Reduced-Complexity Per-Survivor Iterative Timing Recovery for Coded Partial Response Channels

Abstract

A (full-complexity) per-survivor iterative timing recovery scheme, which jointly performs timing recovery, equalization, and error-correction decoding, was recently proposed [1] to deal with the problem of timing recovery operating at low signal-to-noise ratio. Although it outperforms other iterative timing recovery schemes, it has very high complexity. In this paper, we propose a reduced-complexity per-survivor iterative timing recovery scheme to make it implementable in real-life applications. Simulation results indicate that for low to moderate complexity, the reducedcomplexity scheme provides a better performance than the full-complexity scheme.

Robustness of per-survivor iterative timing recovery in perpendicular recording channels

“…Then, the proposed scheme iteratively exchanges soft information between PSP-SOVA and a soft-in soft-out (SISO) decoder. As studied in [2], the proposed scheme outperforms the conventional receiver, especially when the timing error is severe. In practice, noise in magnetic recording channels is also data-dependent [5], whose severity depends on the data pattern written on the disk.…”

“…PSP-SOVA-PDNP works in a same manner as PSP-SOVA does [2], except that we replace the branch metric calculation in (A-6) [2] with (4), and keep track of the predicted noisen k (u, v) of each transition. Clearly, PSP-SOVA-PDNP has high complexity because it requires trellis expansion.…”

“…For other mathematical functions, e.g., log(x), exp(x), etc., we assume they are implemented as lookup tables, and that we ignore their complexity. It can be shown that PSP-SOVA-PDNP has (4N s + N e + 4L + 7)Q ( p) + δ 2 +9δ+9 2 +1 additions and (N s +N e +2L+11)Q ( p)+1 multiplications, whereas PSP-SOVA-PDNP-MO has additions and multiplications, where δ = 5(v + 1) is a SOVA decoding depth [3], Q = 2 v is the number of trellis states, N s is the number of interpolation taps used to refine the samples [2], and N e is the number of equalizer taps. It is apparent that PSP-SOVA-PDNP-MO has lower complexity than PSP-SOVA-PDNP.…”

“…Per-survivor iterative timing recovery was proposed in [2] to deal with the problem of timing recovery operating at low signal-to-noise ratio (SNR). It is realized by first developing a per-survivor softoutput Viterbi algorithm (SOVA) [3], denoted as "PSP-SOVA," [2] by embedding the timing recovery step inside the SOVA equalizer using per-survivor processing (PSP) [4], a technique of jointly estimating a data sequence and unknown parameters.…”

“…It is realized by first developing a per-survivor softoutput Viterbi algorithm (SOVA) [3], denoted as "PSP-SOVA," [2] by embedding the timing recovery step inside the SOVA equalizer using per-survivor processing (PSP) [4], a technique of jointly estimating a data sequence and unknown parameters. Then, the proposed scheme iteratively exchanges soft information between PSP-SOVA and a soft-in soft-out (SISO) decoder.…”