When using Pollard's method, we choose an , compute , then compute . This is divisible exactly by the primes such that . To reinterpret Pollard's method using the above isomorphism, let . Then , and the that divide are exactly the such that . By Theorem 2.1.19, these include the primes such that is -power smooth, where .
We will not define when is composite, since this is not needed for the algorithm (where we assume that is prime and hope for a contradiction). However, for the remainder of this paragraph, we pretend that is meaningful and describe a heuristic connection between Lenstra and Pollard's methods. The significant difference between Pollard's method and the elliptic curve method is that the isomorphism is replaced by an isomorphism (in quotes)
where is , and the of Pollard's method is replaced by . We put the isomorphism in quotes to emphasize that we have not defined . When carrying out the elliptic curve factorization algorithm, we attempt to compute and if some components of are , for some point that appears during the computation, but others are nonzero, we find a nontrivial factor of .
William 2007-06-01