Design of Low-Complexity 2-D SOVA Detector for Shingled Magnetic Recording

Abstract

Shingled magnetic recording emerges as a promising near-term option to sustain the historical areal density growth of hard-disk drives while retaining conventional heads and media. Highly scaled shingled magnetic recording is subject to severe intertrack interference, which naturally demands 2-D read channel signal processing. By concurrently detecting multitrack read-back signals from a read head array, joint 2-D signal detection can fully exploit the 2-D interference to maximize the detection performance at the penalty of very high computational complexity. As modern read channel employs iterative signal detection and decoding and hard-disk drives are very cost sensitive, it is highly desirable to effectively reduce the 2-D signal detection computational complexity and hence silicon cost at minimal detection performance loss. This paper presents two low-complexity 2-D soft-output Viterbi algorithm (SOVA) detector design strategies for realizing multitrack joint 2-D detection, which represent different complexity versus performance tradeoffs. By carrying out simulations and application-specific integrated circuit design, this paper shows that, compared with the best symbol-based 2-D SOVA, it is feasible to achieve almost the same detection performance and meanwhile reduce the silicon area by more than 60% in three-track joint detection.