0
$\begingroup$

I am learning elliptic curves theorem and I have read in more papers that for two distinct points $P$ and $Q$ there is always point $R$ such that $P+Q+R = \infty$. I know that this point should be unique according to all possible shapes of the curve and intersections (http://en.wikipedia.org/wiki/Elliptic_curve). But is there a simple way how to prove uniqueness of such a point $R$?

$\endgroup$
  • 5
    $\begingroup$ The points on an elliptic curve form an abelian group...and everything in a group has an inverse. $\endgroup$ – fretty Nov 17 '13 at 12:38
  • 1
    $\begingroup$ @fretty The question likely being asked in the context of trying to prove exactly this. Maybe the OP can clarify what the context is (in particular, how "$+$" is defined in this context)? $\endgroup$ – Jan Ladislav Dussek Nov 18 '13 at 19:44
  • $\begingroup$ What is true is that as divisors on the elliptic curve $E$, for all $P,Q$, there exists $R$ such that $P+Q \sim \infty + R$ (linear equivalence). This results from Riemann-Roch theorem for $E$. Moreover $R$ is unique because if another $R'$ satisfies the same relation, then $R\sim R'$. This can happen with $R\ne R'$ if and only if we are in a curve of genus $0$, but $E$ has genus $1$. So $R=R'$. $\endgroup$ – Cantlog Nov 18 '13 at 22:22
  • $\begingroup$ I just had a look at the wikipedia page you linked to. The $R$ is such that $P+Q+R \sim 3 \infty$. It is given by the intersection of the cubic (elliptic curve) with the line in $\mathbb P^2$ passing through $P, Q$. By Bezout, this intersection consists in general in three points $P, Q, R$. But $R$ could coincide with $P$ or $Q$. $\endgroup$ – Cantlog Nov 18 '13 at 22:26
1
$\begingroup$

A sketch of a proof may look like this:

As what was said in the comments, it is obvious that $P+Q\in E$. So let's just say that $S=P+Q$. Now, since points on an elliptic curve form a group (abelian group), there is an inverse for $S$. Notice that $S$ must satisfy the equation of $E$ to lie on $E$.

So say $E$ is given by the equation $y^2=x^3+ax+b$. Now, since $\mathcal{O}$ is the point of infinity, the inverse of our point $S$ must lie on the line $$x=\text{$x$-coordinate of $S$}.$$

Note that plugging the line into the equation of $E$ will give us at most two solutions, so one solution belongs to $S$, while the other gives us the unique inverse.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.