Improved Maximum Likelihood Syncmark Detection for Magnetic Recording Channels
Shaohua Yang, Dahua Qin · IEEE Transactions on Magnetics · 2013
Data framing errors can defeat advanced error correction coding schemes and cause read failures in a magnetic recording system. Acquisition and locating the syncmark is one of the first essential steps during a read event. The conventional syncmark detection scheme is based on comparing the received signal to a reference signal, and the syncmark location is declared once the distance between the two signals is smaller than a predetermined threshold. We propose a novel syncmark detection scheme based on hypothesis test and apply the maximum likelihood (ML) detection principle. The improved detection algorithm declares the syncmark location once the signal is more like the syncmark reference signal than any other possible presyncmark reference signals. With the same syncmark pattern, this new detection algorithm outperforms the conventional detection algorithm by multiple dBs using the same Euclidean metrics. For syncmark patterns affected by defects, the absolute value of the Euclidean metric can be used as detection quality indicator. For false detections due to random noise and other distortion, a likelihood metric ratio can be used to indicate detection reliability. The complexity of the new detection algorithm is comparable to the conventional threshold detector.