Cryptography Reference
In-Depth Information
Example 14.2
Let E be the elliptic curve y 2 + y = x 3
x 2 considered in the previous example.
If we compute the number N p of points on E mod p for various primes, we
obtain, with a p = p +1
N p ,
a 2 = 2 ,
3 = 1 ,
5 =1 ,
7 = 2 ,
13 =4 ,...
(except for p =2 , 3 , 5, the numbers a p can be calculated using any of the
equations in the previous example). The value
a 11 =1
is specified by the formulas for bad primes. We then calculate the coecients
for composite indices. For example,
4 = a 2 2=2
a 6 = a 2 a 3 =2 ,
(see (14.2) and (14.1)). Therefore,
f E ( τ )= q − 2 q 2
− q 3 +2 q 4 + q 5 +2 q 6
2 q 7 + ··· .
It can be shown that
f ( τ )= q
(1 − q j ) 2 (1 − q 11 j ) 2
j =1
is a cusp form of weight 2 and level N = 11. In fact, it is the only such form,
up to scalar multiples. The product for f can be expanded into an infinite
series
f ( τ )= q − 2 q 2
− q 3 +2 q 4 + q 5 +2 q 6
2 q 7 + ··· .
It can be shown that f = f E (see [61]).
The L -series for E satisfies the functional equation
( 11 / 2 π ) s Γ( s ) L E ( s )=+( 11 / 2 π ) 2 −s Γ(2 − s ) L E (2 − s ) .
In the early 1960s, Birch and Swinnerton-Dyer performed computer exper-
iments to try to understand the relation between the number of points on
an elliptic curve mod p as p ranges through the primes and the number of
rational points on the curve. Ignoring the fact that the product for L E ( s )
doesn't converge at s = 1, let's substitute s = 1 into the product (we'll ignore
the finitely many bad primes):
p
1
1 − a p p 1 + p 1 1 =
p
=
p
a p +1
p
p
N p .
p
 
Search WWH ::




Custom Search