Soft-Information Improvement using Multilayer Perceptron (MLP) in Bit-Patterned Magnetic Recording (BPMR) Systems

被引:0
|
作者
Praiwan, J. [1 ]
Warisarn, C. [1 ]
Suwansawang, S. [2 ]
Koonkarnkhai, S. [2 ]
机构
[1] King Mongkuts Inst Technol Ladkrabang, Coll Adv Mfg Innovat AMI, Bangkok 10520, Thailand
[2] Nakhon Pathom Rajabhat Univ, Adv Signal Proc Disrupt Innovat Res Ctr, Mueang Nakhon Pathom Dis 73000, Nakhon Pathom, Thailand
关键词
Multilayer perceptron (MLP); Soft-information; Bit-patterned magnetic recording (BPMR); Log-likelihood ratio (LLR); MEDIA;
D O I
10.1109/ITC-CSCC52171.2021.9501426
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The effects of both inter-track interference and inter-symbol interference are the main hindrances in alternated magnetic recording technologies such as bit-patterned magnetic recording. To increase an areal density of hard disk drives over 1 terabit per square inch; therefore, these two effects must be coped with some powered signal processing techniques. In this paper, we propose to use the multilayer perceptron, which is one of the artificial neural networks, to improve the soft information or log-likelihood ratios that are obtained from parallel soft-output Viterbi algorithm (SOVA) detectors. The computer-simulated results reveal that our proposed recording system can efficiently yield better bit error rate performance when it was compared with the previously proposed magnetic recording systems.
引用
收藏
页数:3
相关论文
共 36 条