Robustness analysis for dynamic watermarks

In this paper we propose new attack on dynamic watermark scheme Radix-n widely known as one of the most secure. It was shown that we are able to reveal dynamically allocated memory blocks and then a digital watermark can be discovered by hooking API function calls. During the experiment it was ascertained that often an amount of such structures is low because of widely used optimization. This feature increases a probability of successful attack on dynamic watermark.