind_gx^n=ind_ga\Rightarrow nind_gx\equiv ind_ga(\mod \varphi(m)) indgxn=indganindgxindga(modφ(m))( Pay attention to the mold m The cyclic group of is only φ ( m ) \varphi(m) φ(m) Elements , Therefore, it is necessary to mold φ ( m ) \varphi(m) φ(m)!), Make X=indgx, Then there are n X ≡ i n d g a ( m o d    φ ( m ) ) nX\equiv ind_ga(\mod \varphi(m)) nXindga(modφ(m))
The necessary and sufficient condition for the first congruence to have a solution is (n,φ(m))|indga, namely d|indga, Equivalent to indga ≡ \equiv 0(mod d)
From theorem 2.4(4) Yes φ ( m ) d i n d g a ≡ 0 ( m o d    φ ( m ) ) \frac{\varphi(m)}{d}ind_ga\equiv 0(\mod \varphi(m)) dφ(m)indga0(modφ(m)). Take... On both sides “ Index ” have to a φ ( m ) d ≡ 1 ( m o d    m ) a^{\frac{\varphi(m)}{d}}\equiv 1(\mod m) adφ(m)1(modm), So the original proposition holds .

notes : The theorem can also help to solve the number of solutions of high-order congruence . For congruence a x ≡ b ( m o d    m ) ax\equiv b(\mod m) axb(modm), The necessary and sufficient condition for its solution is ( a , m ) ∣ b (a,m)|b (a,m)b, And the general solution can be written as x = x 0 + m ( a , m ) t , t = 0 , 1 , . . . , ( a , m ) − 1 x=x_0+\frac{m}{(a,m)}t,t=0,1,...,(a,m)-1 x=x0+(a,m)mt,t=0,1,...,(a,m)1 In the form of , So the number of solutions is ( a , m ) (a,m) (a,m). that n X ≡ i n d g a ( m o d    φ ( m ) ) nX\equiv ind_ga(\mod \varphi(m)) nXindga(modφ(m)) The solution of should have ( n , φ ( m ) ) (n,\varphi(m)) (n,φ(m)) individual

原网站

版权声明
本文为[L3H_ CoLin]所创,转载请带上原文链接,感谢
https://yzsam.com/2022/162/202206111713249835.html