KMITL

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

Browse

Settings

Access type: UnknownSubject: search.filters.subject.Viterbi detector

Search Results

Now showing 1 - 4 of 4
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Adaptive Designing Process of 2-D GPR Target and Equalizer based on BER in BPMR Systems
    (2020-06-01)
    Rueangnetr, Natthakan
    ;
    Warisarn, Chanon
    ;
    Myint, Lin M.M.
    In high areal density magnetic recording systems, two-dimensional (2-D) equalizer and 2-D generalized partial response (GPR) target are proposed to tackle the 2-D interference problem. However, the minimum mean squared error (MMSE) technique cannot provide the optimality of the 2D detector, particularly only partial side-track information are available in the designing process. Therefore, we develop an adaptive designing process to obtain the optimal equalizers and targets based on bit error rate (BER) for two-track two-head (2T2H) bit-pattern media recording (BPMR) read channel model. In the process, the coefficients of the 2-D equalizer and GPR target are computed using the MMSE technique first. Then these MMSE targets and equalizers are adaptively adjusted based on BER at the low SNR levels. In each iteration, the main side-track coefficients of the 2×3 GPR targets are adjusted based on the BER. The iterative process is continually operated until there will be enough low level of error. Finally, the performance of those equalizers and targets are evaluated under the 2T2H BPMR system.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Study of Detectors with Noise Predictor for High Areal Density BPMR Systems
    (2019-04-10)
    Myint, Lin Min Min
    ;
    Supnithi, Pornchai
    In this paper, we study the performance of Viterbi detectors embedded with noise predictor for the 2-D interference channel of high areal density bit patterned media recording (BPMR) system. After the readback signal is processed by a equalizer, the received signal at the input of detector is corrupted by the colored/correlated noise. Therefore, the noise whitening process using a noise predictor is required to improve the performance of the detector. When the noise predictor is incorporated into the detector, it needs to extend the structure of trellis in the detector. For the 2-D interference channel of BPMR system, the complexity of the detector will grow beyond the practical level. Therefore, we design a detector embedded by a noise prediction technique without extending the trellis. In the proposed technique, the most likely noise samples generated from the branch metric calculation are stored along the surviving path of the trellis in Viterbi algorithm. The proposed technique requires less number of memory and arithmetic operations at each state of the trellis compared to conventional technique. We consider both finite impulse response (FIR) filter and infinite impulse response (IIR) filter for the noise prediction filter. The simulation results show that the proposed noise prediction can improve the performance of detector, especially for the high areal density at 4 Tbit/in<sup>2</sup>.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    Sidetrack data estimation using multi-track joint 2-D viterbi detector in bit patterned media storage
    (2018-07-02)
    Myint, Lin Min Min
    ;
    Supnithi, Pornchai
    In the high areal density, bit patterned media recording (BPMR) system, the inter-track interference (ITI) effects become severe when the areal density is increased. In the literature, a multi-track joint 2-D detector with the aid of a multi-reader array technology is proposed to mitigate the ITI effect in BPMR systems because of its significant performance gains achieved by processing the multiple readback signals concurrently. In practice, the multi-track joint 2-D detector can estimate the recorded data from not only the detecting tracks, but also the sidetracks when the ITI effect is considerably high. Although the estimates of the data on the sidetracks are not as good as those on the detecting tracks, they can be applied in or improved by further coding process. In this work, we study the performance of a multi-track joint 2-D Viterbi detector for estimating the data on the sidetracks by adjusting the offset position of heads in the multi-head multi-track (MHMT) BPMR system. The detector recovers the recorded data on the detecting and sidetracks by processing the readback signals from the detecting tracks. We also investigate the performance changes due to the head offsets. The results show that the performance of detector for estimating the data on the sidetracks can be improved by adjusting the head offsets well.
  • Loading...
    Thumbnail Image
    Some of the metrics are blocked by your 
    consent settings
    Item type:Publication,
    A comparative study on reduced complexity 2-D detectors for bit patterned media
    (2011-12-01)
    Myint, Lin M.M.
    ;
    Supnithi, Pornchai
    In this paper, we compare the performance of the reduced complexity 2-D detection methods: Graph-based detection and Viterbi-based detection for the 2-D interference bit patterned media channels at various areal densities. In the graph-based detection, the complexity is reduced by ignoring the ITI from the cornered islands in designing the factor graph. In Viterbi-based detector, a trellis for the ISI only is considered. However, both methods improve the ITI mitigation with the help of multi-track processing and iterative processing. Between them, the graph-based detection needs higher computational complexity than the Viterbi-based detection. However, the graph-based detection method provides a significant performance gain over the Viterbi-based detect.on © 2011 IEEE.