WO2015123970A1 - 一种频偏估计方法、装置及帧同步子系统 - Google Patents

一种频偏估计方法、装置及帧同步子系统 Download PDF

Info

Publication number
WO2015123970A1
WO2015123970A1 PCT/CN2014/083831 CN2014083831W WO2015123970A1 WO 2015123970 A1 WO2015123970 A1 WO 2015123970A1 CN 2014083831 W CN2014083831 W CN 2014083831W WO 2015123970 A1 WO2015123970 A1 WO 2015123970A1
Authority
WO
WIPO (PCT)
Prior art keywords
channel
frequency offset
bit data
frame
frame header
Prior art date
Application number
PCT/CN2014/083831
Other languages
English (en)
French (fr)
Inventor
蔡轶
周伟勤
耿敏明
孟繁雪
贾振生
史学明
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2015123970A1 publication Critical patent/WO2015123970A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00Modulated-carrier systems
    • H04L27/0014Carrier regulation

Definitions

  • the present invention relates to the field of communications, and in particular, to a frequency offset estimation method, apparatus, and frame synchronization subsystem.
  • a frequency offset estimation method, apparatus, and frame synchronization subsystem In a communication system, the synchronization of a local carrier and a modulated carrier is one of the main factors that affect the demodulation performance of the system.
  • the frequency offset error caused by the inconsistent laser frequencies at the transmitting end and the receiving end is that the carrier synchronization phase of the receiving end must be eliminated.
  • the carrier synchronization frequency offset is also eliminated in the design of the coherent receiver.
  • the oscillation frequency of the local oscillator laser in the optical coherent receiver may be deviated from the carrier frequency. Taking a common laser as an example, if the nominal range of the oscillation frequency of the laser is ⁇ 1 GHz of the set frequency, the maximum possible between the local oscillator laser and the carrier is ⁇ 2 GHz. The frequency deviation is reflected on the symbol and is the phase offset.
  • the frequency offset estimation is an indispensable key module in the receiver.
  • An application of the 100G PM-QPSK optical transmission system structure is shown in Figure 1.
  • Various damages in the fiber channel such as dispersion, carrier frequency offset, phase offset, etc., can be caused by signal impairments at the receiver.
  • the digital signal processing DSP in the electrical domain is flexibly compensated.
  • the estimation algorithm of the carrier frequency offset in the receiver there are generally four power estimation and frequency sweep estimation methods.
  • the quadratic estimation algorithm and the sweep estimation algorithm have the same ability to estimate the frequency offset range, both of which are [- /8, + /8].
  • a method of estimating and compensating the frequency offset fuzzy is introduced, which is divided into two stages: In the first stage, the accurate frequency offset value needs to be calculated by the quadratic square frequency offset estimation method or the frequency sweep estimation algorithm; The method of detecting the training sequence of the receiving end is used to detect the state of the training sequence, estimate the frequency offset fuzzy value, and use the estimated frequency offset fuzzy value to correct the frequency offset value obtained by the fourth power estimation or the frequency sweep estimation.
  • the introduction of training sequences increases the optical transmission load.
  • an object of embodiments of the present invention is to provide a frequency offset estimation method, apparatus, and frame synchronization subsystem to implement correction of frequency offset estimation values without increasing optical transmission load.
  • the embodiment of the present invention provides the following solutions:
  • the embodiment of the present invention provides a frequency offset estimation method, which is used for the optical receiving end, and includes: the frame data of the frequency offset compensation and the phase offset compensation are in the same polarization state.
  • the corresponding I-channel and Q-path soft data respectively perform hard decision to obtain I-channel and Q-channel bit data, and the frequency offset compensation is performed based on the frequency offset estimation value;
  • the I-channel and Q-channel bit data, and the The pre-stored frame headers of the first frame data for frame synchronization are separately subjected to differential decoding to obtain differentially decoded I-channel and Q-way bit data and pre-stored frame headers; according to the differentially decoded I-channel and Q-way bit data and pre-stored frame headers Determining whether a frame header in the I-channel and the Q-channel bit data after the differential decoding is inverted, respectively, and obtaining a determination result; determining, according to the determination result, a phase offset corresponding to the frame data in the same polarization state And correcting the frequency offset estimation value by using the phase offset amount as a frequency offset fuzzy value to obtain a corrected frequency offset estimation value.
  • the determining, according to the differentially decoded I-channel and Q-way bit data and the pre-stored frame header, whether the frame headers in the differentially decoded I-channel and Q-channel bit data are inverted respectively, and obtaining the determination result includes: The differentially decoded I-channel and Q-channel bit data and the pre-stored frame header are used to detect a frame header in each of the differentially decoded I-channel and Q-channel bit data; and respectively determine the differentially decoded I-channel and Q-channel bit data in each Whether the frame header is inverted or not, and the judgment result is obtained.
  • the differentially decoded I-channel and Q-channel bit data and the pre-stored frame header are detected, and the differentially decoded I-channel and Q are detected.
  • the frame header in the any bit bit data is detected by: performing a sliding search on the any bit bit data according to the pre-stored frame header after the differential decoding, when searching When the bit data of the frame header length is compared with the differentially decoded pre-stored frame header, when the same number of bits is greater than a threshold value or smaller than a difference between the frame header length and the threshold value, the currently searched frame header
  • the bit data of the length is the frame header in the detected any of the way bit data; and the determining whether the frame header in each of the differentially decoded I channel and the Q way bit data is inverted, and obtaining the judgment result
  • the step if the frame header in the detected any bit bit data is compared with the differentially decoded pre-stored frame header, the same number of bits is greater than the threshold value, indicating that none of the paths are reversed.
  • the frame headers in the differentially decoded I channel and the Q channel bit data respectively occur when the inversion and the inversion do not occur, the corresponding phase offset of the frame data in the same polarization state is pi/2; the frame headers in the differentially decoded I channel and the Q channel bit data respectively do not reverse When the rotation and the inversion occur, the frame data is in the The corresponding phase shift amount in the same polarization state is -pi/2; when the frame headers in the differentially decoded I channel and the Q channel bit data are inverted, the frame data corresponds in the same polarization state.
  • the phase offset is pi.
  • the frequency offset estimation value is obtained by performing frequency offset estimation on the data corresponding to the frame data before the frequency offset compensation according to a fourth power estimation algorithm or a frequency sweep estimation algorithm.
  • the corrected frequency offset estimation value is used for feedback control of the local oscillator laser at the light receiving end.
  • the method further includes: performing frequency offset blur compensation on the I channel and the Q channel soft data respectively according to the frequency offset blur value.
  • the embodiment of the present invention further provides a frequency offset estimation apparatus, configured for the optical receiving end, comprising: a hard decision unit, configured to correspond to the I path and the Q corresponding to the frame data of the frequency offset compensation and the phase offset compensation in the same polarization state.
  • the road soft data is respectively subjected to a hard decision to obtain I and Q way bit data, the frequency offset compensation is performed based on the frequency offset estimation value;
  • the differential decoding unit is connected to the hard decision unit, and is set to the I path and the Q way bit Data, and pre-stored frame headers set to perform frame synchronization on the first frame data are separately differentially decoded to obtain differentially decoded I-channel and Q-way bit data and pre-stored frame headers;
  • a header state determination unit The differential decoding unit is connected, and is configured to determine whether the frame headers in the differentially decoded I-channel and Q-channel bit data are respectively inverted according to the differentially decoded I-channel and Q-channel bit data and the pre-stored frame header, and obtain a determination result;
  • a phase offset determining unit coupled to the frame header state determining unit, configured to determine, according to the determining result, a phase corresponding to the frame data in the same polarization state a shift amount estimation unit is connected to the
  • the frame header state determining unit includes: a frame header detecting subunit connected to the differential decoding unit, configured to detect the differential decoded image according to the differentially decoded I channel and Q channel bit data and the pre-stored frame header. a frame header in each of the I channel and the Q channel bit data; a frame header state determining subunit connected to the frame header detecting subunit, configured to respectively determine a frame header in each of the differentially decoded I channel and the Q channel bit data Whether or not inversion occurs, the judgment result is obtained.
  • the frame header in the any bit-bit data is detected as follows: According to differential decoding a pre-stored frame header, performing a sliding search on the any bit bit data, when the bit data of the searched frame header length is compared with the differentially decoded pre-stored frame header, the same number of bits is greater than a threshold value or less than the said When the difference between the frame header length and the threshold value, the bit data of the currently searched frame header length is the detected frame header in any of the way bit data; and the frame header state determining subunit detects The number of bits of any of the above roads " ⁇ ⁇ Compared with the pre-stored frame header after differential decoding, if the same number of bits is greater than the threshold, it indicates that no reverse occurs in any of the paths; if the same number of bits is smaller than the difference between the length of the frame header and the threshold, All the way reversed.
  • the frame data when the frame headers in the differentially decoded I channel and the Q channel bit data are respectively inverted and not inverted, the frame data is in the same polarization state.
  • the corresponding phase offset is pi/2; when the frame headers in the differentially decoded I-channel and Q-channel bit data are not inverted and inverted, respectively, the frame data corresponds to the same polarization state.
  • the embodiment of the present invention further provides a frame synchronization subsystem, which is configured as an optical receiving end of the DWDM high-speed optical transmission system, and includes the frequency offset estimating apparatus described above, and further includes: a receiving unit, and data in the optical receiving end
  • the signal processor DSP is connected to the hard decision unit, and configured to receive frame data after frequency offset compensation and phase offset compensation from the DSP; perform hard decision on the received data, and perform sliding correlation with the pre-stored frame header sequence, Frame header sequence Header state, the fuzzy inference offset value, and blur compensation received data frequency offset, while the fuzzy output offset value; an output unit configured to output a frequency offset estimation value after the correction.
  • FIG. 1 is a structural diagram of a 100G PM-QPSK optical system
  • FIG. 2 is a flow chart showing the steps of a frequency offset estimation method according to an embodiment of the present invention
  • FIG. 3 is a diagram showing a frequency offset of a preferred embodiment of the present invention.
  • Figure 4 is a block diagram showing a frequency offset estimation and compensation algorithm according to a preferred embodiment of the present invention
  • Figure 5 is a block diagram showing a structure of a frequency offset estimation apparatus according to an embodiment of the present invention.
  • the embodiments of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
  • 2 is a flow chart showing the steps of a frequency offset estimation method according to an embodiment of the present invention. Referring to FIG.
  • an embodiment of the present invention provides a frequency offset estimation method, which is used in an optical receiving end, and includes the following steps: Step 201, The frame data subjected to the frequency offset compensation and the phase offset compensation respectively performs hard decision on the corresponding I channel and Q channel soft data in the same polarization state to obtain I channel and Q channel bit data, and the frequency offset compensation is performed based on the frequency offset estimation value. Step 202: Perform differential decoding on the I channel and the Q channel bit data, and the pre-stored frame header used for frame synchronization of the first frame data, to obtain the differentially decoded I channel and Q channel bit data and the pre-stored frame.
  • Step 203 Determine, according to the differentially decoded I-channel and Q-channel bit data and the pre-stored frame header, whether the frame headers in the differentially decoded I-channel and Q-channel bit data are respectively inverted, and obtain a determination result;
  • Step 204 Determining, according to the determination result, a phase offset corresponding to the frame data in the same polarization state;
  • Step 205 using the phase offset as a frequency offset fuzzy value to the frequency offset estimation value Correction is performed to obtain a corrected frequency offset estimate.
  • the inversion of the frame header in different road bit data in the same polarization state is determined by differential decoding, and the phase offset corresponding to the frame data is determined according to the frequency offset value, and the frequency offset estimation value is used as the frequency offset fuzzy value. Corrected so that the correction of the frequency offset estimate can be achieved without increasing the optical transmission load.
  • the light receiving end is in a DWDM high speed optical transmission system.
  • the method may include: detecting, according to the differentially decoded I channel and Q channel bit data and the pre-stored frame header, a frame header in each of the differentially decoded I channel and the Q channel bit data; respectively determining the differentially decoded I channel and the Q channel bit Whether or not the frame header in each of the data is inverted, the judgment result is obtained. Wherein, for any one of the I-channel and the Q-bit data after the differential decoding, the differentially decoded I-channel and Q-bit data and the pre-stored frame header are detected, and the differentially decoded I and Q channels are detected.
  • the frame header in the any bit bit data can be detected by: Performing a sliding search on the any bit bit data according to the pre-stored frame header after the differential decoding.
  • the bit data of the searched frame header length is compared with the differentially decoded pre-stored frame header, the same number of bits is greater than a threshold value or When the difference between the frame header length and the threshold value is smaller, the bit data of the currently searched frame header length is the frame header in the detected any path bit data; and the differentially decoded I is respectively determined.
  • the step of determining the phase offset corresponding to the frame data in the same polarization state according to the determination result may include: differentially decoded I channel and Q channel bit data.
  • the corresponding phase offset of the frame data in the same polarization state is pi/2; in the differentially decoded I channel and Q channel bit data When the frame headers are not inverted and inverted, respectively, the corresponding phase offset of the frame data in the same polarization state is -pi/2; frames in the differentially decoded I channel and Q channel bit data When the heads are inverted, the corresponding phase shift amount of the frame data in the same polarization state is pi.
  • pi is the pi.
  • the frequency offset estimation value may be obtained by performing frequency offset estimation on the data corresponding to the frame data before the frequency offset compensation according to a fourth power estimation algorithm or a frequency sweep estimation algorithm.
  • the corrected frequency offset estimation value may be used to perform feedback control on the local oscillator laser at the optical receiving end.
  • the method may further include: performing frequency offset blur compensation on the I channel and the Q channel soft data according to the frequency offset blur value, and outputting the signal.
  • BEST MODE FOR CARRYING OUT THE INVENTION In order to make the embodiments of the present invention more clearly understood, preferred embodiments of the embodiments of the present invention are provided below.
  • the preferred embodiment provides a method and apparatus for frequency offset estimation and compensation applied in a DWDM high-speed optical transmission system, and relates to a frequency offset by monitoring a frame header state in a DWDM high-speed optical transmission system.
  • the method of estimating and estimating the fuzzy quantity can enlarge the range of the frequency offset estimation compensation, and the accuracy of the frequency offset estimation is not lost, and the added operation amount and the logic amount are few. And it is easy to implement ASIC.
  • the technical problem to be solved by the preferred embodiment is: In the DWDM high-speed optical transmission system, a method for widening the frequency offset estimation range is provided, and the frequency offset value estimated by the fourth-order estimation algorithm or the frequency sweep calculation method is provided. Make corrections to get the correct frequency offset estimate.
  • a method for widening the frequency offset estimation range is provided for the coherent receiver demodulation algorithm of the DWDM high-speed optical transmission system, including the following main steps: performing hard decision on the soft data after phase recovery, and obtaining a hard Data; differential decoding of hard data to eliminate absolute phase deviation, after the differential data is blurred, the deflection phase is fixed; the differentially decoded data and the differentially decoded pre-stored frame header sequence Perform sliding correlation, find the position of the frame header; monitor the state of the frame header, observe the inversion state of the I and Q signals in the same polarization state, and determine whether the true value of the frequency offset exceeds the pre-frequency offset estimation compensation module.
  • the fuzzy compensation range if the frequency offset compensation range is exceeded, the frequency offset estimation value is corrected, and the current frame data is compensated accordingly to ensure the data flow is normal.
  • the frequency offset estimation range can be expanded by many times, the frequency offset estimation accuracy is not lost, and the increased computational amount and logic amount are small, which is convenient for ASIC implementation.
  • the apparatus in the preferred embodiment includes: a frequency offset estimation unit, a hard decision unit, a differential decoding unit, a frame header sliding monitoring unit, a frame header state determining unit, a frequency offset compensation unit, and a frequency offset estimation value modifying unit.
  • the preferred embodiment provides a method for widening the frequency offset estimation range in the DWDM high-speed optical transmission system.
  • the frequency offset estimation range can be expanded by many times, the frequency offset estimation accuracy is not lost, and the increased computation amount and logic amount are small. Easy to implement ASIC. Algorithm principle: Since the quadratic estimation algorithm and the sweep estimation algorithm have periodicity, the frequency offset value beyond the range of [- /8, + /8] will be reduced to [- /8, + /8].
  • the frequency offset value -/ is evenly different from /8 (the even number can be positive or negative).
  • Am Am*exp(j*m*(-fn*(+ /8)))) 0
  • the result of the /c compensation is more rotated than the result of the -/compensation exp(/* *w*( /8) ) (or exp(/* *(-w*(+ /8))))).
  • Converting /8 to an angle is pi/4, ie more compensation for exp /'*w* «* p ⁇ 4)) (or exp( * *(— «*(+ ⁇ 4))).
  • the angle value compensates for the angle of rotation in the data stream, and the frequency offset can be compensated back to obtain a normal data stream.
  • the angle value is sent to the frequency offset compensation value correction module for correcting the fourth power estimation algorithm or frequency sweeping. Estimating the exact frequency offset value obtained by the algorithm, and using the modified frequency offset value to compensate the frequency offset generated by the local oscillator laser and the originating laser. Specific examples of the preferred embodiment are given below: As shown in FIG. 1, the optical transmitting end The laser emits light and splits into two beams of deflected light through the PBS beam splitter.
  • the two polarized lights are respectively orthogonally modulated with the electrical signal, and two sets of orthogonal signals are obtained through the PBC combiner to obtain a polarization-multiplexed optical signal, which reaches the optical receiving end through the channel; the coherent receiver receives the originating signal After that, the two polarized light signals are separated by the PBS beam splitter, and respectively demodulated with the optical signal emitted by the local oscillator laser, photoelectrically converted, and the ADC is collected to obtain a digital signal, and digital signal processing is performed.
  • the dispersion value is estimated, and dispersion compensation is performed; the data after the dispersion is removed is subjected to clock recovery and interpolation filtering, the clock phase value is calculated and fed back to the ADC, and the acquired signal is restored to an ideal state.
  • the sampling point blindly equalizing to eliminate residual dispersion and polarization mode dispersion, and polarization demultiplexing the two polarization states, that is, separating the X and Y polarization states; due to the frequency offset of the local oscillator laser and the originating laser, the laser There are factors such as line width, and it is necessary to perform frequency offset estimation and phase estimation on the data, and frequency offset compensation.
  • phase offset compensating partial elimination phase In addition to the frequency offset, phase offset compensating partial elimination phase; data phase after partial compensation can already be recovered constellation points have the normal, the data passes into the skew sync frame synchronization subsystem.
  • the frame synchronization subsystem hardly judging the obtained soft data stream, performing differential decoding, performing sliding correlation with the pre-stored frame header sequence, slidingly searching for the position of the frame header, and detecting the state of the frame header; performing frequency offset Fuzzy compensation and other operations. Because as the laser ages, the frequency offset range will become larger than the range of [- ⁇ 74, ⁇ /4]. At this time, the quadratic estimation algorithm and the sweep algorithm can only estimate [- /4, /4].
  • the frequency offset therefore, uses the state of the frame header detection to estimate and compensate for the integer multiple of the /2 offset.
  • the block diagram of the frequency offset estimation and compensation algorithm is shown in FIG. 3: the frequency offset estimation obtains the frequency offset compensation value, and the frequency offset compensation value is sent.
  • correction module frequency offset compensation for this step will appear frequency offset blur for frequency offsets greater than [- /8, + /8] (see algorithm principle); frame synchronization subsystem acquires soft data stream, after hard decision, difference Decoding; After the differential decoding, the data has no frequency offset fuzzy information, and the phase change of the whole data reflects the value of the frequency offset blur, and the value of the frequency offset blur is determined by calculating the phase change; [_ ⁇ / 4 , ⁇ The frequency offset value within / 4 ] has been compensated in the frequency offset estimation module, so the value of the frequency offset blur in the frame synchronization subsystem is an integer multiple of /2. In one polarization state, there are two paths of I and Q. After the differential decoding, the overall phase is a fixed offset.
  • the phase offset value can be judged by the deflection of the polarization state. If the I path is inverted and the Q path is unchanged, it is ⁇ /2. Q road is reversed, I road Change, then ⁇ /2, if both I and Q are unchanged, it is 0. If both I and Q are reversed, then the phase offset obtained by this is the frame synchronization.
  • the frequency offset fuzzy value of the soft data stream of the subsystem, the X and Y polarization states are calculated in the same way as the frequency offset blur; the calculated frequency offset fuzzy compensation data stream is used, and the frequency offset fuzzy compensation value is sent to the frequency offset compensation value.
  • the correction module corrects the frequency offset compensation value calculated by the frequency offset estimation module; and controls the local oscillator laser with the modified frequency offset value.
  • Step 401 hard decision.
  • the frame synchronization subsystem acquires two soft-state soft data streams from the phase-off output, each of which has two data of I and Q, and performs hard decision on the soft data to obtain a binary bit stream.
  • Step 402 differential decoding. Differential quadrature phase shift keying (DQPSK) is performed on a bit stream in a polarization state.
  • DQPSK Differential quadrature phase shift keying
  • the bit stream before decoding has a frequency offset blur, and the decoded bit stream is in the form of a difference.
  • Step 403 sliding the search frame header. Perform a sliding search on the bit stream after decoding the frame header and the bit stream after data decoding, and slide one bit at a time to compare how many bit streams are the same. Set the frame header length to M, and the threshold to A. If the same number of bits If it is greater than A or less than MA, it means that the frame header is found, otherwise it will be searched until the frame header is found, otherwise the frequency offset fuzzy estimation compensation is exited. Step 404, monitoring the frame header status of the I and Q paths.
  • Step 405 determining a value of the frequency offset blur. If the I path is reversed and the Q path is not inverted, the frequency offset blur value is pi/2; if the I path is not inverted and the Q path is inverted, the value of the frequency offset blur is -pi/2; If the Q path is reversed, the frequency offset blur value is pi; if the I path and the Q path are not reversed, the value of the frequency offset blur is 0.
  • Step 406 Compensating the soft data by using the obtained value of the frequency offset blur, and correcting the value of the frequency offset estimation, and using the corrected value to feedback control the local oscillator laser.
  • a DCW high-speed optical transmission system is used to optimize the frequency offset estimation and cancellation of the frequency offset blur compensated by the fourth power algorithm or the frequency sweep algorithm in the manner of monitoring the frame header state.
  • the method can realize the estimation and compensation of the frequency offset blur by adding the function of monitoring and judging the state of the frame header in the frame synchronization subsystem, which is very easy to implement by the ASIC hardware.
  • FIG. 5 is a structural block diagram of a frequency offset estimation apparatus according to an embodiment of the present invention. Referring to FIG.
  • an embodiment of the present invention further provides a frequency offset estimation apparatus, which is used for an optical receiving end, and includes: a hard decision unit, configured to The frame data of the frequency offset compensation and the phase offset compensation are respectively subjected to hard decision in the corresponding I channel and Q channel soft data in the same polarization state to obtain I channel and Q channel bit data, and the frequency offset compensation is based on the frequency offset estimation value.
  • a hard decision unit configured to The frame data of the frequency offset compensation and the phase offset compensation are respectively subjected to hard decision in the corresponding I channel and Q channel soft data in the same polarization state to obtain I channel and Q channel bit data, and the frequency offset compensation is based on the frequency offset estimation value.
  • a differential decoding unit connected to the hard decision unit, configured to differentially decode the I channel and the Q channel bit data, and the pre-stored frame header set to frame synchronization of the first frame data to obtain differential decoding After I and Q bit data and pre-stored frame headers;
  • the frame header state determining unit is connected to the differential decoding unit, and is configured to determine, according to the differentially decoded I channel and Q channel bit data and the pre-stored frame header, the frame headers in the differentially decoded I channel and the Q channel bit data respectively Whether the inversion occurs, and the determination result is obtained;
  • the phase offset determining unit is connected to the frame header state determining unit, and configured to determine, according to the determining result, a phase offset corresponding to the frame data in the same polarization state a shift amount estimation unit is connected to the phase shift amount determining unit, and is configured to correct the frequency offset estimated value by using the phase shift amount as a frequency offset estimation ambiguity, and obtain the corrected Frequency offset estimate.
  • the frame header state determining unit may include: a frame header detecting subunit, connected to the differential decoding unit, configured to detect according to the differentially decoded I channel and Q channel bit data and the pre-stored frame header.
  • a frame header in each of the differentially decoded I channel and the Q channel bit data a frame header state determining subunit connected to the frame header detecting subunit, configured to respectively determine the differentially decoded I channel and the Q channel bit data Whether the frame header in the inversion is inverted, and the judgment result is obtained.
  • the header of any one of the bit data is detected by the following method: a pre-stored frame header, performing a sliding search on the any bit bit data, when the bit data of the searched frame header length is compared with the differentially decoded pre-stored frame header, the same number of bits is greater than a threshold value or smaller than the frame
  • the bit data of the currently searched frame header length is the frame header in the detected any of the way bit data
  • the detected in the frame header state determining subunit The frame header in any one of the bit data is compared with the pre-stored frame header of the differential decoding, and the same number of bits is greater than the threshold value, indicating that the reverse path is not reversed; the number of bits corresponding to the frame length is smaller than the frame length
  • the difference between the threshold and the threshold indicates that any of the paths is reversed.
  • the phase offset amount determining unit when the frame headers in the differentially decoded I channel and the Q channel bit data are respectively inverted and not inverted, the frame data is in the same
  • the corresponding phase offset in the polarization state is pi/2;
  • the corresponding phase offset of the frame data in the same polarization state is -pi/2;
  • the phase offset corresponding to the frame data in the same polarization state is pi.
  • the embodiment of the present invention further provides a frame synchronization subsystem, which is configured as a light receiving end of the DWDM high-speed optical transmission system, and includes the frequency offset estimating apparatus described above, further comprising: a receiving unit, and the optical receiving end
  • the data signal processor DSP is connected to the hard decision unit, and is configured to receive frame data subjected to frequency offset compensation and phase offset compensation from the DSP; and an output unit configured to output the corrected frequency offset estimation value.
  • the received data is hard-decised, and is correlated with the pre-stored frame header sequence, the frame header sequence is found, the frame header state is determined, the frequency offset fuzzy value is estimated, and the received data is subjected to frequency offset fuzzy compensation and output simultaneously. Frequency offset fuzzy value.
  • the inversion of the frame header in different road bit data in the same polarization state is determined by differential decoding, and the phase offset corresponding to the frame data is determined accordingly.
  • the amount is used as a frequency offset fuzzy value to correct the frequency offset estimation value, so that the frequency offset estimation value can be corrected without increasing the optical transmission load.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Optical Communication System (AREA)

Abstract

本发明实施例提供一种频偏估计方法、装置及帧同步子系统。所述方法包括:对经过频偏补偿和相偏补偿的帧数据在同一偏振态内对应的I路和Q路软数据分别进行硬判决,得到I路和Q路比特数据;对I路和Q路比特数据、和预存帧头分别进行差分解码,得到差分解码后的I路和Q路比特数据和预存帧头;根据差分解码后的I路和Q路比特数据和预存帧头,判断差分解码后的I路和Q路比特数据中的帧头分别是否发生反转,获取判断结果;根据所述判断结果,确定所述帧数据在所述同一偏振态内对应的相位偏移量;将所述相位偏移量作为频偏模糊值对频偏估计值进行修正,得到修正后的频偏估计值。本发明实施例不需要增加光传输负荷就能实现频偏估计值的修正。

Description

一种频偏估计方法、 装置及帧同步子系统 技术领域 本发明实施例涉及通信领域,尤其涉及一种频偏估计方法、装置及帧同步子系统。 背景技术 通信系统中,本地载波与调制载波的同步性是影响系统解调性能的主要原因之一。 信号在传播过程中, 因发送端和接收端的激光器频率不一致导致的频偏误差因素, 是 接收端的载波同步阶段必须要完成消除的。 对于应用在骨干网承载大容量数据传输的密集波分复用 (DWDM , Dense Wavelength Division Multiplexing) 高速光传输系统, 例如 100G PM-QPSK系统, 相干 接收机设计时也同样存在载波同步频偏消除的要求。 由于激光器的非理想特性, 光相 干接收机中本振激光器的振荡频率, 可能会与载波频率之间存在一定的偏差。 以常见 的激光器为例,若激光器的振荡频率标称范围为设置频率的 ± 1GHZ,则本振激光器与 载波之间的最大可能为 ± 2GHZ。 该频率偏差反映在符号上, 是相位的偏移, 对于 PM-QPSK这种相位调制系统来说,必须去除频偏带来的相位偏移,才有可能解调出最 后的数据符号。 所以, 频偏估计是接收机中不可缺少的一个关键模块。 一种应用的 100G PM-QPSK光传输系统结构如图 1所示, 光纤 (fiber) 信道中的 各种损伤, 如色散、 载波频偏、 相位偏移等导致的信号损伤, 都能在接收机中通过电 域的数字信号处理 DSP来灵活地补偿。对于接收机中的载波频偏的估计算法, 一般有 四次方估计和扫频估计方法。 四次方估计算法和扫频估计算法对频偏估计范围的能力 是一致的, 均为 [- /8,+ /8]。 但是随着激光器的老化, 频偏范围会变大, 有可能超出四次方估计算法和扫频估 计算法的估计范围 [- /8,+ /8], 造成上述频偏估计算法失效。 因此, 引出一种频偏 模糊的估计和补偿的方法, 分为两个阶段: 第一阶段, 需要通过四次方频偏估计法或 者扫频估计算法计算出精确的频偏值; 第二阶段:利用对接收端训练序列检测的方式, 检测记录训练序列的状态, 估计出频偏模糊值, 用估计得到的频偏模糊值去修正四次 方估计或者扫频估计得到的频偏值。 但是, 训练序列的引入, 增加了光传输负荷。 发明内容 有鉴于此,本发明实施例的目的是提供一种频偏估计方法、装置及帧同步子系统, 以在不增加光传输负荷的情况下实现频偏估计值的修正。 为解决上述技术问题, 本发明实施例提供方案如下: 本发明实施例提供一种频偏 估计方法, 用于光接收端, 包括: 对经过频偏补偿和相偏补偿的帧数据在同一偏振态 内对应的 I路和 Q路软数据分别进行硬判决, 得到 I路和 Q路比特数据, 所述频偏补 偿基于频偏估计值进行; 对 I路和 Q路比特数据、 和用于对所述第一帧数据进行帧同 步的预存帧头分别进行差分解码,得到差分解码后的 I路和 Q路比特数据和预存帧头; 根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后的 I路和 Q路比 特数据中的帧头分别是否发生反转, 获取判断结果; 根据所述判断结果, 确定所述帧 数据在所述同一偏振态内对应的相位偏移量; 将所述相位偏移量作为频偏模糊值对所 述频偏估计值进行修正, 得到修正后的频偏估计值。 优选地, 所述根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码 后的 I路和 Q路比特数据中的帧头分别是否发生反转, 获取判断结果包括: 根据差分 解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I路和 Q路比特数据各 自中的帧头;分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反转, 获取所述判断结果。 优选地, 对于差分解码后的 I路和 Q路比特数据中的任一路比特数据, 所述根据 差分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I路和 Q路比特数 据各自中的帧头的步骤中, 所述任一路比特数据中的帧头通过如下方式检测得到: 根 据差分解码后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜索到的帧头长 度的比特数据与差分解码后的预存帧头相比, 相同的比特数大于门限值或者小于所述 帧头长度与所述门限值之差时, 当前搜索到的帧头长度的比特数据为检测到的所述任 一路比特数据中的帧头; 所述分别判断差分解码后的 I路和 Q路比特数据各自中的帧 头是否发生反转, 获取所述判断结果的步骤中, 检测到的所述任一路比特数据中的帧 头与差分解码后的预存帧头相比, 相同的比特数大于门限值, 则表明所述任一路没有 发生反转; 相同的比特数小于帧头长度与门限值之差, 则表明所述任一路发生反转。 优选地, 所述根据所述判断结果, 确定所述帧数据在所述同一偏振态内对应的相 位偏移量的步骤中, 差分解码后的 I路和 Q路比特数据中的帧头分别发生反转和没有 发生反转时, 所述帧数据在所述同一偏振态内对应的相位偏移量为 pi/2; 差分解码后 的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转时, 所述帧数据在所述 同一偏振态内对应的相位偏移量为 -pi/2; 差分解码后的 I路和 Q路比特数据中的帧头 均发生反转时, 所述帧数据在所述同一偏振态内对应的相位偏移量为 pi。 优选地, 所述频偏估计值为根据四次方估计算法或扫频估计算法对所述帧数据在 所述频偏补偿前对应的数据进行频偏估计得到。 优选地, 所述修正后的频偏估计值用于对所述光接收端的本振激光器进行反馈控 制。 优选地, 还包括: 根据所述频偏模糊值对 I路和 Q路软数据分别进行频偏模糊补 偿后输出。 本发明实施例还提供一种频偏估计装置, 用于光接收端, 包括: 硬判决单元, 设 置为对经过频偏补偿和相偏补偿的帧数据在同一偏振态内对应的 I路和 Q路软数据分 别进行硬判决, 得到 I路和 Q路比特数据, 所述频偏补偿基于频偏估计值进行; 差分 解码单元, 与所述硬判决单元连接, 设置为对 I路和 Q路比特数据、 和设置为对所述 第一帧数据进行帧同步的预存帧头分别进行差分解码, 得到差分解码后的 I路和 Q路 比特数据和预存帧头; 帧头状态判断单元, 与所述差分解码单元连接, 设置为根据差 分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后的 I路和 Q路比特数据 中的帧头分别是否发生反转, 获取判断结果; 相位偏移量确定单元, 与所述帧头状态 判断单元连接, 设置为根据所述判断结果, 确定所述帧数据在所述同一偏振态内对应 的相位偏移量; 频偏估计值修正单元, 与所述相位偏移量确定单元连接, 设置为将所 述相位偏移量作为频偏估计模糊度对所述频偏估计值进行修正, 得到修正后的频偏估 计值。 优选地, 所述帧头状态判断单元包括: 帧头检测子单元, 与所述差分解码单元连 接, 设置为根据差分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I 路和 Q路比特数据各自中的帧头;帧头状态判断子单元,与所述帧头检测子单元连接, 设置为分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反转, 获取 所述判断结果。 优选地, 对于差分解码后的 I路和 Q路比特数据中的任一路比特数据, 所述帧头 检测子单元中, 所述任一路比特数据中的帧头通过如下方式检测得到: 根据差分解码 后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜索到的帧头长度的比特数 据与差分解码后的预存帧头相比, 相同的比特数大于门限值或者小于所述帧头长度与 所述门限值之差时, 当前搜索到的帧头长度的比特数据为检测到的所述任一路比特数 据中的帧头; 所述帧头状态判断子单元中, 检测到的所述任一路比特数《Λ ^τ^ 差分解码后的预存帧头相比, 相同的比特数大于门限值, 则表明所述任一路没有发生 反转; 相同的比特数小于帧头长度与门限值之差, 则表明所述任一路发生反转。 优选地, 所述相位偏移量确定单元中, 差分解码后的 I路和 Q路比特数据中的帧 头分别发生反转和没有发生反转时, 所述帧数据在所述同一偏振态内对应的相位偏移 量为 pi/2; 差分解码后的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转 时,所述帧数据在所述同一偏振态内对应的相位偏移量为 -pi/2;差分解码后的 I路和 Q 路比特数据中的帧头均发生反转时, 所述帧数据在所述同一偏振态内对应的相位偏移 量为 pio 本发明实施例还提供一种帧同步子系统, 设置为 DWDM高速光传输系统的光接 收端, 包括以上所述的频偏估计装置, 还包括: 接收单元, 与所述光接收端中的数据 信号处理器 DSP和所述硬判决单元连接, 设置为从所述 DSP接收经过频偏补偿和相 偏补偿的帧数据; 对接收的数据进行硬判决, 并与预存帧头序列进行滑动相关, 找到 帧头序列, 判断帧头状态, 推断频偏模糊值, 并对接收数据进行频偏模糊补偿, 同时 输出频偏模糊值; 输出单元, 设置为将修正后的频偏估计值输出。 从以上所述可以看出, 本发明实施例至少具有如下有益效果: 通过差分解码的方 式来判断同一偏振态内不同路比特数据中帧头的反转情况, 据此确定帧数据对应的相 位偏移量, 将其作为频偏模糊值对频偏估计值进行修正, 从而不需要增加光传输负荷 就能实现频偏估计值的修正。 附图说明 图 1表示 100G PM-QPSK光系统结构图; 图 2 表示本发明实施例提供的一种频偏估计方法的步骤流程图; 图 3表示本发明实施例的较佳实施方式的频偏估计和补偿算法的框图; 图 4表示本发明实施例的较佳实施方式的频偏估计和补偿算法的流程图; 图 5 表示本发明实施例提供的一种频偏估计装置的结构框图。 具体实施方式 为使本发明实施例的目的、 技术方案和优点更加清楚, 下面将结合附图及具体实 施例对本发明实施例进行详细描述。 图 2 表示本发明实施例提供的一种频偏估计方法的步骤流程图, 参照图 2, 本发 明实施例提供一种频偏估计方法, 用于光接收端, 包括如下步骤: 步骤 201, 对经过频偏补偿和相偏补偿的帧数据在同一偏振态内对应的 I路和 Q 路软数据分别进行硬判决, 得到 I路和 Q路比特数据, 所述频偏补偿基于频偏估计值 进行; 步骤 202, 对 I路和 Q路比特数据、 和用于对所述第一帧数据进行帧同步的预存 帧头分别进行差分解码, 得到差分解码后的 I路和 Q路比特数据和预存帧头; 步骤 203, 根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后 的 I路和 Q路比特数据中的帧头分别是否发生反转, 获取判断结果; 步骤 204, 根据所述判断结果, 确定所述帧数据在所述同一偏振态内对应的相位 偏移量; 步骤 205, 将所述相位偏移量作为频偏模糊值对所述频偏估计值进行修正, 得到 修正后的频偏估计值。 可见, 通过差分解码的方式来判断同一偏振态内不同路比特数据中帧头的反转情 况,据此确定帧数据对应的相位偏移量,将其作为频偏模糊值对频偏估计值进行修正, 从而不需要增加光传输负荷就能实现频偏估计值的修正。 其中, 所述光接收端在 DWDM高速光传输系统中。 本发明实施例中, 所述根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断 差分解码后的 I路和 Q路比特数据中的帧头分别是否发生反转, 获取判断结果可以包 括: 根据差分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I路和 Q 路比特数据各自中的帧头; 分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反转, 获取所 述判断结果。 其中, 对于差分解码后的 I路和 Q路比特数据中的任一路比特数据, 所述根据差 分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I路和 Q路比特数据 各自中的帧头的步骤中, 所述任一路比特数据中的帧头可以通过如下方式检测得到: 根据差分解码后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜索到的 帧头长度的比特数据与差分解码后的预存帧头相比, 相同的比特数大于门限值或者小 于所述帧头长度与所述门限值之差时, 当前搜索到的帧头长度的比特数据为检测到的 所述任一路比特数据中的帧头; 所述分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反转, 获 取所述判断结果的步骤中, 检测到的所述任一路比特数据中的帧头与差分解码后的预 存帧头相比, 相同的比特数大于门限值, 则可以表明所述任一路没有发生反转; 相同 的比特数小于帧头长度与门限值之差, 则可以表明所述任一路发生反转。 本发明实施例中, 所述根据所述判断结果, 确定所述帧数据在所述同一偏振态内 对应的相位偏移量的步骤中, 可以有: 差分解码后的 I路和 Q路比特数据中的帧头分别发生反转和没有发生反转时, 所 述帧数据在所述同一偏振态内对应的相位偏移量为 pi/2; 差分解码后的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转时, 所 述帧数据在所述同一偏振态内对应的相位偏移量为 -pi/2; 差分解码后的 I路和 Q路比特数据中的帧头均发生反转时, 所述帧数据在所述同 一偏振态内对应的相位偏移量为 pi。 其中, pi即圆周率。 本发明实施例中, 所述频偏估计值可以为根据四次方估计算法或扫频估计算法对 所述帧数据在所述频偏补偿前对应的数据进行频偏估计得到。 本发明实施例中, 所述修正后的频偏估计值可以用于对所述光接收端的本振激光 器进行反馈控制。 本发明实施例中, 还可以包括: 根据所述频偏模糊值对 I路和 Q路软数据分别进行频偏模糊补偿后输出。 <较佳实施方式 > 为将本发明实施例阐述得更加清楚明白,下面提供本发明实施例的较佳实施方式。 本较佳实施方式提供一种应用在 DWDM高速光传输系统中的频偏估计和补偿的 方法和装置, 涉及在 DWDM高速光传输系统中, 通过对帧头状态的监测的方式, 达 到对频偏模糊量的判断和估计的方法, 相比四次方频偏估计算法或者扫频算法, 可以 扩大频偏估计补偿的范围, 频偏估计的精度也没有损失,增加的运算量和逻辑量很少, 而且便于 ASIC实现。 本较佳实施方式所要解决的技术问题是: 在 DWDM高速光传输系统中, 提供一 种扩大频偏估计范围的方法, 该方法对四次方估计算法或扫频计算法估计出来的频偏 值进行修正, 从而得到正确的频偏估计值。 该方法相比传统四次方估计算法和扫频算 法, 其估计范围更大, 精度没有损失, 增加的运算量和逻辑量很少, 便于 ASIC实现。 本较佳实施方式中对 DWDM高速光传输系统的相干接收机解调算法中, 提供一 种扩大频偏估计范围的方法,包括以下主要步骤:对相位恢复后的软数据进行硬判决, 得到硬数据; 对硬数据进行差分解码, 以消除绝对的相位偏差, 存在频偏模糊的输入 数据经过差分解码后, 偏转相位是固定的; 对差分解码后的数据与差分解码后的预存 的帧头序列进行滑动相关, 找到帧头位置; 监测帧头的状态, 观察同一个偏振态内 I 和 Q两路信号的反转状态, 以此判断频偏的真实值是否超出了前级频偏估计补偿模块 的无模糊补偿范围; 若超出频偏补偿范围则对频偏估计值进行修正, 同时对当前的帧 数据进行相应补偿, 保证数据流正常。 经过实验分析和比较, 采用本较佳实施方式提出的方法, 频偏估计范围可以扩大 很多倍, 频偏估计精度没有损失, 增加的运算量和逻辑量很少, 便于 ASIC实现。 本较佳实施方式中的装置, 包括: 频偏估计单元, 硬判决单元, 差分解码单元, 帧头滑动监测单元, 帧头状态判断单元、 频偏补偿单元和频偏估计值修正单元。 本较佳实施方式提供了 DWDM高速光传输系统中, 一种扩大频偏估计范围的方 法, 频偏估计范围可以扩大很多倍, 频偏估计精度没有损失, 增加的运算量和逻辑量 很少, 便于 ASIC实现。 算法原理: 由于四次方估计算法和扫频估计算法具有周期性, 即超出 [- /8,+ /8]范围的频 偏值都会归算到 [- /8,+ /8]内。 假设频偏真实值为 f, 且 f 的范围满足 (《 + 1)*(+ fsl )> f>(n-\)*(+ fsl ) (或 (n + \)*(-fs/S)< /<(w-l)*(- /8)),n=2,4,6...,则频偏模块估计值 fe = f-n*(+ fs/S) ( 或 fo = f + n*(+ /8) ) , 频偏补偿值 fc , S卩 -/ + "*(+ fsli) (或 -f-n*(+fs/S)), 与实际需要补偿的频偏值 -/相差了 /8的偶数倍(该偶数可以为 正数也可以为负数), 这一现象称作频偏模糊。 数据流上的频偏补偿公式: = *exp(/*∞*/c), m=0, 1, 2...。 将/ c的表达 式 代 人 , Am = Am * exp( *m*(- f + n*(+ fs / 8))) ( 或
Am = Am*exp(j*m*(-f-n*(+ /8))))0 用 /c补偿的结果比用-/补偿的结果多旋 转了 exp(/* *w*( /8)) (或 exp(/* *(-w*(+ /8))) )。 把 /8换算到角度就是 pi/ 4 , 即 多补偿了 exp /'*w*«* p〃4)) (或 exp( * *(— «*(+ρ〃4)》)。 差分角牟码后 ^ = *εχρ(_/'*(¾*(- + «*(+ /8)》- (( - 1)*(- + «*(+ ^/8)))) = B * exp( *{-f + n*{pi/ 4))) (或 *εχρ(/·*(-/-"*(ρ /4))) ), 差分解码后的数据相 比真实值旋转了 /4的偶数倍, 也就是 /2的整数倍的一个固定相位。 这个固定相 位会造成 I路或 Q路反相, 通过帧同步模块可以检测出来, 并通过检测的结果判断频 偏角度值, 用该角度值补偿数据流里面旋转的角度, 即可把频偏补偿回来, 得到正常 的数据流。 同时将该角度值发送给频偏补偿值修正模块, 用来修正四次方估计算法或 者扫频估计算法得到的精确频偏值, 并用修正后的频偏值来补偿本振激光器与发端激 光器产生的频偏。 下面给出本较佳实施方式的具体实例: 如图 1所示, 光发送端激光器发射光经过 PBS分束器分为两束偏转光, 这两束偏 振光分别经过与电信号的正交调制, 得到两组正交信号经过 PBC合束器, 得到偏振复 用的光信号, 经过信道到达光接收端; 相干接收机接收到发端信号后, 经过 PBS光分 束器将两路偏振光信号分开, 并分别与本振激光器发射的光信号进行解调, 经光电转 换, ADC采集, 得到数字信号, 并进行数字信号处理。 信号进入到 DSP中, 经过粗均衡, 估计色散值, 并进行色散补偿; 对消除色散后 的数据进行时钟恢复和插值滤波等处理, 计算时钟鉴相值反馈给 ADC, 并使得采集得 到的信号恢复到理想的采样点上; 通过盲均衡以便消除残余的色散和偏振模色散, 并 将两个偏振态进行偏振解复用, 即将 X和 Y偏振态进行分离; 由于本振激光器和发端 激光器存在频偏, 激光器本身存在线宽等因素, 需要对数据进行频偏估计和相位估计 的处理, 频偏补偿消除频偏, 相偏补偿消除相偏; 经过相偏补偿的数据已经可以正常 得恢复出星座点, 将经过相偏后的数据送入帧同步子系统进行同步处理。 在帧同步子系统中, 对获取的软数据流进行硬判, 并做差分解码, 与预存的帧头 序列进行滑动相关, 滑动搜索到帧头的位置, 并检测帧头的状态; 进行频偏模糊补偿 及其他操作。 因为随着激光器的老化, 频偏范围会变大, 超过 [- ^74,^/4]的范围, 此时四次 方估计算法和扫频算法只能估计 [- /4, /4]以内的频偏, 因此利用帧头检测的状态 对 /2的整数倍频偏进行估计和补偿。 本较佳实施方式中, DWDM高速光传输系统的相干接收机解调算法中,频偏估计 和补偿算法的框图如图 3所示: 频偏估计得到频偏补偿值,送入频偏补偿值修正模块, 这一步骤的频偏补偿对大于 [- /8,+ /8]范围的频偏会出现频偏模糊 (见算法原理); 帧同步子系统获取软数据流, 经过硬判决, 差分解码; 差分解码后数据已经不存在频 偏模糊信息, 而整体数据的相位变化体现了频偏模糊的值, 通过对这个相位变化的计 算来确定频偏模糊的值; [_ ^/ 4,^/ 4]以内的频偏值在频偏估计模块中已经被补偿掉, 所以在帧同步子系统中频偏模糊的值都为 /2的整数倍, 在一个偏振态内, 有 I和 Q 两路信号, 经过差分解码后, 整体相位为一个固定的偏移量, 可以通过偏振态的偏转 来判断这个相位偏移值, 如果 I路被反转, Q路不变, 则为^ /2, 如果 Q路被反转, I路不变, 则为^ /2, 如果 I和 Q两路都不变, 则为 0, 如果 I和 Q两路都反转, 则 为 ,那么由此判断得到的相位偏移量就是送入帧同步子系统的软数据流的频偏模糊 值, X和 Y偏振态计算频偏模糊的方式同理; 用计算得到的频偏模糊补偿数据流, 同 时将频偏模糊补偿值送到频偏补偿值修正模块, 修正频偏估计模块计算得到的频偏补 偿值; 用修正后的频偏值来控制本振激光器。 频偏模糊估计和补偿算法的实现流程图如图 4所示, 此处只对频偏模糊的估计和 补偿做说明, 具体实施时包括下列步骤: 步骤 401, 硬判决。 帧同步子系统从相偏输出端获取两个偏振态的软数据流, 每个偏振态分别有 I和 Q两路数据, 对软数据进行硬判决, 得到二进制的比特 (bit) 流。 步骤 402, 差分解码。 对一个偏振态内两路 bit流进行差分解码 (DQPSK, differential quadrature phase shift keying), 解码前的 bit流存在频偏模糊, 而解码后的 bit流, 由于是差分的形式, 所以不存在频偏模糊, 但是有一个 /2的整数倍的相位; 对于本地的预存的帧头也进 行差分解码。 步骤 403, 滑动搜索帧头。 对帧头解码后的 bit流与数据解码后的 bit流进行滑动搜索, 每次滑动一个 bit, 比 较两个 bit流有多少相同, 设置帧头长度为 M, 门限为 A, 如果相同的 bit数大于 A或 者小于 M-A, 则表示查找到了帧头, 否则一直查找, 直到找到帧头, 否则退出频偏模 糊估计补偿。 步骤 404, 监测 I和 Q两路的帧头状态。 找到帧头后, 记录帧头的位置, 并检测帧头的状态, 如果相同的 bit数大于 A, 则 这一路没有反转, 如果相同的 bit数小于 M-A, 则表示这一路有反转。 步骤 405, 判断频偏模糊的值。 如果 I路反转、 Q路不反转, 则频偏模糊值为 pi/2; 如果 I路不反转、 Q路反转, 则频偏模糊的值为 -pi/2; 如果 I路和 Q路都反转, 则频偏模糊值为 pi; 如果 I路和 Q 路两路都不反转, 则频偏模糊的值为 0。 步骤 406, 利用得到的频偏模糊的值对软数据进行补偿, 并对频偏估计的值进行 修正, 用修正后的值来反馈控制本振激光器。 本较佳实施方式中一种应用在 DWDM高速光传输系统中, 以对帧头状态的监测 的方式对四次方算法或扫频算法补偿后的频偏模糊进行频偏估计和消除的较优化的方 法, 通过在帧同步子系统中增加对帧头状态的监测和判断的功能, 可实现频偏模糊的 估计补偿, 非常易于 ASIC硬件实现。 图 5 表示本发明实施例提供的一种频偏估计装置的结构框图, 参照图 5, 本发明 实施例还提供一种频偏估计装置, 用于光接收端, 包括: 硬判决单元, 设置为对经过频偏补偿和相偏补偿的帧数据在同一偏振态内对应的 I路和 Q路软数据分别进行硬判决, 得到 I路和 Q路比特数据, 所述频偏补偿基于频 偏估计值进行; 差分解码单元, 与所述硬判决单元连接, 设置为对 I路和 Q路比特数据、 和设置 为对所述第一帧数据进行帧同步的预存帧头分别进行差分解码, 得到差分解码后的 I 路和 Q路比特数据和预存帧头; 帧头状态判断单元, 与所述差分解码单元连接, 设置为根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后的 I路和 Q路比特数据中的帧头分别是否 发生反转, 获取判断结果; 相位偏移量确定单元, 与所述帧头状态判断单元连接,设置为根据所述判断结果, 确定所述帧数据在所述同一偏振态内对应的相位偏移量; 频偏估计值修正单元, 与所述相位偏移量确定单元连接, 设置为将所述相位偏移 量作为频偏估计模糊度对所述频偏估计值进行修正, 得到修正后的频偏估计值。 可见, 通过差分解码的方式来判断同一偏振态内不同路比特数据中帧头的反转情 况,据此确定帧数据对应的相位偏移量,将其作为频偏模糊值对频偏估计值进行修正, 从而不需要增加光传输负荷就能实现频偏估计值的修正。 本发明实施例中, 所述帧头状态判断单元可以包括: 帧头检测子单元, 与所述差分解码单元连接, 设置为根据差分解码后的 I路和 Q 路比特数据和预存帧头, 检测差分解码后的 I路和 Q路比特数据各自中的帧头; 帧头状态判断子单元, 与所述帧头检测子单元连接, 设置为分别判断差分解码后 的 I路和 Q路比特数据各自中的帧头是否发生反转, 获取所述判断结果。 其中, 对于差分解码后的 I路和 Q路比特数据中的任一路比特数据, 所述帧头检 测子单元中, 所述任一路比特数据中的帧头通过如下方式检测得到: 根据差分解码后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜索到的 帧头长度的比特数据与差分解码后的预存帧头相比, 相同的比特数大于门限值或者小 于所述帧头长度与所述门限值之差时, 当前搜索到的帧头长度的比特数据为检测到的 所述任一路比特数据中的帧头; 所述帧头状态判断子单元中, 检测到的所述任一路比特数据中的帧头与差分解码 后的预存帧头相比, 相同的比特数大于门限值, 则表明所述任一路没有发生反转; 相 冋的比特数小于帧头长度与门限值之差, 则表明所述任一路发生反转。 本发明实施例中, 所述相位偏移量确定单元中, 差分解码后的 I路和 Q路比特数 据中的帧头分别发生反转和没有发生反转时, 所述帧数据在所述同一偏振态内对应的 相位偏移量为 pi/2; 差分解码后的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转时, 所 述帧数据在所述同一偏振态内对应的相位偏移量为 -pi/2; 差分解码后的 I路和 Q路比特数据中的帧头均发生反转时, 所述帧数据在所述同 一偏振态内对应的相位偏移量为 pi。 本发明实施例还提供一种帧同步子系统, 该系统设置为 DWDM高速光传输系统 的光接收端, 包括以上所述的频偏估计装置, 还包括: 接收单元, 与所述光接收端中的数据信号处理器 DSP和所述硬判决单元连接, 设 置为从所述 DSP接收经过频偏补偿和相偏补偿的帧数据; 输出单元, 设置为将修正后的频偏估计值输出。 该系统中, 对接收的数据进行硬判决, 并与预存帧头序列进行滑动相关, 找到帧 头序列, 判断帧头状态, 推断频偏模糊值, 并对接收数据进行频偏模糊补偿, 同时输 出频偏模糊值。 以上所述仅是本发明实施例的实施方式, 应当指出, 对于本技术领域的普通技术 人员来说, 在不脱离本发明实施例原理的前提下, 还可以作出若干改进和润饰, 这些 改进和润饰也应视为本发明实施例的保护范围。 工业实用性 如上所述, 通过上述实施例及优选实施方式, 即通过差分解码的方式来判断同一 偏振态内不同路比特数据中帧头的反转情况, 据此确定帧数据对应的相位偏移量, 将 其作为频偏模糊值对频偏估计值进行修正, 从而不需要增加光传输负荷就能实现频偏 估计值的修正。

Claims

权 利 要 求 书
1. 一种频偏估计方法, 用于光接收端, 包括:
对经过频偏补偿和相偏补偿的帧数据在同一偏振态内对应的 I路和 Q路软 数据分别进行硬判决, 得到 I路和 Q路比特数据, 所述频偏补偿基于频偏估计 值进行;
对 I路和 Q路比特数据、 和用于对所述第一帧数据进行帧同步的预存帧头 分别进行差分解码, 得到差分解码后的 I路和 Q路比特数据和预存帧头;
根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后的 I 路和 Q路比特数据中的帧头分别是否发生反转, 获取判断结果; 根据所述判断结果, 确定所述帧数据在所述同一偏振态内对应的相位偏移 将所述相位偏移量作为频偏模糊值对所述频偏估计值进行修正, 得到修正 后的频偏估计值。
2. 根据权利要求 1所述的方法, 其中, 所述根据差分解码后的 I路和 Q路比特数 据和预存帧头, 判断差分解码后的 I路和 Q路比特数据中的帧头分别是否发生 反转, 获取判断结果包括:
根据差分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I 路和 Q路比特数据各自中的帧头; 分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反转, 获取所述判断结果。
3. 根据权利要求 2所述的方法, 其中, 对于差分解码后的 I路和 Q路比特数据中 的任一路比特数据, 所述根据差分解码后的 I路和 Q路比特数据和预存帧头, 检测差分解码后的 I路和 Q路比特数据各自中的帧头的步骤中, 所述任一路比 特数据中的帧头通过如下方式检测得到:
根据差分解码后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜 索到的帧头长度的比特数据与差分解码后的预存帧头相比, 相同的比特数大于 门限值或者小于所述帧头长度与所述门限值之差时, 当前搜索到的帧头长度的 比特数据为检测到的所述任一路比特数据中的帧头; 所述分别判断差分解码后的 I路和 Q路比特数据各自中的帧头是否发生反 转, 获取所述判断结果的步骤中, 检测到的所述任一路比特数据中的帧头与差 分解码后的预存帧头相比, 相同的比特数大于门限值, 则表明所述任一路没有 发生反转; 相同的比特数小于帧头长度与门限值之差, 则表明所述任一路发生 反转。
4. 根据权利要求 1所述的方法, 其中, 所述根据所述判断结果, 确定所述帧数据 在所述同一偏振态内对应的相位偏移量的步骤中, 差分解码后的 I路和 Q路比 特数据中的帧头分别发生反转和没有发生反转时, 所述帧数据在所述同一偏振 态内对应的相位偏移量为 pi/2; 差分解码后的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转 时, 所述帧数据在所述同一偏振态内对应的相位偏移量为 -pi/2; 差分解码后的 I路和 Q路比特数据中的帧头均发生反转时, 所述帧数据在 所述同一偏振态内对应的相位偏移量为 pi。
5. 根据权利要求 1所述的方法, 其中, 所述频偏估计值为根据四次方估计算法或 扫频估计算法对所述帧数据在所述频偏补偿前对应的数据进行频偏估计得到。
6. 根据权利要求 1所述的方法, 其中, 所述修正后的频偏估计值设置为对所述光 接收端的本振激光器进行反馈控制。
7. 根据权利要求 1所述的方法, 其中, 还包括: 根据所述频偏模糊值对 I路和 Q路软数据分别进行频偏模糊补偿后输出。
8. 一种频偏估计装置, 设置为光接收端, 包括: 硬判决单元, 设置为对经过频偏补偿和相偏补偿的帧数据在同一偏振态内 对应的 I路和 Q路软数据分别进行硬判决, 得到 I路和 Q路比特数据, 所述频 偏补偿基于频偏估计值进行;
差分解码单元, 与所述硬判决单元连接, 设置为对 I路和 Q路比特数据、 和设置为对所述第一帧数据进行帧同步的预存帧头分别进行差分解码, 得到差 分解码后的 I路和 Q路比特数据和预存帧头; 帧头状态判断单元, 与所述差分解码单元连接, 设置为根据差分解码后的 I路和 Q路比特数据和预存帧头, 判断差分解码后的 I路和 Q路比特数据中的 帧头分别是否发生反转, 获取判断结果; 相位偏移量确定单元, 与所述帧头状态判断单元连接, 设置为根据所述判 断结果, 确定所述帧数据在所述同一偏振态内对应的相位偏移量; 频偏估计值修正单元, 与所述相位偏移量确定单元连接, 设置为将所述相 位偏移量作为频偏估计模糊度对所述频偏估计值进行修正, 得到修正后的频偏 估计值。
9. 根据权利要求 8所述的装置, 其中, 所述帧头状态判断单元包括: 帧头检测子单元, 与所述差分解码单元连接, 设置为根据差分解码后的 I 路和 Q路比特数据和预存帧头,检测差分解码后的 I路和 Q路比特数据各自中 的帧头; 帧头状态判断子单元, 与所述帧头检测子单元连接, 设置为分别判断差分 解码后的 I路和 Q路比特数据各自中的帧头是否发生反转,获取所述判断结果。
10. 根据权利要求 9所述的装置, 其中, 对于差分解码后的 I路和 Q路比特数据中 的任一路比特数据, 所述帧头检测子单元中, 所述任一路比特数据中的帧头通 过如下方式检测得到:
根据差分解码后的预存帧头, 对所述任一路比特数据进行滑动搜索, 当搜 索到的帧头长度的比特数据与差分解码后的预存帧头相比, 相同的比特数大于 门限值或者小于所述帧头长度与所述门限值之差时, 当前搜索到的帧头长度的 比特数据为检测到的所述任一路比特数据中的帧头; 所述帧头状态判断子单元中, 检测到的所述任一路比特数据中的帧头与差 分解码后的预存帧头相比, 相同的比特数大于门限值, 则表明所述任一路没有 发生反转; 相同的比特数小于帧头长度与门限值之差, 则表明所述任一路发生 反转。
11. 根据权利要求 8所述的装置, 其中, 所述相位偏移量确定单元中, 差分解码后 的 I路和 Q路比特数据中的帧头分别发生反转和没有发生反转时, 所述帧数据 在所述同一偏振态内对应的相位偏移量为 pi/2;
差分解码后的 I路和 Q路比特数据中的帧头分别没有发生反转和发生反转 时, 所述帧数据在所述同一偏振态内对应的相位偏移量为 -pi/2; 差分解码后的 I路和 Q路比特数据中的帧头均发生反转时, 所述帧数据在 所述同一偏振态内对应的相位偏移量为 pi。
2. 一种帧同步子系统, 设置为 DWDM高速光传输系统的光接收端, 包括如权利 要求 8至 11中任一权利要求所述的频偏估计装置, 还包括: 接收单元,与所述光接收端中的数据信号处理器 DSP和所述硬判决单元连 接, 设置为从所述 DSP接收经过频偏补偿和相偏补偿的帧数据; 输出单元, 设置为将修正后的频偏估计值输出。
PCT/CN2014/083831 2014-02-18 2014-08-06 一种频偏估计方法、装置及帧同步子系统 WO2015123970A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410054841.0 2014-02-18
CN201410054841.0A CN104852873B (zh) 2014-02-18 2014-02-18 一种频偏估计方法、装置及帧同步子系统

Publications (1)

Publication Number Publication Date
WO2015123970A1 true WO2015123970A1 (zh) 2015-08-27

Family

ID=53852235

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/083831 WO2015123970A1 (zh) 2014-02-18 2014-08-06 一种频偏估计方法、装置及帧同步子系统

Country Status (2)

Country Link
CN (1) CN104852873B (zh)
WO (1) WO2015123970A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109428844B (zh) * 2017-09-04 2021-07-16 瑞昱半导体股份有限公司 通信装置及通信方法
CN109104390B (zh) * 2018-09-12 2021-01-12 北京睿信丰科技有限公司 一种高速信号的捕获和跟踪方法及装置
CN110266380B (zh) * 2019-04-29 2021-01-26 华中科技大学 一种采用单探测器的光场重建与时频同步系统与方法
CN113179234B (zh) * 2021-04-26 2022-01-18 哈尔滨工程大学 一种基于分块多普勒补偿的高动态长信号累积方法
CN113612711B (zh) * 2021-08-04 2024-04-05 重庆两江卫星移动通信有限公司 一种低信噪比下短突发调制信号的频偏估计方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101646232A (zh) * 2008-07-18 2010-02-10 俊茂微电子(上海)有限公司 频偏估计方法、装置以及通信设备
CN101808068A (zh) * 2009-10-29 2010-08-18 清华大学 联合ldpc码msk迭代解调方法及系统
CN103227661A (zh) * 2012-01-30 2013-07-31 展讯通信(上海)有限公司 频偏估计的方法、频偏估计的装置

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2006083324A1 (en) * 2005-02-02 2006-08-10 Lin Wen T A system and method of detecting a phase, a frequency and an arrival-time difference between signals
KR100749804B1 (ko) * 2005-12-09 2007-08-17 한국전자통신연구원 카테시안 피드백 루프를 이용한 초협대역 송신 시스템의이득 제어 장치
CN101286972B (zh) * 2008-06-02 2012-04-11 北京海尔集成电路设计有限公司 一种载波恢复方法及其系统
EP2506516A1 (en) * 2011-03-31 2012-10-03 Alcatel Lucent Method of decoding optical data signals
CN102412866B (zh) * 2011-11-23 2014-09-10 北京泰美世纪科技有限公司 载波频偏、帧头相位和细定时联合估计的方法和系统

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101646232A (zh) * 2008-07-18 2010-02-10 俊茂微电子(上海)有限公司 频偏估计方法、装置以及通信设备
CN101808068A (zh) * 2009-10-29 2010-08-18 清华大学 联合ldpc码msk迭代解调方法及系统
CN103227661A (zh) * 2012-01-30 2013-07-31 展讯通信(上海)有限公司 频偏估计的方法、频偏估计的装置

Also Published As

Publication number Publication date
CN104852873A (zh) 2015-08-19
CN104852873B (zh) 2019-08-27

Similar Documents

Publication Publication Date Title
US10122470B2 (en) Clock recovery for optical transmission systems
JP6057382B2 (ja) デジタル復調器アーキテクチャ
RU2557012C2 (ru) Модуль оценивания расфазировки, модуль компенсации расфазировки и когерентный приемник
JP5391660B2 (ja) 光コヒーレント受信機、周波数オフセット推定装置、及び光コヒーレント受信機の方法
JP5525552B2 (ja) コヒーレント光信号受信機のデータ・パターン依存歪の補正
US8478137B2 (en) Optical receiver
JP5922670B2 (ja) 光信号受信機において硬判定から軟判定の信頼性情報を生成するためのシステムおよび方法
US8655191B2 (en) Symbol timing recovery in polarization division multiplexed coherent optical transmission system
US8705977B1 (en) Equalizer tap correction and reseeding
US8260154B2 (en) Digital equalization apparatus and method for coherent optical receiver
US8335438B2 (en) Estimating frequency offset using a feedback loop
WO2015123970A1 (zh) 一种频偏估计方法、装置及帧同步子系统
US20120082464A1 (en) Coherent optical receiving apparatus, coherent optical communications system employing same, and coherent optical communications method
US20120148265A1 (en) Coherent optical receiving device capable of digital equalization of optical input, digital equalization method for optical input and coherent optical transmitting/receiving device
JP2013530619A (ja) パイロット・シンボル支援のキャリア位相推定
JP2012510762A (ja) 光通信システムにおけるサイクルスリップの影響を修正する方法、及びコヒーレント受信機
EP2930865B1 (en) Transmitter quadrature imbalance compensation at a coherent optical receiver
US9906307B2 (en) Apparatus for characterizing a chromatic dispersion of an optical receive signal
JPWO2018180912A1 (ja) 信号処理装置、及び信号処理方法
JP2011199657A (ja) 光msk変調/任意偏移量cpfskの光サンプリング復調方法
WO2012144108A1 (ja) 光受信方法および光受信機
JP2014154926A (ja) 受信信号処理装置および受信信号処理方法
WO2014187742A1 (en) Apparatus, method and computer program for recovering a phase of a received signal
EP3605883B1 (en) Method and apparatus for correcting phase jump
WO2016101668A1 (zh) 一种载波相位估计方法、装置及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14882814

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14882814

Country of ref document: EP

Kind code of ref document: A1