Exercise 2.9.3* (Number of solutions modulo $p^2$)

Let f ( x ) = a x 2 + bx + c , and let p be an odd prime that does not divide all the coefficients a , b , c . Show that the congruence f ( x ) ≡ 0 ( mod p 2 ) has either 0 , 1 , 2 ,or p solutions.

Answers

Note first that if a x 2 + bx + c ≡ 0 mod p 2 , then a x 2 + bx + c ≡ 0 mod p

∙
Assume first that p ∤ a , and p ∤ D = b 2 − 4 ac . Then the equation f ( x ) ≡ 0 ( mod p ) has two distinct solutions x 0 , x 0 ′ modulo p (see Exercise 2.9.2). If x 1 is a solution of f ( x ) ≡ 0 ( mod p 2 ) , then f ( x 1 ) ≡ 0 ( mod p 2 ) , a fortiori f ( x 1 ) ≡ 0 ( mod p ) , thus x 1 ≡ x 0 ( mod p ) or x 1 ≡ x 0 ′ ( mod p ) .

If x 1 ≡ x 0 ( mod p ) , then x 1 = x 0 + kp , k ∈ ℤ , where a x 0 2 + b x 0 + c = λp , λ ∈ ℤ . Since

f ( x 1 ) = a ( x 0 + kp ) 2 + b ( x 0 + kp ) + c = ( a x 0 2 + b x 0 + c ) + ( 2 a x 0 + b ) kp = λp + ( 2 a x 0 + b ) kp ,

we obtain

f ( x 1 ) ≡ 0 ( mod p 2 ) ⟺ λp + ( 2 a x 0 + b ) kp ≡ 0 ( mod p 2 ) ⟺ λ + f ′ ( x 0 ) k ≡ 0 ( mod p ) .

By Exercise 2.9.2, in the case p ∤ a , p ∤ D , we know that p ∤ f ′ ( x 0 ) . Therefore the equation λ + f ′ ( x 0 ) k ≡ 0 ( mod p ) has a unique solution k .

We can conclude that the equation f ( x ) ≡ 0 ( mod p 2 ) has two solutions modulo p 2 , which are x 0 + kp , x 0 ′ + k ′ p , where k ≡ − λ f ′ ( x 0 ) − 1 ( mod p ) , k ′ ≡ − λ ′ f ′ ( x 0 ′ ) − 1 ( mod p ) , and λ = f ( x 0 ) ∕ p , λ ′ = f ( x 0 ′ ) ∕ p .

∙
Assume now that p ∤ a , but p ∣ D = b 2 − 4 ac . Then the equation f ( x ) ≡ 0 ( mod p ) has a unique solution x 0 = 2 − 1 a − 1 b mod p , where f ′ ( x 0 ) ≡ 0 ( mod p ) (see Exercise 2.9.2).

A solution x 1 of f ( x ) ≡ 0 ( mod p 2 ) is of the form x 1 = x 0 + kp , k ∈ ℤ . Since f ( x 0 ) = a x 0 2 + b x 0 + c = λp for some λ ∈ ℤ , and f ′ ( x 0 ) = 0 , we obtain

f ( x 1 ) ≡ 0 ( mod p 2 ) ⟺ λ ≡ 0 ( mod p ) ⟺ p 2 ∣ f ( x 0 ) .

If p 2 ∤ f ( x 0 ) , there is no solution, and if p 2 ∣ f ( x 0 ) any value of k gives a solution, so that there are p solutions x 0 , x 0 + p , ⋯ , x 0 + ( p − 1 ) p distinct modulo p 2 .

∙
if p ∣ a , p ∤ b , the equation f ( x ) = bx + c ≡ 0 ( mod p ) has a unique solution x 0 ≡ − c b − 1 ( mod p ) . if x 1 = x 0 + kp is a solution of f ( x ) ≡ 0 ( mod p 2 ) , then b x 0 + c = λp for some integer λ . f ( x 1 ) ≡ 0 ( mod p 2 ) ⟺ b ( x 0 + kp ) + c ≡ 0 ( mod p 2 ) ⟺ λp + bkp ≡ 0 ( mod p 2 ) ⟺ λ + bk ≡ 0 ( mod p ) .

Since p ∤ b , this equation has exactly one solution modulo p , which give a unique solution x 0 + kp of f ( x ) ≡ 0 ( mod p 2 ) .

∙
If p ∣ a , p ∣ b , then p ∤ c , and the equation f ( x ) = c ≡ 0 mod p 2 has no solution.

To conclude, the number N of solutions modulo p 2 of a x 2 + bx + c ≡ 0 mod p 2 is given by

a b c D ( D p ) f ( x 0 ) N p ∤ a p ∤ D − 1 0 p ∤ a p ∤ D 1 2 p ∤ a p ∣ D 0 p 2 ∤ f ( x 0 ) 0 p ∤ a p ∣ D 0 p 2 ∣ f ( x 0 ) p p ∣ a p ∤ b 1 p ∣ a p ∣ b p ∤ c 0

where x 0 = 2 − 1 a − 1 b .

User profile picture
2024-05-18 09:21
Comments