The CT watermark has many
advantages over the static ones, for instance: it can survive from most
semantic-reserve transformations and it is hard to be removed by automatic
attacks. However, CT watermark alone
cannot provide adequate defences against attacks because it also has
vulnerabilities. The CT watermark can be distorted once the attacker locates
the code that builds the watermark | simple changes of one statement could make
the recognisor fail to return the correct value. Consequently, another layer of
protection besides watermarking techniques is needed. Because tamper-proofing
techniques are responsible for detecting and preventing code modifications, it
is natural to consider applying them to protect watermarks.