Constrained Code and Partial-Response Maximum-Likelihood Detection for High Density Multi-Level Optical Recording Channels

Abstract

A new efficient four-ary ð1; 4Þ run-length limited (RLL) code is presented, which is intended for high density multi-level optical recording channels. By using proposed hybrid DC control scheme, the new code can achieve acceptable DC suppressing characteristics while maintaining high code rate. The bit error rate (BER) performances of this new code on blue laser recording systems are evaluated by using partial-response maximum-likelihood (PRML) detection. The feasibility of using four-ary RLL modulation and PRML detection is studied in this paper, and these techniques show great potential to be applicable in high-density multi-level optical recording channels.

Theory and practice of multilevel method in high-definition discs

“…The capacity of single disc is reduced to 8.5GB. If VML-RLL (1,9) codes are adopted in DVD structure, because of d = 1, each channel-bit length is T = L / (1 + d) = L/2, it is 1.5 times that of DVD. Therefore, it has enough margins to compress channel density (the minimum symbol size) with no change of the manufacturing process and read-out system.…”

“…Table 7 is the comparison of some minimum symbol size after compressed. It is easy to see from the table, when d = 1, without taking into account ISI of the minimum symbol, even if the minimum symbol is compressed to 4/5 of that of DVD 5, the method of VML-RLL (1,9) encoding still has 10% allowance compared to the jitter standard of the single-sided double-layer structure of DVD 9. Therefore, both of DVD 5 and DVD 9 can use the same minimum symbol size if VML-RLL (1,9) coding is adopted.…”

“…It is easy to see from the table, when d = 1, without taking into account ISI of the minimum symbol, even if the minimum symbol is compressed to 4/5 of that of DVD 5, the method of VML-RLL (1,9) encoding still has 10% allowance compared to the jitter standard of the single-sided double-layer structure of DVD 9. Therefore, both of DVD 5 and DVD 9 can use the same minimum symbol size if VML-RLL (1,9) coding is adopted. If the minimum symbol length is Lc=8/9 after compression, DR of VML-RLL is 2, its capacity is 2/1.5*(9/8)=1.5 times that of DVD, so single-sided single-layer capacity is 4.7GB*1.5=7.05GB; single-sided double-layer no longer need to change the minimum symbol length, so its capacity is twice that of single-sided single-layer, that is 14.1GB.…”

“…Table 8 is the comparison of capacity after compressed. [9] , 2T nonlinear compensation, and so on. Figure 3a ~ figure 8a show the RF signal of different bands on a same experimental disc.…”

“…Figure 3b ~ figure 8b are the statistical distribution of run-length recognized by slicer. The Figure 3 is the DVD signals of DK (2,10), Figure 4 ~ figure 8 are the RF signals of DK (1,9), of which the minimum symbol size are respectively compressed to 8/9, 7/8, 6/7,5/6,4/5 of that of DVD 5. The sampling accuracy of statistical run-length is 1/12T.…”