Enhancing Areal Density Based on Signal Processing Techniques in Staggered Dual-Layer Bit-Patterned Magnetic Recording Systems
Natthakan Rueangnetr, Satra Tor. Wattanaphol, Santi Koonkarnkhai, Simon John Greaves, Chanon WarisarnDual-layer bit-patterned magnetic recording (DL-BPMR) is a promising approach to extend areal-density scaling by exploiting the vertical dimension; however, reliable retrieval is challenged by the inherently weaker lower-layer readback and additional interference mechanisms, inter-symbol interference, inter-track interference, and inter-layer interference, that distort the mixed readback signal. This study presents a co-designed dual-layer BPMR scheme with an areal density of 2.0 to 6.0 Terabits per square inch (Tbits/in2), based on two complementary approaches. First, we vary the head–media spacing (HMS) to reduce the distance between the read head and the recording layers, especially the lower layer, strengthening the lower-layer contribution and making the mixed readback clearer and more reliable. Second, conditioned-HMS readback is processed using a combined-processing receiver tailored to a staggered (down-track-offset) dual-layer layout, where oversampling enables a unified one-dimensional sequence representation for dual-layer detection without enforcing early layer separation. To further improve detection efficiency, we also introduce a two-step minimum mean square error design that first determines a short monic generalized partial response target and then computes a minimum mean square error equalizer matched to the resulting readback, producing an equalized sequence suitable for soft-output sequence detection and low-density parity-check coded decoding. The simulation results indicated that reducing the head-media spacing improves readback separability (i.e., reliability) and that the proposed two-step combined-processing partial-response maximum-likelihood method achieves a consistently lower mean-squared error than conventional dual-layer processing across signal-to-noise ratios (SNRs). More importantly, in end-to-end bit-error-rate (BER) evaluation under electronic noise, e.g., the proposed dual-layer BPMR system delivers a clear reliability advantage over a single-layer BPMR baseline, achieving SNR gains of 4.2 dB at BER = 10−4 and 3.7 dB at BER = 10−5 in the favorable head-media spacing regime, thereby demonstrating robust high-density dual-layer retrieval beyond the single-layer baseline.