A new embedding method for robust ISS — watermarking

In this paper, We propose a new algorithm for the improved spread spectrum (ISS) embedding/decoding system to increase its robustness against malicious attacks. The ISS receiver can only detect the watermark bit and as a result, a pirate has the ability to befuddle the ISS detector by sending a pure signal and making the detector incorrectly detect it as a watermarked signal. For preventing such attacks, we modify the detector of ISS receiver in a way that it distinguishes the watermarked signal from a pure one. To achieve this, we introduce a threshold such that the signal is considered pure if the correlation amount is within the threshold region, and considered watermarked otherwise. The false alarm probability of our system is much lower than that in ISS. This mechanism has the same distortion as that in ISS and, under special conditions; our system has the decoding error probability smaller than the same error in ISS. At the end, our system efficiency is confirmed via experimental results too.