[1]张银行,胡庆生.应用于高速串行链路的噪声预测部分响应最大似然均衡[J].东南大学学报(自然科学版),2017,47(1):6-11.[doi:10.3969/j.issn.1001-0505.2017.01.002]
 Zhang Yinhang,Hu Qingsheng.Noise prediction partial response maximum likelihood equalization for high speed serial link[J].Journal of Southeast University (Natural Science Edition),2017,47(1):6-11.[doi:10.3969/j.issn.1001-0505.2017.01.002]
点击复制

应用于高速串行链路的噪声预测部分响应最大似然均衡()
分享到:

《东南大学学报(自然科学版)》[ISSN:1001-0505/CN:32-1178/N]

卷:
47
期数:
2017年第1期
页码:
6-11
栏目:
信息与通信工程
出版日期:
2017-01-18

文章信息/Info

Title:
Noise prediction partial response maximum likelihood equalization for high speed serial link
作者:
张银行胡庆生
东南大学射频与光电集成电路研究所, 南京 210096
Author(s):
Zhang Yinhang Hu Qingsheng
Institute of RF- & OE-ICs, Southeast University, Nanjing 210096, China
关键词:
信号完整性 噪声预测 硬件复杂度 串行链路
Keywords:
signal integrity noise prediction hardware complexity serial link
分类号:
TN919
DOI:
10.3969/j.issn.1001-0505.2017.01.002
摘要:
为了解决25 Gbit/s及以上的高速串行信号在背板信道传输中的信号完整性问题,研究了能够在高速率和低误码率(BER)之间取得折中效果的部分响应最大似然(PRML)均衡.基于噪声预测法,推导出能够有效降低硬件开销的噪声预测PRML均衡.在具有不同衰减特性的背板信道和不同调制方式下,采用判决反馈均衡(DFE)、PRML均衡和噪声预测PRML均衡3种均衡方式对误码率性能和硬件复杂度进行了研究分析.结果表明,对于25 Gbit/s 的非归零(NRZ)信号,噪声预测PRML均衡的误码率性能优于DFE和PRML均衡;而对于12.5 Gbit/s 的四电平脉冲振幅调制(4PAM)信号,PRML均衡的误码率性能优于DFE均衡;噪声预测PRML均衡的硬件复杂度优于PRML均衡,更适合应用于高速串行链路中.
Abstract:
To solve the problem of signal integrity of high speed serial signal transmission in the backplane channels at or above 25 Gbit/s, the partial response maximum likelihood(PRML)equalization scheme,which can achieve a tradeoff between the high speed and the low bit error rate(BER), is studied. The noise prediction PRML equalization, which can effectively reduce the hardware cost, is derived based on the noise prediction method. The BER performance and the hardware complexity are analyzed by the decision feedback equalization(DFE), the PRML equalization and the noise prediction PRML equalization, respectively, in the backplane channels with different attenuation characteristics and on the different modulation modes. The results show that for the 25 Gbit/s non return zero(NRZ)signal, the BER performance of the noise prediction PRML equalization is better than those of the DFE and the PRML equalization. For the 12.5 Gbit/s 4 level pulse amplitude modulation(4PAM)signal, the PRML equalization can obtain better BER performance than the DFE. The hardware complexity of the noise prediction PRML equalization is better than that of the PRML equalization, making it more suitable for high speed serial link.

参考文献/References:

[1] Momtaz A, Green M. An 80 MW 40 Gb/s 7-tap T/2-spaced feed-forward equalizer in 65 nm CMOS [J]. IEEE Journal of Solid-State Circuits, 2010, 45(3): 629-639. DOI:10.1109/jssc.2009.2039268.
[2] Ali T, Rao L, Singh U. A 3.8 mW/Gbps quad-channel 8.5-13 Gbps serial link with a 5-tap DFE and a 4-tap transmit FFE in 28 nm CMOS[C]//2015 Symposium on VLSI Circuits. Kyoto, Japanese, 2015: 348-349. DOI:10.1109/vlsic.2015.7231318.
[3] Francese A P, Toifl T, Braendli M, et al. 10.6 continuous-time linear equalization with programmable active-peaking transistor arrays in a 14nm FinFET 2mW/Gb/s 16Gb/s 2-Tap speculative DFE receiver[C]//2015 IEEE International Solid-State Circuits. San Francisco, California, USA, 2015: 1-3. DOI:10.1109/isscc.2015.7062988.
[4] Zhang B, Khanoyan K, Hatamkhani H, et al. A 28 Gb/s multistandard serial link transceiver for backplane applications in 28 nm CMOS[J]. IEEE Journal of Solid-State Circuits, 2015, 50(12): 3089-3100. DOI:10.1109/jssc.2015.2475180.
[5] Alaa R, AL T, Yuan F, et al. Minimum jitter adaptive decision feedback equalizer for 4PAM serial links[C]//IEEE International Circuits and Systems. Lisbon, Germany, 2015: 2868-2871. DOI:10.1109/iscas.2015.7169285.
[6] Bulzaccheli F J. Equalization for electrical links: Current design techniques and future directions [J]. IEEE Journal of Solid-State Circuits, 2015, 7(4): 23-31. DOI:10.1109/mssc.2015.2475996.
[7] Toifl T, Buchmann P, Beukema T, et al. A 3.5pJ/bit 8-tap-feed-forward 8-tap-decision feedback digital equalizer for 16Gb/s I/Os[C]//European Solid State Circuits Conference. Venice, Lido, 2014: 455-458. DOI:10.1109/esscirc.2014.6942120.
[8] Parikh S, Kao T, Hidaka Y, et al. A 32Gb/s wireline receiver with a low-frequency equalizer, CTLE and 2-tap DFE in 28nm CMOS [C]//IEEE International Solid-State Circuits Conference Digest of Technical Papers. San Francisco, California, USA, 2013: 28-29.
[9] Eleftheriou E, Ocler S, Hutchins R A. Adaptive noise-predictive maximum likelihood(NPML)data detection for magnetic tape storage systems[J]. IBM Journal of Research and Development, 2010, 54(2): 7-1-7-10. DOI:10.1147/jrd.2010.2041034.
[10] Olcer S, Hutchins R A. NPML detection employing IIR noise-predictive with applications to magnetic tape storage[C]//2011 IEEE Global Telecommunications Conference. Houston, Texas, USA, 2011: 1-6. DOI:10.1109/glocom.2011.6133811.
[11] Arslan S S, Lee J, Goker T. Error event corrections using list-NPML decoding and error detection codes[J]. IEEE Transactions on Magnetics, 2013, 49(7): 3775-3778. DOI:10.1109/tmag.2013.2249658.
[12] Megha S. Strada whisper measured backplane channel [EB/OL].(2011-05-12)[2015-08-01]. http://www.ieee802. org/3/100GCU/public/channel.html.
[13] Veigel T, Alpert T, Lang F, et al. A Viterbi equalizer chip for 40 Gb/s optical communication links[C]//European Microwave Integrated Circuits Conference. Nuremberg, Germany, 2013: 49-52.

相似文献/References:

[1]宋晓东,李奇,吴定俊.一种轨道交通高架线路噪声衰减规律的预测方法[J].东南大学学报(自然科学版),2017,47(5):1055.[doi:10.3969/j.issn.1001-0505.2017.05.034]
 Song Xiaodong,Li Qi,Wu Dingjun.Prediction method for noise attenuation in elevated rail transit lines[J].Journal of Southeast University (Natural Science Edition),2017,47(1):1055.[doi:10.3969/j.issn.1001-0505.2017.05.034]

备注/Memo

备注/Memo:
收稿日期: 2016-05-27.
作者简介: 张银行(1982—), 男, 博士生; 胡庆生(联系人), 女, 博士, 教授, 博士生导师, qshu@seu.edu.cn.
基金项目: 国家自然科学基金资助项目(61471119).
引用本文: 张银行,胡庆生.应用于高速串行链路的噪声预测部分响应最大似然均衡[J].东南大学学报(自然科学版),2017,47(1):6-11. DOI:10.3969/j.issn.1001-0505.2017.01.002.
更新日期/Last Update: 2017-01-20