Re: question about multi-prime rsa



On Nov 11, 2:26 pm, yawnmoth <terra1...@xxxxxxxxx> wrote:
From <http://tools.ietf.org/html/rfc3447#page-8>:

-----------------------

   Each CRT coefficient t_i (i = 3, ..., u) is a positive integer less
   than r_i satisfying

      R_i * t_i == 1 (mod r_i) ,

   where R_i = r_1 * r_2 * ... * r_(i-1).

-----------------------

So t_3 satisfies the following:

R_i * t_i == 1 (mod r_3)

You mean R_3 * t_3 == 1 (mod r_3)


Where R_i = r_1 * r_2

R_3 = r_1 * r_2.

And t_2, the following:

R_i * t_i == 1 (mod r_2)

Where R_i = r_1

Ehr... the definition above is for t_i, i=3,...,u. It does not specify
how to define t_1 or t_2. Could that be the source of the problem?

--
Arturo Magidin
.