Page 278 -
P. 278

§1. The Riemann Hypothesis for Elliptic Curves over a Finite Field  255

                                                 2
                                                          3
        (1.4) Example. The elliptic curve E defined by y + y = x + x has five points
        over F 2 , namely
                           {∞ = 0,(0, 0), (0, 1), (1, 0), (1, 1)}.

        Hence E(F 2 ) is isomorphic to Z/5and N 1 = 5. Note that N 1 −1−q = 5−1−2 =
             √
        2 ≤ 2 2.
        (1.5) Definition. Let E be an elliptic curve defined over a finite field F q . The char-
        acteristic polynomial of Frobenius is
                  f E (T ) = det(1 − π E T ) = 1 − (Tr(π))T + qT 2  in Z[T ],

        and the zeta function ζ E (s) is the rational function in q s

                               f E (q −s )   1 − (Tr(π))q −s  + q 1−2s
                  ζ E (s) =                =                     .
                         (1 − q −s )(1 − q 1−s )  (1 − q −s )(1 − q 1−s )
        (1.6) Remark. The zeta function ζ E (s) has poles at s = 0and s = 1. The inequality
        in (1.2), called the Riemann hypothesis, is equivalent to the assertion that the roots of
         f E (T ) are complex conjugates of each other. These roots have absolute value equal
            √
        to 1/ q. In turn this condition is equivalent to the assertion that the zeta function
        ζ E (s) has zeros only on the line Re(s) = 1/2.

           For f E (q −s ) = 1 − (Tr π)q −s  + q 1−2s , we replace s by 1 − s, and we obtain the
        relation

            f E (q −(1−s) ) = 1 − (Tr π)q s−1  + q 2s−1  = q 2s−1 (1 − (Tr π)q −s  + q 1−2s )
                       = q 2s−1  f E (q −s ).

        This functional equation for f E and the invariance of the denominator of the zeta
        function under s changedto1 − s yield the following immediately.

        (1.7) Proposition. The zeta function ζ E (s) of E over F q satisfies the functional
        equation
                                           2s−1
                                ζ E (1 − s) = q  ζ E (s).
           There is a far reaching generalization of this zeta function to zeta functions for
        any projective variety over a finite field due to A. Weil. In fact, it is the focus of
        a vast conjectural program announced by Weil [1949], carried out up to the Rie-
        mann hypothesis primarily by Grothendieck and Artin [1963–64], and completed
        with Deligne’s [1973] proof of the Riemann hypothesis for smooth projective vari-
        eties. Note, for example, the generalization of (1.7) is a consequence of Poincar´ e du-
        ality in  -adic cohomology theory. For a survey of the whole story, see Katz [1976].
   273   274   275   276   277   278   279   280   281   282   283