Posted on

Table of Contents

Proposition 1

Definition

If a,b,mZa,b,m\in \Z and m0m \neq 0, then we can say that a is congruent to bb modulo mm if m divides bab-a. $a \equiv b (m) $

  • aa(m)a\equiv a(m)
  • ab(m)    ba(m)a\equiv b(m) \implies b \equiv a(m)
  • ab(m),bc(m)    ac(m)a\equiv b(m),b\equiv c(m) \implies a\equiv c(m)

Proof:

  • m(aa)=0m|(a-a)=0
  • m(ab)    m(ba)m|(a-b) \implies m|(b-a)
  • m(ba),m(cb)    m(cb+ba)=cam|(b-a),m|(c-b)\implies m|(c-b+b-a)=c-a

That implies, congruence modulo is an equivalence relation on a set of integer.

Proposition 2

Definition

let aˉ\bar a as a set of integers that aˉa(m)\bar a \equiv a(m). aˉ=a+km\bar a = a+km​. aˉ\bar a is called a congruence class modulo mm.

  • aˉ=bˉ    ab(m)\bar a=\bar b \iff a \equiv b(m)
  • aˉbˉ    aˉbˉ=\bar a \neq \bar b \iff \bar a \cap \bar b = \empty
  • There are m distinct congruence classes modulo m

Proof:

  • If aˉ=bˉ\bar a = \bar b, aaˉ=bˉa\in \bar a = \bar b that is ab(m)a\equiv b(m). If ab(m),a\equiv b(m), abˉa\in \bar b so aˉbˉ\bar a \subset \bar b.
  • If aˉbˉ\bar a \cap \bar b \neq \empty, we shall have ca(m),cb(m)c\equiv a(m), c\equiv b(m). That is ab(m)a\equiv b(m) so aˉ=bˉ\bar a = \bar b.
  • Suppose 0k<l<m0 \leq k < l < m and kˉ=lˉ\bar k = \bar l. That implies kl(m)    m(lk)k \equiv l(m) \implies m|(l-k). But (lm)<m(l-m) < m, that is a contradiction. So for 0k<l<m0 \leq k < l < m, kˉlˉ\bar k \neq \bar l. So there are m distinct congruence classes.

Proposition 3

Definition

If aˉ1,aˉ2,aˉ3...aˉm\bar a_1,\bar a_2, \bar a_3...\bar a_m are a complete set of congruence classes modulo m, then {aˉ1,aˉ2,aˉ3...aˉm}\{\bar a_1,\bar a_2, \bar a_3...\bar a_m\} is called a complete set of residues modulo m.

The set of congruence classes modulo m is Z/mZ\Z/m\Z.

This set can be made into a ring.

If ac(m),bd(m)a\equiv c(m), b\equiv d(m), then a+bc+d (m)a+b\equiv c+d\ (m) also abcd(m)ab\equiv cd(m).

Proof:

By definition, mca,mdbm|c-a, m| d-b. m(ca)+(db)=(c+d)(a+b)    a+bc+d (m) m|(c-a)+(d-b) = (c+d)-(a+b) \implies a+b \equiv c+d \ (m)

mc(db)+b(ca)=cdab    abcd (m) m|c(d-b)+b(c-a) = cd -ab \implies ab \equiv cd\ (m)

Application

If p(x)Z[x]p(x)\in \Z[x], assume ab(m)a\equiv b(m), we have p(a)p(b)(m)p(a)\equiv p(b)(m). So when m=2m=2, we have p(a)p(0)(2)p(a)\equiv p(0)(2) or p(a)p(1)(2)p(a)\equiv p(1)(2).

For p(x)=a0xn+a1xn1...an1x+anp(x)=a_0x^n+a_1x^{n-1}...a_{n-1}x+a_n. p(0)=anp(0)=a_n and p(1)=a0+a1+...anp(1)=a_0 + a_1+...a_n.

If p(x)Z[x],p(0),p(1)p(x)\in \Z[x], p(0),p(1) are both odd, there are no integer roots.