Maximum-likelihood detection for information channels with colored noise

In data storage systems, information channels have generally two different noise origins: system noise and media noise; and total noise contained in a playback signal is not necessarily white noise. Therefore, conventional partial-response maximum-likelihood (PRML) detection methods are not optimal for data storage systems. We developed a new detection method suitable for colored noise channels and obtained better performance than with the conventional method. In this paper, we report the mechanism of the new method and the experimental results of its performance.