Data-dependent noise estimation in digital recording systems

In high-density recording systems performance is largely determined by noise, n general different noise sources are present, where each source has direct impact on the overall performance. Therefore the characterization of a noise source provides insightful information about the recording system. In general noise characterization involves the choice of an appropriate stochastic model for it and furthermore the estimation of the parameters of the selected stochastic model. The characterization results can be used as diagnostic information to evaluate existing recording systems and they play an important role in state of the art bit-detection techniques. In these techniques the estimated stochastic parameters are used as side information to improve the detection reliability. For this reason an accurate estimation is very important. The estimation algorithm proposed in this paper, achieves a high estimation accuracy for systems where both data-dependent media noise and correlated additive noise are present. This algorithm makes use of the difference in data-dependency between the two noise sources to jointly estimate the stochastic parameters of the two noise sources.