The industry standard for single-track detection in magnetic recording is partial-response equalization followed by a trellis-based sequence detector. We extend for the first time the partial-response paradigm to the case of multitrack detection when the multiple tracks being jointly detected were written asynchronously, with different bit phases and bit rates. We propose a multiple-input multiple-output (MIMO) partial-response equalizer that equalizes the unsynchronized samples of the multiple readback waveforms to a time-varying MIMO target, thereby enabling a trellis-based rotating-target (ROTAR) detector that accounts for the asynchrony. We evaluate the proposed equalization strategy on a two-dimensional magnetic-recording channel, and find that the proposed receiver outperforms a conventional receiver that detects one track at a time, by a 30% reduction in the bit-error rate, and that it closely matches the performance of a fictitious system in which the tracks are perfectly synchronous.

The industry standard for single-track detection in magnetic recording is partial-response equalization followed by a trellis-based sequence detector. We extend for the first time the partial-response paradigm to the case of multitrack detection when the multiple tracks being jointly detected were written asynchronously, with different bit phases and bit rates. We propose a multiple-input multiple-output (MIMO) partial-response equalizer that equalizes the unsynchronized samples of the multiple readback waveforms to a time-varying MIMO target, thereby enabling a trellis-based sequence detector that is based on the resulted time-varying target to account for the asynchrony. We evaluate the proposed equalization strategy on a two-dimensional magnetic-recording channel, and find that the proposed receiver outperforms a conventional receiver that detects one track at a time, and that it closely matches the performance of a fictitious system in which the tracks are perfectly synchronous.

Two-dimensional magnetic recording promises to increase areal density through the joint detection of multiple tracks of interest. This paper concerns the problem of joint detection of multiple tracks that are written asynchronously, meaning that neither the bit boundaries (phase) nor the bit rate (frequency) are aligned between neighboring tracks. We propose the rotating-target (ROTAR) algorithm for jointly detecting multiple asynchronous tracks from one or more readback waveforms. The proposed approach is based on the joint Viterbi algorithm and a time-varying target that results when the asynchrony of the tracks is absorbed into the underlying target. Timing estimation for the tracks being detected is embedded inside the joint Viterbi detector using per-survivor processing. Performance results show that the proposed algorithm closely matches the performance of a fictitious system in which neighboring tracks are synchronous, and further that it significantly outperforms a previously reported detector that separately detects the two tracks.

Two-dimensional magnetic recording promises to increase areal density through the joint detection of multiple tracks of interest. This paper concerns the problem of joint detection of multiple tracks that are written asynchronously, meaning that neither the bit boundaries (phase) nor the bit rate (frequency) are aligned between neighboring tracks. We propose the rotating-target algorithm for jointly detecting multiple asynchronous tracks from one or more readback waveforms. The proposed approach is based on the joint Viterbi algorithm and a time-varying target that results when the asynchrony of the tracks is absorbed into the underlying target. Timing estimation for the tracks being detected is embedded inside the joint Viterbi detector using per-survivor processing. Performance results show that the proposed algorithm closely matches the performance of a fictitious system in which neighboring tracks are synchronous, and further that it significantly outperforms a previously reported detector that separately detects the two tracks.