KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 3 of 3
  • Some of the metrics are blocked by your 
    Item type:Publication,
    An MLP-Based ITI Suppression Method for Multi-Head Multi-Track Bit-Patterned Magnetic Recording
    (2025-01-01)
    Koonkarnkhai, Santi
    ;
    Kovintavewat, Piya
    ;
    Warisarn, Chanon
    Inter-track interference (ITI) is a critical challenge in bit-patterned magnetic recording (BPMR) systems, particularly at high areal densities (ADs) where reduced bit period and track pitch lead to severe interference. This article introduces a novel multi-layer perceptron (MLP)-based ITI suppression method for the three-head three-track (3H3T) BPMR system. Our approach uses an MLP to estimate the main track data and uses iterative decoding to generate soft information, which will then be used to reconstruct and remove ITI for the adjacent tracks. At an AD of 3 Tb/in2 and a bit-error rate (BER) of 10<sup>-5</sup> , simulation results show that the proposed system achieves performance gains of 1 and 6.5 dB compared with the previously proposed 3H3T system and the conventional system with one-head one-track detection, respectively. In addition, our system demonstrates robust performance under challenging conditions, maintaining effectiveness with track mis-registration (TMR) up to 10% and media noise up to 5%. These results indicate that the proposed method can be considered as one of the promising solutions for ultrahigh-density BPMR systems.
  • Some of the metrics are blocked by your 
    Item type:Publication,
    An Iterative Two-Head Two-Track Detection Method for Staggered Bit-Patterned Magnetic Recording Systems
    (2019-07-01)
    Koonkarnkhai, Santi
    ;
    Warisarn, Chanon
    ;
    Kovintavewat, Piya
    One of the major problems in bit-patterned magnetic recording (BPMR) systems is an inter-track interference (ITI) that can significantly degenerate the system performance, especially at high areal densities (ADs). Generally, the BPMR system with staggered islands experiences less ITI effect than that with regular islands. Therefore, we propose an iterative two-head two-track detection method for a staggered BPMR system to further reduce the ITI effect, which employs two read heads to read two adjacent tracks and decodes them simultaneously. To lessen the ITI, the weighted readback signal of one track, serving as an ITI signal, is subtracted from the readback signal of the other track, and vice versa, before passing the refined readback signal to an iterative decoding process. Hence, the decoded data will be utilized to estimate the ITI signal for the next global iteration, NG. Supposing that the refined readback signal has only one-side ITI effect. We also present the design of the 2× 3 target and its compatible equalizer that will be employed in the proposed system. Results demonstrate that the proposed system not only is superior to the conventional system but also has low complexity. Specifically, the proposed system can perform close to the conventional system whose readback signal encounters only one-side ITI effect when NG = 2, 3 and 5 iterations for AD = 2, 2.5, and 3 Tb/in<sup>2</sup>, respectively.
  • Some of the metrics are blocked by your 
    Item type:Publication,
    A simple parity check matrix LDPC code for perpendicular magnetic recording channels
    (2013-09-02)
    Timakul, Sekson
    ;
    Choochuay, Somsak
    In LDPC code design, the structure of code's parity check matrix can be very crucial to code performance. In this paper, we have investigated a simple designed parity check matrix when the code is applied to a PMR channel of a storage system. Belief propagation algorithm and min-sum algorithm are two major decoding algorithms used to benchmarked the code's merits. Additive white Gaussian noise channel (AWGN) and perpendicular magnetic recording (PMR) channel with SISO equalizer in PR2 target are the conditions of our investigation. The LDPC code itself is a modified array type. Its construction fairly simple compared to normal array and quasi-cyclic code. We used two decision variables to define the permutation matrix. A method proposed by Tanner's et.al had been adapted to the modified array code. The decoders have been tested for 5 and 20. iterations. SNR versus BER and BLER were measured. © 2013 IEEE.