短程光通信的数字信号处理：技术创新与未来展望
需积分: 5 80 浏览量
更新于2024-07-09
收藏 2.08MB PDF 举报
"短程光通信的数字信号处理：当前技术和未来趋势的回顾"
这篇研究论文深入探讨了短程光通信领域中的数字信号处理技术。随着云计算服务和数据中心应用的飞速发展，短程光通信已经成为一个关键的市场细分领域，并在近年来吸引了大量的研究关注。在这些系统中，直接检测型接收器配置以及低成本、小型化组件是其主要特征，但同时也引入了一些在相干通信系统中不常见的传输损耗问题。
创新的信号调制和数字信号处理（DSP）在这个领域中扮演着至关重要的角色。DSP技术能够挖掘这些低成本组件的潜力，克服传输过程中产生的各种损耗，以满足日益增长的数据速率需求。文章可能涵盖了以下关键知识点：
1. **短程光通信系统**：这些系统通常用于数据中心内部或数据中心之间的高速数据传输，要求低功耗、低成本和高效率。
2. **直接检测技术**：与相干检测不同，直接检测是一种简单且成本较低的光信号接收方法，它依赖于光强度的变化来获取信息，但对噪声和色散等光学效应敏感。
3. **传输损耗**：在短程光通信中，由于采用低成本组件，可能会面临诸如模态色散、偏振模色散、非线性效应和光纤损耗等问题。
4. **数字信号处理**：通过先进的算法，如均衡器、前向误差校正（FEC）、扩频调制和自适应滤波，DSP可以补偿这些传输损耗，提高系统的误码率性能和数据传输速率。
5. **未来趋势**：论文可能讨论了如何利用机器学习和人工智能技术优化DSP，以适应更高的数据速率需求，以及新兴的技术如硅光子学和量子通信对短程光通信的影响。
6. **市场和研究方向**：随着5G网络、物联网(IoT)和大数据中心的扩张，短程光通信的研究重点可能会转向更高效、更绿色的解决方案。
7. **标准化和许可**：文中提到的IEEE标准和版权信息提示，该领域的研究和实践需要遵循特定的行业规范和知识产权规定。
这篇论文将为读者提供关于短程光通信领域中数字信号处理的最新进展、挑战和未来展望的全面概述，对于理解这个领域的核心技术和发展方向具有很高的参考价值。
0733-8724 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/JLT.2018.2793881, Journal of
Lightwave Technology
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
5
(DD-FTN) algorithm to overcome the drawbacks of
conventional equalizers [14]. We will discuss these algorithms
in more details in the next section.
CAP: An alternative intensity modulation format that can
provide high order modulation is CAP modulation that enables
the generation of QAM-type signals but with simpler
implementation. In this case, baseband QAM signal with a
frequency shift to a subcarrier is required. CAP uses FIR filter
to achieve this up-conversion and modulate QAM signal to a
real signal or demodulate QAM signal from a received real
signal. Although the shaping and matched filters in CAP
system can be implemented using analog techniques [15], we
focus on digital implementation since it is the preferred choice
given today’s DAC and ADC technologies [16].
The DSP structure of CAP is also shown in Fig. 4. At the
transmitter side, the in-phase (I) and quadrature (Q)
components of the QAM signal are sent into two shaping filters
with impulse responses
( ) ( ) cos(2 )
Ic
h t g t f t
and
( ) ( ) sin(2 )
Qc
h t g t f t
that form a Hilbert transform pair.
A square-root-raised-cosine shaping filter
()gt
with a roll-off
coefficient of
is used as the baseband impulse response. The
center frequency
c
f
is given by
(1 ) /2Bf
where
f
is the
frequency offset in order to avoid spectrum aliasing. Fig. 5(a)
shows the time domain response of the shaping filter. Fig. 5(b)
depicts the baseband QAM signal. After the shaping filter, the
electrical spectrum of generated CAP signal is shown in Fig.
5(c). The signal is shaped to be a Nyquist-like signal and
up-converted to an intermediate frequency (IF) of f
c
. After the
shaping filter, pre-emphasis is employed to compensate for the
bandwidth limitation of DAC and transmitter laser nonlinearity
with a time domain FIR filter. The processed signal is then fed
into a DAC to generate the electrical drive signal. Fig. 5(d)
shows the optical spectrum of optical CAP signal. The optical
signal after transmission would be directly detected by PIN PD
or APD, sampled and sent into two matched filters. The two
matched filters are the time-reversed version of the shaping
filters at the transmitter i.e.
( ) ( ) cos(2 )
Ic
m t g t f t
and
( ) ( ) sin(2 )
Qc
m t g t f t
in order to separate the I and Q
components. The separated I and Q signals are then combined
to re-construct the complex QAM signal in digital domain
which is followed by a re-sampling function. As for complex
high-order QAM signals, common adaptive equalization
algorithms such as constant multi modulus algorithm (CMMA)
[17], multi constant multi modulus algorithm (MCMMA) [18]
or decision directed least-mean-square (DD-LMS) can be used
to compensate various channel impairments (DD-LMS will be
discussed in more detail in the next section).
Fig. 6. Bit- or-power-loading for DMT transmission system. The blue curve is
the estimated SNR as a function of frequency and different amount of signal
powers (represented by the height of the green rectangles) and modulation
format can be encoded in each sub-carrier.
DMT: Also known as direct detected orthogonal frequency
division multiplexing (OFDM), DMT is another attractive
modulation scheme for low cost short-reach systems
[19][20][21]. In contrast with OFDM, only half of the
subcarriers in DMT are coded with information while the other
half are the complex conjugate of the first half to ensure that the
DMT signal is real for intensity modulation in IM/DD systems.
Like all multi-carrier systems, DMT enables bit loading or
power loading [22] that allow a flexible set of modulation
formats for each subcarrier optimized with respect to the
channel transfer function to maximize overall bit rate or power
margin. DMT has advantages such as high spectral efficiency,
high tolerance to impairments and flexible coding. The most
popular bit loading method is Chow’s algorithm [23] and Fig. 6
shows the principle of bit loading. Before signal transmission, a
probe signal would be transmitted and received to estimate the
end-to-end SNR at each frequency sub-carrier as shown by the
blue curve. In this case, high order modulation formats such as
128-QAM, 64-QAM can be coded in the subcarriers at low
frequency which have high SNR. 32-QAM, 16-QAM and
8-QAM can be coded in the subcarriers at the mid-frequencies
and QPSK or BPSK is coded in the subcarriers at high
frequency part to achieve similar bit error ratio performance
across all subcarriers. In this case, the bandwidth usage is
maximized. The DSP configuration for DMT systems is also
shown in Fig. 4. After FEC encoding, the S/P function transfers
Fig. 5. (a) Response of shaping filter (b) Electrical spectrum of baseband signal
(c) electrical spectrum of electrical CAP signal (d) optical spectrum of CAP
signal.
剩余22页未读，继续阅读
点击了解资源详情
点击了解资源详情
点击了解资源详情
2021-09-17 上传
2022-01-04 上传
2021-02-25 上传
2021-08-16 上传
2020-03-20 上传
2014-04-18 上传
weixin_38706824
- 粉丝: 2
- 资源: 892
我的内容管理
展开
