On the Collusion-Resistance of the Fingerprinting Algorithm Based on Additive Watermarking
Digital fingerprinting is the technique used for traitor tracing in multimedia content distribution. It embeds unique customer information (e.g., customer ID) into multimedia content and generates different copies for different customers. The unique customer information can be extracted from the illegal copy and used to tell the traitor who re-distributes his copy to other unauthorized customers. Till now, there exist some fingerprint embedding algorithms, such as additive embedding, multiplicative embedding, quantization embedding, etc. However, the ability to resist collusion attacks is not studied. Collusion attack is the biggest threat to fingerprinting algorithms, which combines several copies together to produce a new copy without customer information.