Attackers can tamper with programs to break usage conditions. Different software protection techniques have been proposed to limit the possibility of tampering. Some of them just limit the possibility to understand the (binary) code, others react more actively when a change attempt is detected. However, the validation of the software protection techniques has been always conducted without taking into consideration a unified process adopted by attackers to tamper with programs.
In this paper we present an extension of the mini-cycle of change, initially proposed to model the process of changing program for maintenance, to describe the process faced by an attacker to defeat software protections. This paper also shows how this new model should support a developer when considering what are the most appropriate protections to deploy.
PDF version of the paper.