Homepage Solution manuals Ivan Niven An Introduction to the Theory of Numbers Exercise 2.3.47* (Number of integers $a,\ 1 \leq a \leq m,$ such that $(f(a),m) = 1$)

Exercise 2.3.47* (Number of integers $a,\ 1 \leq a \leq m,$ such that $(f(a),m) = 1$)

Let f ( x ) be a polynomial with integral coefficients, let N ( m ) denote the number of solutions of the congruence f ( x ) 0 ( mod m ) , and let Φ f ( m ) denote the number of integers a , 1 a m , such that ( f ( a ) , m ) = 1 . Show that if ( m , n ) = 1 then ϕ f ( mn ) = ϕ f ( m ) ϕ f ( n ) . Show that if α > 1 then ϕ f ( p α ) = p α 1 ϕ f ( p ) . Show that Φ f ( p ) = p N ( p ) . Conclude that for any positive integer n , Φ f ( n ) = n p n ( 1 N ( p ) p ) . Show that for an appropriate choice of f ( x ) , this reduces to Theorem 2.19.

Answers

In this solution, I prefer to consider mℤ rather than complete systems of residues modulo m .

Notations:

  • | A | = Card ( A ) refers to the cardinality of a set A .
  • a b is gcd ( a , b )
  • ( a , b ) is an ordered pair.
  • ( mℤ ) × is the group of invertible elements of mℤ . Its cardinality is ϕ ( m ) .
  • u = [ a ] m mℤ denotes the class of some element a .
  • Note that, if f ( x ) = i = 0 d a i x i [ x ] , then f ( u ) = f ( [ a ] m ) = i = 0 d a i [ a ] m i = [ i = 0 d a i a i ] m has a signification.

Proof.

a)
Let f [ x ] . Consider, for any integer m > 1 , the set A m = { u mℤ f ( u ) ( mℤ ) × }

Then, for all a ,

[ a ] m A m f ( a ) m = 1 .

By definition of ϕ f ( m ) , ϕ f ( m ) = | A m | .

Assume that m n = 1 . Consider the map

χ { A mn A m × A n [ a ] mn ( [ a ] m , [ a ] n )

We show first that χ is well defined.

  • If a b ( mod mn ) , then a b ( mod n ) and a b ( mod n ) , so that ( [ a ] m , [ a ] n ) does not depend of the choice of the representative a in the class [ a ] mn .
  • If [ a ] m A mn , then a mn = 1 . Therefore a m = 1 and a n = 1 , so ( [ a ] m , [ a ] n ) A m × A n .

Now we show that χ is bijective.

  • If χ ( u ) = χ ( v ) , where u = [ a ] mn , v = [ b ] mn , then ( [ a ] m , [ a ] n ) = ( [ b ] m , [ b ] n ) . Therefore

    a b ( mod m ) , a b ( mod n ) .

    Since m n = 1 , a b ( mod mn ) , so [ a ] mn = [ b ] mn , and u = v . This shows that f is injective (one-to-one).

  • Let ( u , v ) be any element of A m × A n . There are integers a , b such that u = [ a ] m , v = [ b ] n , where a m = 1 and b n = 1 .

    Since m n = 1 , the Chinese Remainder Theorem shows that there exists some integer x such that

    { x a ( mod m ) , x b ( mod n ) .

    So x = a + km for some integer k , thus x m = ( a + km ) m = a m = 1 . Similarly x n = 1 . Since x m = 1 and x n = 1 , then x mn = 1 , so [ x ] mn A mn . Moreover [ x ] m = [ a ] m = u , and [ x ] n = [ b ] n = v , thus χ ( [ x ] mn ) = ( [ a ] m , [ b ] n ) = ( u , v ) . This shows that χ is surjective (onto).

    We conclude that χ is a bijection. Therefore

    | A mn | = | A m | | A n | .

    This proves

    ϕ f ( mn ) = ϕ f ( m ) ϕ f ( n ) ,  if  m n = 1 .

b)
Suppose that α 1 . Consider the map ξ { A p α A p [ a ] p α [ a ] p .

  • The map ξ is well defined, because a b ( mod p α ) a b ( mod p ) .
  • If u = [ a ] p α A p α , then a p α = 1 . Therefore a p = 1 , so ξ ( v ) = [ a ] p A p .

If v = [ b ] p A p , we count the number of antecedents of v for χ , i.e. we estimate

| ξ 1 ( v ) | = Card { u = [ a ] p α A p α : ξ ( u ) = [ a ] p = v = [ b ] p } .

We can take the representative b of v in the set [ [ 0 , p 1 ] ] , and a [ [ 0 , p α 1 ] ] , which is a complete system of residues modulo p α . Then a = b + kp , where 0 k < p α 1 are the solutions a [ [ 0 , p α 1 ] ] of ξ ( [ a ] p ) = [ b ] p . Therefore

| ξ 1 ( v ) | = p α 1 ,

for all v = [ b ] p A p . This shows (principle of the shepherd, as said in France “principe du berger") that

| A p α | = p α 1 | A p | ,

that is

ϕ f ( p α ) = p α 1 ϕ f ( p ) .

c)
If p is a prime number, the complementary of A p in pℤ is pℤ A p = { [ a ] p pℤ f ( a ) p 1 } = { [ a ] p pℤ p f ( a ) } = { [ a ] p pℤ f ( a ) 0 ( mod p ) } = { u pℤ f ( u ) = 0 } .

Therefore | pℤ A p | = N ( p ) , by definition of N ( p ) . Thus | A p | = | pℤ | N ( p ) = p N ( p ) , so

ϕ f ( p ) = p N ( p ) .

d)
Write n = p 1 a 1 p k a k the decomposition in prime factors of some integer n > 1 , where a i > 0 .

By part (a),

ϕ f ( n ) = i = 0 k ϕ f ( p i a i ) .

By parts (b) and (c),

ϕ f ( p i a i ) = p i a i 1 ϕ f ( p i ) = p i a i 1 ( p i N ( p i ) ) = p i a i ( 1 N ( p i p i ) .

This gives

ϕ f ( n ) = i = 0 k p i a i ( 1 N ( p i p i ) = n i = 0 k ( 1 N ( p i ) p i ) = n p n ( 1 N ( p ) p ) .
e)
If f ( x ) = x , we obtain Theorem 2.17 (the number of solutions of x 0 ( mod p ) is 1 ).

If f ( x ) = x ( x + 1 ) , we obtain the solution of Problem 46 (the number of solutions of x ( x + 1 ) 0 ( mod p ) is 2 ).

Conclusion: if Φ f ( m ) denotes the number of integers a , 1 a m , such that f ( a ) m = 1 , then

ϕ f ( n ) = n p n ( 1 N ( p ) p ) ,

where N ( p ) denotes the number of solutions of the congruence f ( x ) 0 ( mod m )

User profile picture
2024-08-18 09:37
Comments