本篇論文中，主要使用無載波振幅相位調變來當作訊號的主要調變方式，並
且還會介紹無載波振幅相位調變與正交多頻分工相比所具有的優勢，接著將會使
用兩種濾波器以及不同的資料格式來調變訊號，兩種濾波器分為濾波器1 號以及
濾波器2 號。濾波器1 號為使用正餘弦與根升餘弦函數相乘後的函數作為濾波
器，此濾波器具有不隨根餘弦函數中的滾降係數升高而產生碼際間干擾的優勢，
濾波器2 號為直接將根升餘弦函數做希爾伯特轉換來做為濾波器使用，此種濾波
器具有較高的頻譜利用率，但是會隨滾降係數提高而產生碼際間干擾，接著使用
伏爾泰拉等化器以及選擇性反饋式等化器來比較出兩種濾波器的複雜度，之後進
行接收端的解調。
首先會以傳輸距離0 公尺的數據來對兩種濾波器做比較，則比較出在符元速
率低時濾波器1 號具有較大的優勢，而在符元速率提高時則是濾波器2 號具有優
勢，經過傳輸後能在200 公尺以及500 公尺的部分能夠發現到，在資料格式為16
得到的最高資料傳輸量是由濾波器2 號所達成，而在資料格式為32 時是由濾波
器1 號所達成，在傳輸後因濾波器2 號在符元速率提高時所具有的優勢而得到最
高資料傳輸量皆由濾波器2 號所達成，之後能夠以此討論來得到這兩種濾波器在
未來的可行性。

This dissertation studies the application of carrier-less amplitude phase modulation
(CAP) in short-reach MMF transmission. In addition to two different modulation orders
(i.e., CAP-16 and CAP-32), we adopted two kinds of filters in CAP modulation, which
are named by Filter 1 and Filter 2. The frequency responses of two data paths in Filter
1 are root raised cosine functions multiplied by sine and cosine, and this scheme is not
only inter symbol interference- (ISI-) free but also maintains orthogonality between two
paths. As to Filter 2, the root raised cosine function and its Hilbert transform are directly
used as the frequency responses of two data paths. Although higher spectral efficiency
can be achieved by Filter 2, the orthogonality between two paths does not hold unless
the roll-off factor is zero. We used the Volterra and decision feedback equalizer (DFE)
to equalize the CAP signals, and the corresponding improvement can demonstrate the
characteristics of the used filters.
First of all, we experimentally compared the CAP signals based on the two filters
at optical back-to-back. When the symbol rate is low, Filter 1 can outperform due to the
orthogonality. However, when the symbol rate increases, Filter 2 can perform better due
to its better spectral efficiency. In particular, when the system bandwidth was reduced
after 200- or 500-m MMF transmission, the highest data rate of CAP-16 was achieved
by Filter 2. Nonetheless, if CAP-32 was used, the highest data rate was accomplished
by Filter 1 because symbol rate was low to meet the requirement of higher SNR. Since
CAP-16 can achieve higher data rate compared to CAP-32, this work demonstrates that
Filter 2 is a potential candidate for future high-capacity short-reach MMF transmission.

論文審定書..................................................................................................................... i
致謝................................................................................................................................ ii
摘要............................................................................................................................... iii
Abstract ......................................................................................................................... iv
圖次............................................................................................................................... vi
第一章 緒論.................................................................................................................. 1
1-1 前言 ................................................................................................................. 1
1-2 研究動機 ......................................................................................................... 2
1-2-1 短距離傳輸系統 (Short reach, SR) .................................................... 2
1-2-2 IEEE 400G 乙太網路通訊協定 .......................................................... 4
1-2-3 主動式光纖纜線系統 (Active optical cable, AOC) ........................... 6
第二章 數位訊號調變技術.......................................................................................... 7
2-1 正交分頻多工調變.................................................................................... 7
2-1-1 正交分頻多工簡介 .............................................................................. 7
2-1-2 正交分頻多工原理 .............................................................................. 7
2-1-3 正交多頻分工優缺點 .......................................................................... 8
2-2 無載波振幅相位調變.............................................................................. 10
2-2-1 無載波振幅相位調變簡介 ................................................................ 10
2-2-2 希爾伯特轉換 (Hilbert Transform) .................................................. 11
2-2-3 無載波振幅相位調變以及濾波器之差異 ........................................ 12
第三章 數位訊號處理技術........................................................................................ 19
3-1 前饋式等化器 ............................................................................................... 19
3-2 選擇性反饋式等化器 ................................................................................... 20
3-3 伏爾泰拉等化器 ........................................................................................... 21
第四章 實驗架構與實驗結果.................................................................................... 23
4-1 實驗架構 ....................................................................................................... 23
4-2 實驗結果 ....................................................................................................... 28
第五章 結論................................................................................................................ 49
5-1 總結與討論 ................................................................................................... 49
參考文獻...................................................................................................................... 50

[1] sagoon, https://blog.sagoon.com/how-internet-has-changed-communication-overthe-
time/
[2] T.P. Morgan, (2016) “Construction zones on the ethernet roadmap”, March 24.
[3] A. Liu, (2015) “Direct Attach Copper Cable VS Active Optical Cable (AOC)”,
August 15.
[4] IEEE 802.3 Ethernet, http://www.ieee802.org/3/bs/
[5] R.W. Chang (1966).“Synthesis of Band‐Limited Orthogonal Signals for
Multichannel Data Transmission.” Bell System Technical Journal, 45(10), 1775-1796.
[6] Y.H. Wang. (2016) “ IM/DD Single-Sideband OFDM Long-Distance Transmission
Using Cascaded DML/EAM”, Department of Photonics National Sun Yat-sen
University Master Thesis.
[7] A short tutorial on the significance of cyclic prefix in OFDM systems
[8]Y.S. Huang(2014), “Nonlinear Volterra Filter in Visible Light Communication” ,
Department of Photonics National Sun Yat-sen University Master Thesis.
[9] wiki Hilbert transform , https://en.wikipedia.org/wiki/Hilbert_transform
[10] J. J. Werner,(1992) “Tutorial on Carrierless AM/PM - Part 1 - Fundamentals and
digital CAP transmitter,” Document for ANSI X3T9.5 TP/PMD, St. Petersburg Beach.
[11] J.J. Werner, (1993)“Tutorial on Carrierless AM/PM - Part 2 - Performance of
bandwidth-efficient line codes,” Document for ANSI X3T9.5 TP/PMD, Austin.
[12] A. Shalash and K. K. Parhi, (1996) “Comparison of discrete multitone and
carrierless AM/PM techniques for line equalization,” IEEE International Symposium
on Circuits and Systems, vol. 2, pp. 560-563.
[13] Diniz P. (2008)Adaptive Filtering 3ed., Springer.
[14]K.C. Chen(2015), “Application of Carrier-Less Amplitude-Phase Modulation in
Short-Reach Optical Interconnection and Long-Reach PONs Supplemented by Digital
Signal Processing”, Department of Photonics National Sun Yat-sen University Master
Thesis.
[15] D. A. George, R. R. Bowen, and J. R. Storey. (June 1971)“An adaptive decision
feedback equalizer,” IEEE Transactions on Communication Technology, vol. com-19,
pp. 281-93.
[16] C. Kaiyun, C. Gang, H. Qunfeng, and X. Zhengyuan. (2009) “Indoor optical
wireless communication by ultraviolet and visible light,” Proceedings of the SPIE -
The International Society for Optical Engineering, vol. 7464, pp. 74640D (9 pp.)-
74640D (9 pp.).
[17] G. P. Agrawal, Nonlinear Fiber Optics, the 3rd edition
[18] J.D. Yang. (2016) “ Application of Duobinary & PAM-4 Signal in Short-Reach
Optical Interconnection with 850-nm VCSELs and Multimode Fiber”, Department of
Photonics National Sun Yat-sen University Master Thesis.