检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:徐文婧 李岩 刘宇旸[1,2] 李永富 郭宏翔 邱吉芳 洪小斌[1,2] 李蔚 左勇[1,2] 伍剑 Xu Wenjing;Li Yan;Liu Yuyang;Li Yongfu;Guo Hongxiang;Qiu Jifang;Hong Xiaobin;Li Wei;Zuo Yong;Wu Jian(School of Electronic Engineering,Beijing University of Posts and Telecommunications,Beijing 100876,China;State Key Laboratory of Information Photonics and Optical Communications,Beijing University of Posts and Telecommunications,Beijing 100876,China)
机构地区:[1]北京邮电大学电子工程学院,北京100876 [2]北京邮电大学信息光子学与光通信国家重点实验室,北京100876
出 处:《光学学报》2021年第12期63-73,共11页Acta Optica Sinica
基 金:国家重点研发计划(2019YFB1803601);国家自然科学基金(61875019,61675034);中央高校基本科研业务基金。
摘 要:针对目前相干光通信正交相移键控(QPSK)系统载波相位恢复算法的两种实现方式——基于复数运算和基于角度运算,就基于复数运算的Viterbi-Viterbi相位估计(VVPE)算法和基于角度运算的重心相位估计(BCPE)算法的恢复性能和硬件实现复杂度两个方面进行分析和对比。仿真证明VVPE算法比BCPE算法对抗加性噪声的能力更强,产生周期滑动的概率更低,两种算法在无周期滑动时性能接近。进行了速率为2.5 GBaud(符号率线宽积4×10^(-5))和10 GBaud(符号率线宽积1×10^(-5))的QPSK相干接收实验,采用Matlab离线和Verilog离线方式,对两种算法的性能进行了验证。结果表明:对于2.5 GBaud的QPSK信号,优化均值滤波器长度后,两种算法的接收灵敏度均为-52 dBm@1×10^(-3),VVPE算法和BCPE算法分别在接收功率为-54 dBm和-53 dBm时产生周期滑动;对于10 GBaud的QPSK信号,优化均值滤波器长度后,两种算法的接收灵敏度均为-47 dBm@1×10^(-3),BCPE算法在接收功率为-52 dBm时产生周期滑动。基于Xilinx公司的Virtex Ultrascale+系列现场可编程门阵列(FPGA)开发平台,对系统速率为10 GBaud、并行路数为64的两种算法进行了对比,结果表明VVPE算法在输入信号位数为8、BCPE算法在输入信号位数为16时,接收灵敏度均为-51 dBm@2×10-2,BCPE算法相对于VVPE算法,硬件查找表资源占用量降低了约5.78%。In current quadrature phase shift keying(QPSK)systems for coherent optical communications,there are two algorithms based on complex operation and angle operation to recover carrier phase.This paper analyzed and compared the recovery performance and implementation complexity of the Viterbi-Viterbi phase estimation(VVPE)algorithm based on complex operation and the barycenter phase estimation(BCPE)algorithm based on angle operation.The simulation proves that the VVPE algorithm is stronger in resisting additive noise than the BCPE algorithm and has a lower probability of cycle slip.The performances of the two algorithms are close when there is no cycle slip.In this paper,we carried out QPSK coherent reception experiments with rates of 2.5 GBaud(the product of symbol rate and linewidth is 4×10^(-5))and 10 GBaud(the product of symbol rate and linewidth is 1×10^(-5)).Then,we evaluated the performance of the two algorithms using the offline Matlab and Verilog methods.For the 2.5 GBaud QPSK signal,after the length of the average filter is optimized,the receiving sensitivity of the two algorithms is-52 dBm@1×10^(-3),and when the cycle slip occurs,the receiving power of the VVPE algorithm and BCPE algorithm is-54 dBm and-53 dBm,respectively.For the 10 GBaud QPSK signal,after the length of the average filter is optimized,the receiving sensitivity of the two algorithms is-47 dBm@1×10^(-3),and in the case of cycle slip,the receiving power of the BCPE algorithm is-52 dBm.Relying on Xilinx Virtex Ultrascale+FPGA,this paper compared the two algorithms with 10 GBaud system rate and 64 parallel channels.Regarding 8-bit and 16-bit input signals respectively for the VVPE algorithm and BCPE algorithm,the receiving sensitivity is-51 dBm@2×10^(-2).The occupancy of the hardware look-up table of the BCPE algorithm is about 5.78%lower than that of the VVPE algorithm.
分 类 号:TN914[电子电信—通信与信息系统]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.7