New Adaptive Partial Response Maximum Likelihood Detection for High-Density Optical Recording

Abstract

Adaptive partial response maximum likelihood (PRML) detection is proposed to improve the detection performance for nonlinearity of the readout signal, such as an asymmetry compared with normal PRML detection. However, the target levels of the equalizer are defined according to predetermined PR polynomial, the same as in the normal PRML. Such an equalization causes an undesirable noise enhancement when the readout signal characteristics is quite different from the selected PR. To resolve this problem, we have proposed a new adaptive PRML detection in which both of the target levels of the equalizer and the reference levels of branch metrics are defined by the equalizer input. We have confirmed that our proposed method can effectively improve bit error rate (bER) compared with the conventional adaptive PRML and the normal PRML detections. The bER of the new PRML detection is lower than one-half of those of previous methods for a high-density recording signal including a significant nonlinearity.