Oral
03 Nov 2023
To support the exponential growth of data storage demand, bit-patterned magnetic recording (BPMR) [1] has been presented as a new alternated recording technology using the patterned isolated magnetic islands in place of granular media. The signal-to-noise ratio can be gained, so an areal density (AD) can be also increased up to 4.0 Tb/in2. To reach the expected AD; however, the distance between bit-island must be shrunk closer to each neighboring island, which means two-dimensional (2D) interference and media noise will be also dramatically increased. Therefore, both 2D modulation encoding techniques [2-4] and deep neural networks [4-5] were wildly utilized to cope with those problems under the multi-reader systems [6]. we first propose to use the multilayer perceptron (MLP) for refining the soft information under the rate-5/6 2D modulation coded three-track/three-head BPMR systems as illustrated in Fig. 1, which has been continuously developed from prior soft-information flipper based on long-short term memory (LSTM) networks [4]. The data patterns and input numbers of the soft information are carefully investigated. Moreover, to cope with the codeword border effect, we also propose the MLP-based array reader optimization method. Among all three coded tracks, the upper- and lower-most tracks are still suffered from neighboring track interference; therefore, we present the suitable reader positions of three readers in three cases as shown in Fig. 1. The moving distance is considered in the percentage of the track pitch. Then, their readback signals are sent to the read channel for processing. As shown in Fig. 2, the proposed MLP soft information flipper can provide better recording performance in terms of bit-error rate (BER), which can gain 1.5 dB at BER = 10-4 over the LSTM soft information flipper [4]. Moreover, results also show that an appropriate position of an array reader can significantly yield better BER performance utilizing the MLP-based array reader optimization method.References: [1] B. D. Terriset et al., Microsyst. 13(2), 189–196 (2006) [2] C. Warisarn, A. Arrayangkool, and P. Kovintavewat, IEICE Trans. Electron. E98.C, 528–533 (2015) [3] C. D. Nguyen and J. Lee, IEEE Trans. Magn. 53(3), 3101207 (2017) [4] N. Rueangnetr, L. M. Myint, and C. Warisarn1, AIP Advances 11, 015307 (2021) [5] A. Aboutaleb et al., Appl. Phys. Lett. 119, 010502 (2021) [6] G. Mathew et al., IEEE Trans. Magn., vol. 50, no. 3, pp. 155-161, (2014)
