非参数直方图估计优化IMDD-OQAM-OFDM系统的低分辨率性能
34 浏览量
更新于2024-08-26
收藏 2.19MB PDF 举报
"这篇研究论文探讨了在IMDD-OQAM-OFDM(直插式强度调制-直接检测正交频分复用)系统中，如何通过非参数直方图估计来优化低分辨率的性能。该研究发表于2018年8月的IEEE Photonics Journal，由Meihua Bi、Tiancheng Huang、Ling Liu、Xin Miao、Shilin Xiao和Weisheng Hu共同完成，并给出了DOI：10.1109/JPHOT.2018.2848248。文章强调了在光通信领域，尤其是在IMDD-OQAM-OFDM系统中，提高低分辨率接收机性能的重要性。
IMDD（直插式强度调制-直接检测）是一种常见的光通信技术，其中光信号的强度被用来直接携带信息，然后通过简单的光电探测器进行检测。这种方法相对于其他调制方式如相位或频率调制，具有成本低、实现简单的优点。然而，它也存在一些挑战，例如在高数据速率传输下容易受到噪声和色散的影响，导致性能下降。
OQAM-OFDM（正交频分复用）是无线通信和光通信中广泛采用的一种多载波调制技术，尤其适合宽带系统的频谱效率提升。OQAM使得实部和虚部数据流可以并行传输，从而提高信道利用率。然而，低分辨率的ADC（模数转换器）在处理OQAM-OFDM信号时可能会引入额外的量化噪声，降低系统性能。
非参数直方图估计是统计学中的一个方法，它不依赖于特定的概率分布假设，而是通过对观测数据的统计特性直接建模来估计未知分布。在本文中，研究者应用这一方法来估计和优化IMDD-OQAM-OFDM系统的接收机性能。通过非参数直方图估计，可以更准确地捕捉数据的特性，减少量化误差，从而改善系统在低分辨率下的性能。
作者们可能采用了不同的算法和策略来实现这一目标，比如改进的量化策略、自适应阈值选择或优化的数据预处理步骤。这些技术有助于在不增加硬件复杂度或成本的情况下，提高系统的误码率性能和整体效率。
这篇论文为IMDD-OQAM-OFDM系统的性能优化提供了一个新的视角，利用非参数直方图估计在低分辨率条件下实现更好的通信效果。这对于提升光通信系统在实际应用中的性能和可靠性具有重要意义，特别是在高数据速率和资源受限的环境中。"
IEEE Photonics Journal Performance Optimization by NPHE
been widely demonstrated in optical and fiber-wireless communication systems [1]–[6]. Owing to its
inherent merits such as high sidelobe suppression ratio, large asynchronous transmission feature,
high spectrum efficiency due to the non-requirement of cyclic prefix (CP) and time-frequency
domain well-localized pulse shapes, increasing system robustness against inter-carrier interference
(ICI) and frequency offset due to lower out-of-band power leakage [6] etc., it has been considered
as one of the most promising candidate waveform for intensity modulated direct detection-based
(IMDD) optical network transmission system [7]. For the OQAM-OFDM, it is essential to employ a
digital-to-analog convertor (DAC) to convert the OQAM-OFDM samples to time-domain waveforms.
Apparently, a high bit resolution is desirable for DACs implement, but such a resolution would lead
to high power consumption and cost [8], [9]. Besides, based on our previous work [10], it has
been verified that, the bit precisions of inverse fast Fourier transform (IFFT) and FFT were also the
important limiting factors, and the quantization noise caused by these operations would worsen
the system performance.
To deal with these issues, the common amplitude limiting [10], and the encoded peak-to-average
power ratio (PAPR) reduction [11], [12] schemes were proposed in IMDD-OFDM system to reduce
the bit resolution. However, for these schemes, only few quantization bits were reduced or the
limited performance improvements were achieved, and either extra complexity or clipping noises
were introduced which would be not suitable for the practical system. In fact, it is easy to get that,
the system performance strongly depends on the statistical distribution of the low-bit resolution
multicarrier signals’ samples, namely the quantization bit of small signal [13]. To deal with this, the
nonuniform quantization scheme with the Gaussian-like distribution assumption was proposed in
OFDM-based and OQAM-OFDM system [14], [15]. Besides, different from the common OFDM,
owing to the filtering effect of OQAM-OFDM signal a non-perfect Gaussian distri bution of signal
would be achieved, in which a sharp peak would occur on top of gaussian spectr um. In this way,
the nonuniform quantization with the Gaussian-like assumption would not be the optimal selection
for the IMDD-based OQAM-OFDM system.
Therefore, we propose the nonuniform quantization scheme with a novel nonparametric his-
togram signal sample estimation to quantify the signals for further reducing the DAC bits with
the acceptable system performance in the IMDD-based system. Due to no any requirements of
signal distribution in advance, this scheme would be more suitable for the irregular distribution
OQAM-OFDM system. In addition, by selecting the appropriate kernel density function and win-
dow width, the OQAM-OFDM signals’ distribution can be accurately estimated. Then, by combin-
ing nonlinear programming technique, the optimal nonuniform-distr ibuted discrete output levels of
OQAM-OFDM signal are achieved. In this way, more effective quantization bits can be assigned to
the small OQAM-OFDM signals hence reducing quantization noise and improving system perfor-
mance. To verify the feasibility of this scheme, the simulation (the commercial software OptiSystem
13) and experiment with the 10-Gb/s 16-OQAM-OFDM signal for 20-km fiber transmission with
different DACs and IFFT/FFT bit resolution is constructed. The simulation results show t hat, for
the 4-bit and 6-bit DAC case, ∼9.1 dB and ∼3.4 dB receiver sensitivity improvement @BER =
3.8e-3 can be achieved in comparison with the common uniform scheme, and compared to the
Gaussian-like distribution nonuniform method, our scheme also demonstrates better performance.
Meanwhile, by experiment, we can easily get that, our scheme with the 5-bit DAC can achieve
the same performance as the common scheme with 7-bit DAC. Therefore, our demonstrated
scheme can be considered as a promising way to improve the performance of the IMDD-based
OQAM-OFDM optical system, especially with the low-bit resolution DAC.
2. Principle
Firstly, we give the schematic diagram of IMDD OQAM-OFDM system to fur ther illustrate the
proposed NPHE scheme as shown in Fig. 1, which includes optical OQAM-OFDM transmitter,
fiber transmission link and OQAM-OFDM receiver. At the transmitter, as configured in our previous
work [15], the pseudo-random binary sequence (PRBS) data is firstly mapped into M-order offset
quadrature amplitude modulation (M-OQAM) complex symbols, then is divided into the odd and
Vol. 10, No. 4, August 2018 7905013
剩余13页未读，继续阅读
2021-05-10 上传
2021-03-01 上传
2021-03-01 上传
2021-02-07 上传
2021-03-14 上传
2021-04-01 上传
115 浏览量
weixin_38687968
- 粉丝: 7
- 资源: 969
我的内容管理
展开
