基于低采样率模数转换器的延时复用频分多址无源光网络

发布时间：2018-06-28 08:10

本文选题：无源光网络 + 正交频分多址技术　；参考：《物理学报》2017年19期

【摘要】：基于正交频分复用技术的无源光网络中,光网络单元为了获得其所属小部分下行数据,需高采样率模数转换器将所有频宽的信号恢复才能分出其所需要数据.同时正交频分信号峰均比很高,传输中容易引起非线性效应.为此,本文提出一种基于低采样模数转换器的延时复用频分多址无源光网络.在光线路终端将数据序列交错排序并在时域映射为正交幅度调制信号;再通过离散傅里叶变换扩频技术,将信号转换为频域信号并映射到子载波上.通过预先发送和回传训练信号,估测包括延时采样和低采样接收在内的信道频响;再将频域信号利用估测信息在光线路终端做预处理,从而使信号传输中的失真得到有效预补偿.本文实验演示了含有多个光网络单元的系统,对于含有M个光网络单元的无源光网络,模数转换器的采样率可以降低到1/M Nyquist采样率,实验中模数转换器的采样率可以降低到1/32 Nyquist采样率;由于下行信号通过光线路终端预处理实现失真预补偿,光网络单元接收到的信号不需要均衡,不需要傅里叶变换和傅里叶逆变换,避免了与之对应的相关计算量,降低了光网络单元的计算复杂度;由于使用了扩频技术,信号波形具有更低的峰均比,从而降低了非线性对信号的影响,增加了功率预算.此外,随着光网络单元的增加,信号的误码率几乎没有增加,光网络单元个数增加到32时,向前纠错极限为10~(-3)的功率代价小于0.5 dB;系统对光网络单元采样时刻偏离具有一定容限;25 km光纤传输的功率代价大约0.5 dB.理论和实验均证明本方案能够简化光网络单元,降低无源光网络的成本;与传统的无源光网络相比具有明显优势.
[Abstract]:In a passive optical network based on orthogonal frequency division multiplexing (OFDM), in order to obtain a small part of the downlink data to which the optical network unit belongs, a high sampling rate A / D converter is required to recover all the bandwidth signals in order to separate out the required data. At the same time, the peak to average ratio of orthogonal frequency division signal is very high, which is easy to cause nonlinear effect in transmission. In this paper, a delay multiplexing frequency division multiple access passive optical network based on low sampling A / D converter is proposed. The channel frequency response, including delay sampling and low sampling reception, is estimated by sending and returning the training signal in advance, and the frequency domain signal is preprocessed at the optical line terminal by using the estimation information. Thus the distortion in signal transmission can be precompensated effectively. In this paper, we demonstrate experimentally that the sampling rate of A / D converters can be reduced to 1 / M Nyquist sampling rate for passive optical networks with M optical network elements. In the experiment, the sampling rate of A / D converter can be reduced to 1 / 32 Nyquist sampling rate. Because the downlink signal is preprocessed by optical line terminal to realize distortion precompensation, the signal received by optical network unit does not need to be equalized. There is no need for Fourier transform and inverse Fourier transform, which avoids the corresponding computation and reduces the computational complexity of optical network units. Because of the use of spread spectrum technology, the signal waveform has a lower peak to average ratio (PAPR). Thus, the nonlinear effect on the signal is reduced, and the power budget is increased. The power cost of the forward error correction limit of 10 ~ (-3) is less than 0.5 dB, and the power cost of the optical network unit is about 0.5 dB when the sampling time deviates from the optical network unit with a certain tolerance of 25 km optical fiber transmission. Both theoretical and experimental results show that the proposed scheme can simplify the optical network unit and reduce the cost of the passive optical network, and it has obvious advantages over the traditional passive optical network.
【作者单位】：贵州大学大数据与信息工程学院;贵州大学物理学院;
【基金】：国家自然科学基金(批准号:11264006,61465002,61650403) 贵州省留学人员科技创新项目(批准号:2016-23) 贵州省社发公关项目(批准号:2013-3125)资助的课题~~
【分类号】：TN929.1

【相似文献】