Under the auspices of the Computational Complexity Foundation (CCF)

REPORTS > DETAIL:

Paper:

TR04-035 | 10th April 2004 00:00

Complexity of Inverting the Euler Function

TR04-035
Authors: Scott Contini, Ernie Croot, Igor E. Shparlinski
Publication: 14th April 2004 17:32
We present an algorithm to invert the Euler function $\varphi(m)$. The algorithm, for a given integer $n\ge 1$, in polynomial time on average'', finds theset $\Psi(n)$ of all solutions $m$ to the equation $\varphi(m) =n$. In fact, in the worst case the set $\Psi(n)$ is exponentially large and cannot be constructed by a polynomial time algorithm. In the opposite direction, we show, under some widely accepted number theoretic conjecture, that the problem of deciding whether $\varphi(m) =n$ for some $m$ is {\bf NP}-complete. Finally, we establish close links between the problem of inverting the Euler function and the integer factorisation problem.