WO2018195722A1 - 一种光接收机及延时估计方法 - Google Patents

一种光接收机及延时估计方法 Download PDF

Info

Publication number
WO2018195722A1
WO2018195722A1 PCT/CN2017/081706 CN2017081706W WO2018195722A1 WO 2018195722 A1 WO2018195722 A1 WO 2018195722A1 CN 2017081706 W CN2017081706 W CN 2017081706W WO 2018195722 A1 WO2018195722 A1 WO 2018195722A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
component
frequency
sub
frequency domain
Prior art date
Application number
PCT/CN2017/081706
Other languages
English (en)
French (fr)
Inventor
卢彦兆
李良川
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to CN201780082891.1A priority Critical patent/CN110168967B/zh
Priority to PCT/CN2017/081706 priority patent/WO2018195722A1/zh
Publication of WO2018195722A1 publication Critical patent/WO2018195722A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication

Definitions

  • the embodiments of the present invention relate to the field of communications technologies, and in particular, to an optical receiver and a delay estimation method.
  • the transmission bandwidth and capacity requirements of optical fiber transmission systems are getting higher and higher.
  • the 100G-Wavelength Division Multiplexing (WDM) transmission system has been commercialized on a large scale.
  • the single-wave transmission rate will evolve to a single-wave 400Gbps or even 1Tbps, and high-order modulation and high symbol modulation rate can be used.
  • the single-wave transmission rate is improved, but in the coherent detection system, both the high-order modulation and the high-symbol modulation rate have higher requirements on the in-phase and quadrature (IQ) delay.
  • IQ in-phase and quadrature
  • the modulated optical signal is expressed as
  • the received modulated optical signal is demodulated and photoelectrically converted to obtain an in-phase component I' and a quadrature component Q' of the modulated optical signal in the form of an electrical signal, and the in-phase component I' and the quadrature component are obtained.
  • the delay generated by Q' during the transmission of the respective channels is the orthogonal delay.
  • the prior art uses a 4 ⁇ 4 Multiple Input Multiple Out (MIMO) filter to blindly compensate the orthogonal delay, that is, It is necessary to estimate the orthogonal delay size, and only use the adaptive FIR (Finite Impulse Response) filter 4x4-MIMO for delay compensation.
  • MIMO Multiple Input Multiple Out
  • the FIR filter can converge to the optimal state of the compensated orthogonal delay.
  • the compensation accuracy of this method is related to the complexity of the filter, and it is necessary to update each modulation symbol in real time, which is more complicated than the conventional 2x2-MIMO filter.
  • the delays generated by the in-phase component I' and the quadrature component Q' in the respective channel transmissions may be estimated first, and the obtained orthogonal delay pair in-phase component I' or quadrature component may be used.
  • the delay of Q' is compensated.
  • Embodiments of the present invention provide an optical receiver and a delay estimation method for providing a new optical receiver.
  • an optical receiver In a first aspect, an optical receiver is provided.
  • an analog signal processor is used to photoelectrically process the received optical signal to obtain an electrical signal.
  • the digital signal processor performs time-frequency transform processing on the electrical signal to obtain a frequency domain signal.
  • the first sub-frequency domain signal of the in-phase component of the optical signal and the second sub-frequency domain signal of the orthogonal component of the optical signal are obtained by the delay estimator according to the frequency domain signal, and according to the first sub-frequency domain signal and the second
  • the sub-frequency domain signal calculates the phase difference between the in-phase component and the quadrature component.
  • the delay compensator uses the phase difference to compensate the delay of the optical signal.
  • the frequency domain signal is obtained by the delay estimator, and the first sub-frequency domain signal of the in-phase component of the optical signal received by the optical receiver and the orthogonal component of the received optical signal are obtained according to the frequency domain signal.
  • the second sub-frequency domain signal further calculates a phase difference between the in-phase component and the quadrature component according to the first sub-frequency domain signal and the second sub-frequency domain signal, so that the delay compensator delays the received optical signal by using the phase difference Compensation, it can be seen that the optical receiver first estimates the post-compensation, so there is no need to set a higher complexity filter to blindly compensate the orthogonal delay, thereby reducing the compensation complexity.
  • the delay estimator calculates the phase difference between the in-phase component and the orthogonal component according to the first sub-frequency domain signal and the second sub-frequency domain signal, And: obtaining, according to the first sub-frequency domain signal, a first signal component and a second signal component; wherein the first signal component and the second signal component are used to extract a first phase of the in-phase component Obtaining a third signal component and a fourth signal component according to the second sub-frequency domain signal; wherein the third signal component and the fourth signal component are used to extract a second phase of the orthogonal component And calculating a phase difference between the in-phase component and the quadrature component according to the first signal component, the second signal component, the third signal component, and the fourth signal component.
  • the first phase angle of the in-phase component is extracted by using the first signal component and the second signal component
  • the second phase angle of the orthogonal component is extracted by using the third signal component and the fourth signal component
  • the delay estimator when the first estimator and the second signal component are obtained according to the first sub-frequency domain signal, the delay estimator is specifically configured to: according to the first sub-frequency domain signal a frequency component at the nth frequency point and a frequency component at the n+th frequency point, obtaining the first signal component and the second signal component; the delay estimator is according to the second
  • the frequency domain component is used to: according to the frequency component of the second sub-frequency domain signal at the nth frequency point and at the n+th frequency point
  • the frequency component obtains the third signal component and the fourth signal component; wherein L is at least half of a center frequency of the spectrum of the first sub-frequency domain signal or the second sub-frequency domain signal
  • L is at least half of a center frequency of the spectrum of the first sub-frequency domain signal or the second sub-frequency domain signal
  • L is at least a distance of a half symbol rate from a center frequency of the spectrum of the first sub-frequency domain signal or the second sub-frequency domain signal, and can effectively distinguish the nth frequency.
  • the second phase angle of the quadrature component is at least a distance of a half symbol rate from a center frequency of the spectrum of the first sub-frequency domain signal or the second sub-frequency domain signal, and can effectively distinguish the nth frequency.
  • N fft is the number of points of the time-frequency transform
  • R s is the symbol rate
  • f s is the sampling rate
  • the delay estimator is based on the frequency component of the first sub-frequency domain signal at the n-th frequency point
  • the frequency component at the n+th frequency points when the first signal component and the second signal component are obtained, specifically used to: use the first sub-frequency domain signal in the nth frequency
  • the frequency component at the point is multiplied by the conjugate of the frequency component at the n+L frequency points to obtain the first signal component;
  • the first sub-frequency domain signal is at the n+Lth frequency
  • the frequency component at the point is multiplied by the conjugate of the frequency component at the nth frequency point to obtain the second signal component.
  • the frequency component of the first sub-frequency domain signal at the nth frequency point is multiplied by the conjugate of the frequency component at the n+L frequency points to obtain the first a signal component is obtained by multiplying the amplitude of the first sub-frequency domain signal at the nth frequency point and the n+th frequency point, and the first sub-frequency domain signal at the nth frequency point
  • the phase at the n+L frequency points is subtracted, and the phase difference between the nth frequency point and the n+th frequency point can be relatively easily obtained to extract the first phase angle of the in-phase component.
  • the delay estimator conjugates a frequency component of the first sub-frequency domain signal at the nth frequency point with a frequency component at an n+th frequency point Multiplying, when obtaining the first signal component, specifically for: traversing n from 1 to N, obtaining N first sub-signal components; adding the N first sub-signal components to obtain the first a signal component; the delay estimator frequency of the first sub-frequency domain signal at the n+th frequency points The rate component is multiplied by the conjugate of the frequency component at the nth frequency point, and when the second signal component is obtained, specifically used to: traverse n from 1 to N to obtain N second sub-signal components And adding the N second sub-signal components to obtain the second signal component.
  • the frequency components of the N-pair frequency points are conjugate multiplied, and the N first sub-signal components are added and the N second sub-signal components are added, which is actually the N-pair frequency.
  • the phase difference of the points is averaged and added to improve the accuracy of extracting the first phase angle of the in-phase component according to the first signal component and the second signal component, and the accuracy of the delay estimation can be further improved.
  • L is equal to L 1 +M, Where N fft is the number of points of the time-frequency transform, R s is the symbol rate, f s is the sampling rate, and M is an integer greater than zero; the delay estimator is based on the first sub-frequency domain signal at the nth a frequency component at a frequency point and a frequency component at the n+th frequency point, when the first signal component and the second signal component are obtained, specifically for: for each value of M, Multiplying a frequency component of the first sub-frequency domain signal at the nth frequency point by a conjugate of a frequency component at the n+L 1 +M frequency points to obtain the first signal component, Collecting a plurality of first signal components; for each value of M, frequency components of the first sub-frequency domain signal at the n+L 1 frequency point and at the n+M frequency points Multiplying the conjugate of the frequency components to obtain the second signal component, and obtaining a plurality of second signal components; the delay estimator
  • a plurality of different first signal components, a plurality of different second signal components, a plurality of different third signal components, and a plurality of different fourth signals can be obtained.
  • the component is capable of obtaining an in-phase component and a quadrature component using a plurality of different first signal components, a plurality of different second signal components, a plurality of different third signal components, and a plurality of different fourth signal components A different phase difference, and then a plurality of different phase differences are processed to reduce the error of the phase difference between the in-phase component and the quadrature component, thereby improving the accuracy of the delay estimation.
  • L is equal to L 1 +M, Where N fft is the number of points of the time-frequency transform, R s is the symbol rate, f s is the sampling rate, and M is an integer greater than zero; the delay estimator is based on the first sub-frequency domain signal at the nth
  • the frequency component at the frequency point and the frequency component at the n+th frequency point are obtained when the first signal component and the second signal component are obtained, specifically for: the first value for M, n traversing from 1 to N, multiplying a frequency component of the first sub-frequency domain signal at the nth frequency point by a conjugate of a frequency component at the n+L 1 +M frequency points, Obtaining N first sub-signal components; adding the N first sub-signal components to obtain the first signal component; the first value is any value of M; a value, traversing n from 1 to N, multiplying the frequency component of the first sub-frequency domain signal at the n+L 1 frequency point by the conjugate of the frequency component at the
  • the delay estimator adds the N first sub-signal components when traversing n, obtains the first signal component, and adds the N second sub-signal components to obtain the second signal.
  • Component adding N third sub-signal components, obtaining a third signal component, adding N fourth sub-signal components to obtain a fourth signal component, which reduces extraction of different frequency points of the first sub-frequency domain signal
  • the phase difference between the phase difference and the phase difference between the different frequency points of the second sub-frequency domain signal thereby improving the accuracy of extracting the first phase angle of the in-phase component and the second phase angle of the quadrature component, thereby increasing the delay Estimating accuracy, further, when M takes different values, a plurality of different first signal components, a plurality of different second signal components, a plurality of different third signal components, and a plurality of different fourth signal components can be obtained.
  • multiple phases of the in-phase component and the quadrature component can be obtained by using a plurality of different first signal components, a plurality of different second signal components, a plurality of different third signal components, and a plurality of different fourth signal components difference, And a plurality of different phase difference, so as to reduce the error is eligible to take the phase difference between the in-phase and quadrature components, to further improve the accuracy of the delay estimation from another dimension.
  • the delay estimator calculates the in-phase component and the location according to the first signal component, the second signal component, the third signal component, and the fourth signal component.
  • the method is specifically configured to: obtain the in-phase according to the first signal component and the second signal component corresponding to the same M of the plurality of first signal components and the plurality of second signal components a plurality of first phase angles of components, and obtaining the quadrature component according to the third signal component and the fourth signal component corresponding to the same M of the plurality of third signal components and the plurality of fourth signal components a plurality of second phase angles; subtracting the plurality of first phase angles from the first phase angle and the second phase angle corresponding to the same M of the plurality of second phase angles to obtain a plurality of phase differences; And an average phase difference of the plurality of phase differences, determining that the average phase difference is a phase difference between the in-phase component and the quadrature component.
  • a phase difference between the plurality of in-phase components and the orthogonal components is obtained, and then an average phase difference of the plurality of phase differences is calculated, and the average phase difference is taken as the orthogonal component and the orthogonal component.
  • the phase difference that is, the obtained phase difference is filtered by the filter. Filtering multiple phase differences to reduce the error of obtaining the phase difference between the in-phase component and the quadrature component, thereby improving the accuracy of the delay estimation.
  • the delay compensator performs delay compensation on the optical signal by using the phase difference, specifically, using the phase difference to the first sub-frequency domain signal or The phase of the second sub-frequency domain signal is adjusted to complete the delay compensation of the optical signal.
  • the delay compensator uses the phase difference to adjust the phase of the frequency domain signal in the frequency domain. Since the phase of the frequency domain signal is directly adjusted, the compensation process is relatively simple.
  • the delay compensator uses the phase difference to delay compensation of the optical signal, specifically, the delay obtained by the delay estimator according to the phase difference is used.
  • the delay of the analog converter in the analog signal processor is adjusted to complete the delay compensation of the optical signal.
  • the delay compensator can also adjust the delay of the analog-to-digital converter in the analog signal processor in the time domain by using the delay, because the delay of the analog converter is directly adjusted.
  • the compensation process is relatively simple.
  • a delay estimation method is provided, the steps included in the method being performed by a delay estimator of an optical receiver in the first aspect.
  • a delay estimator comprising a receiving module, an obtaining module and a calculating module, and the module included in the delay estimator is configured to perform the delay estimating method described in the second aspect.
  • a delay estimator in a fourth aspect, includes a processor configured to support the delay estimator to perform the delay in the second aspect. The corresponding function in the time estimation method.
  • the delay estimator can also include a memory coupled to the processor for storing program instructions and data necessary for the delay estimator.
  • Embodiments of the present invention provide an optical receiver in which a frequency domain signal is obtained by a delay estimator, and a first sub-frequency of an in-phase component of an optical signal received by an optical receiver is obtained according to a frequency domain signal. a second sub-frequency domain signal of a quadrature component of the domain signal and the received optical signal, and then calculating a phase difference between the in-phase component and the quadrature component according to the first sub-frequency domain signal and the second sub-frequency domain signal, so that the delay compensation The phase difference is used to compensate the received optical signal for delay compensation. It can be seen that the optical receiver first estimates the post-compensation, so there is no need to set a complex filter to blindly compensate the orthogonal delay. Thereby reducing the compensation complexity.
  • 1 is a system architecture diagram of an optical network system
  • FIG. 2 is a schematic structural diagram of an optical receiver in an optical network system
  • FIG. 3 is a schematic structural diagram of an optical receiver according to an embodiment of the present disclosure.
  • FIG. 5 is a flowchart of a method for estimating a delay according to an embodiment of the present invention.
  • FIG. 6 is a schematic structural diagram of a delay estimator according to an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of a delay estimator according to an embodiment of the present invention.
  • Optical Line Terminal is an important central office equipment.
  • the OLT and the front-end equipment such as the aggregation layer switch, are connected by a network cable, and the OLT is connected to the optical splitter of the user end through a single optical fiber.
  • the OLT can implement functions such as control, management, and ranging of the user equipment, and the user equipment is, for example, an optical network unit (ONU).
  • ONU optical network unit
  • the OLT can convert the received electrical signal into an optical signal, which is an optoelectronic integrated device.
  • An optical network unit is divided into an active optical network unit and a passive optical network unit, and the ONU can also be an optical integrated device.
  • the in-phase component I(t) and the quadrature component Q(t) are ideally orthogonal.
  • an application scenario of the embodiment of the present invention that is, the system architecture of the optical network system, is first introduced.
  • An optical transmitter and an optical receiver are included in Fig. 1, and the optical transmitter transmits the optical signal to the optical receiver without distortion through the optical fiber.
  • the optical network system in the embodiment of the present invention may be a Nyquist system or a Faster-Than-Nyquist (FTN) system, which is not limited in the embodiment of the present invention.
  • FTN Faster-Than-Nyquist
  • the optical receiver is located in the OLT and/or the ONU and includes an analog signal processor and a digital signal processor in the optical receiver:
  • the analog signal processor includes the following parts:
  • 90 degree mixer (90° Hybrid)
  • 90 degree mixer is coupled with four photoelectric converters in the optical receiver, then there are four channels between the 90 degree mixer and the four photoelectric converters, the output 4-channel signal component.
  • light of the same wavelength generates two independent and orthogonal polarized lights through a polarizer in the optical transmitter, and respectively modulates the two polarized lights to obtain two polarized light signals, which are recorded as first polarized light signals.
  • x and second polarized light signal y The optical transmitter transmits the first polarized light signal x and the second polarized light signal y to the optical receiver.
  • the optical receiver receives the first polarized light signal x and the second polarized light signal y through the 90-degree hybrid interface, and outputs four optical signals corresponding to the four-channel signal component, where the four-channel signal component is the first polarized optical signal
  • the in-phase component x i , the orthogonal component x q of the first polarized light signal, the in-phase component y i of the second polarized light signal, and the orthogonal component y q of the second polarized light signal is the first polarized optical signal.
  • An optical to electrical converter coupled to a 90-degree mixer for detecting four optical signals output by a 90-degree mixer and converting four optical signals into four electrical signals After the output.
  • An analog to digital converter coupled to the photoelectric converter for converting four electrical signals output by the photoelectric converter into a digital signal, where the digital signal is also a quantized digital data stream x i , x q , y i and y q .
  • a digital signal processor coupled to an analog to digital converter in the analog signal processor for further processing of the four quantized digital data streams x i , x q , y i , and y q .
  • the optical receiver shown in FIG. 2 cannot guarantee that the lengths of the four channels from the output of the 90-degree mixer to the input of the analog-to-digital converter are equal, that is, the 90-degree mixing cannot be guaranteed.
  • the length of the fiber output to the input of the photoelectric converter, and the length of the cable from the output of the photoelectric converter to the input of the analog-to-digital converter are equal, resulting in a delay in the transmission of the optical signal, thereby
  • the order modulation and high symbol modulation rate have an effect.
  • an embodiment of the present invention provides an optical receiver, after receiving an optical signal transmitted by an optical transmitter, by an optical receiver, a 90-mixer, a photoelectric converter, and an analog-to-digital converter. Processing, obtaining a processed signal, wherein the processed signal is a time domain signal, the processed signal enters a digital signal processor, and the processed signal is subjected to Fast Fourier Transform by a digital signal processor. FFT), that is, converting the time domain signal into a frequency domain signal, the frequency domain signal enters a delay estimator, and the delay estimator obtains the first sub-frequency domain signal and the reception of the in-phase component of the received optical signal according to the frequency domain signal.
  • FFT Fast Fourier Transform
  • the delay compensator uses the phase difference to obtain the delay compensation of the received optical signal, that is, the embodiment of the present invention provides an optical receiver, which is estimated by the delay in the optical receiver.
  • the meter receives the received optical signal The phase difference between the in-phase component and the quadrature component, so that the delay compensator uses the phase difference to delay-compensate the received optical signal without blindly compensating the orthogonal delay by setting a filter with higher complexity, thereby Reduced compensation complexity.
  • an embodiment of the present invention provides an optical receiver, which can be applied to the application scenario shown in FIG. 1.
  • the receiver includes:
  • the analog signal processor 301 is configured to perform photoelectric conversion processing on the received optical signal to obtain an electrical signal.
  • the light of the same wavelength passes through the polarizing plate in the optical transmitter to generate two mutually independent and orthogonal polarized lights, and the optical transmitter separately modulates the two polarized lights to obtain two polarized light signals, which are recorded as the first polarized light.
  • the signal x and the second polarized light signal y the optical transmitter transmits the first polarized light signal x and the second polarized light signal y to the optical receiver.
  • the optical signal received by the optical receiver may include two polarized optical signals, that is, the first polarized optical signal x and the second polarized optical signal y, wherein the first polarized optical signal x corresponds to
  • the signal expressions corresponding to the signal expression and the second polarized light signal y are all in the plural form, and the polarization directions of the first polarized light signal x and the second polarized light signal y are perpendicular to each other.
  • an optical receiver provided by an embodiment of the present invention will be described by taking one of the two polarized optical signals as an example.
  • the analog signal processor includes the 90-degree mixer, the photoelectric converter, and the analog-to-digital converter shown in FIG. 2, wherein the 90-degree mixer in the optical receiver receives the optical signal and outputs Two optical signals corresponding to two channel signal components, for example, the received optical signal is exemplified by a first polarized optical signal x, which is an in-phase component x i and orthogonal of the first polarized optical signal x, respectively.
  • the component x q , the in-phase component x i and the quadrature component x q enter the photoelectric converter, and after photoelectric conversion, two electrical signals are output, and further, the electrical signal output by the photoelectric converter is processed by an analog-to-digital converter, and the output is quantized. Two signals x i and x q .
  • the digital signal processor 302 is configured to perform time-frequency transform processing on the electrical signal to obtain a frequency domain signal.
  • the digital signal processor 302 then receives the electrical signal output by the analog signal processor, where the electrical signal refers to the quantized two signals x i and x q and performs an FFT transformation on the quantized two electrical signals.
  • a delay estimator 303 configured to acquire, according to the frequency domain signal, a first sub-frequency domain signal of an in-phase component of the optical signal and a second sub-frequency domain signal of a quadrature component of the optical signal, and according to the first sub-frequency domain signal and the first The two sub-frequency domain signals calculate the phase difference between the in-phase component and the quadrature component.
  • the delay estimator 303 may be a chip or an integrated circuit, and receives the frequency domain signal output by the digital signal processor 302. A method for the delay estimator 303 to acquire the first sub-frequency domain signal and the second sub-frequency domain signal according to the frequency domain signal is described below.
  • the frequency domain signal after the FFT processing is still a complex signal
  • the delay estimator 303 A frequency domain representation of the in-phase component x i of the received optical signal and the orthogonal component x q of the received optical signal can be obtained by mathematically transforming the frequency domain signal.
  • the delay estimator 303 adds the conjugate of the sequence of the frequency domain signal to the conjugate of the symmetric sequence of the sequence and divides by 2 to obtain the first sub-frequency domain signal of the in-phase component x i , ie
  • the delay estimator 303 subtracts the conjugate of the frequency domain signal from the conjugate of the symmetric sequence of the sequence and divides by 2j to obtain a second sub-frequency domain signal of the orthogonal component x q , ie
  • FXI(N fft ) is the frequency domain signal of the in-phase component x i of the time domain signal x
  • FXQ(N fft ) is the frequency domain signal of the orthogonal component x q of the time domain signal x.
  • FX R (N fft ) is a symmetric sequence of FX, wherein the coordinate 1 on the FX sequence corresponds to the center frequency of the FX spectrum.
  • the digital signal processor 302 if the quantized two signals x i and x q are directly subjected to FFT transformation, the frequency domain signal FXI and the time domain signal x of the time domain signal x i are obtained.
  • the frequency domain signal FXQ of q is such that, in the delay estimator 303, the above-described step of mathematically transforming the frequency domain signal is not performed.
  • the delay estimator 303 acquires the first sub-frequency domain signal and the second sub-frequency domain signal, if the phase difference between the in-phase component and the quadrature component is to be calculated, the first component of the in-phase component is obtained. a phase angle and a second phase angle of the quadrature component. Specifically, the delay estimator 303 needs to calculate, according to the first sub-frequency domain signal, two signal components, a first signal component and a second signal component, which are related to the delay.
  • the signal component acquires the phase difference between the in-phase component and the quadrature component.
  • the optical receiver provided by the embodiment of the invention can be applied to the Nyquist system, and can also be applied to the super Nyquist system. Among them, when applied to the super Nyquist system, Inter-Symbol Interference (ISI) is caused by the super Nyquist system transmitting at a super Nyquist symbol rate.
  • ISI Inter-Symbol Interference
  • the analog signal processor 301 in the optical receiver performs photoelectric conversion processing on the received optical signal to obtain an electrical signal.
  • the electrical signal enters the digital signal processor 302, and the digital signal processor 302 performs time-frequency transform processing on the electrical signal, that is, performs FFT transformation on the complex signal x to obtain a frequency domain signal FX of x, which is used by the delay estimator 303.
  • the domain signal FX acquires the phase difference between the in-phase component and the quadrature component of the received optical signal. Since the number of frequency-frequency transforms is not fixed, the more the number of frequency points, the better the original electrical signal can be recovered. The phase angles of the in-phase component and the quadrature component can be better extracted to improve the accuracy of the delay estimation.
  • the delay estimator 303 may also adopt a manner of extracting the first phase angle of the in-phase component by using the first signal component or the second signal component, and the foregoing is to adopt the first signal.
  • the manner in which the component and the second signal component extract the first phase angle of the in-phase component is because the first phase angle of the in-phase component may not be accurately extracted due to the presence of inter-code crosstalk in the super Nyquist system.
  • the delay estimator 303 extracts the first phase angle of the in-phase component by extracting the phase angle by the first signal component or the second signal component, it is only necessary to calculate the first signal component by using the first sub-frequency domain signal, without simultaneously calculating The first signal component and the second signal component, therefore, can reduce computational complexity.
  • delay estimation 303 can be taken
  • the manner in which the phase angle is extracted by the first signal component and the second signal component may also be employed in such a manner that the phase angle is extracted by the first signal component or the second signal component, and may be selected according to circumstances.
  • the delay estimator 303 may adopt a manner of extracting the first phase angle of the in-phase component by using the first signal component or the second signal component, Or extracting the first phase angle of the in-phase component by either of the two methods of extracting the first phase angle of the in-phase component by the first signal component and the second signal component.
  • a method of acquiring the signal component by the delay estimator 303 will be described below.
  • the first signal component and the second signal component are taken as an example.
  • the third signal component and the fourth signal component can also be obtained by using the method described below, and no further description is provided.
  • the delay estimator 303 can obtain the first signal component and the frequency component according to the frequency component of the first sub-frequency domain signal at the nth frequency point and the frequency component at the n+th frequency point. Two signal components.
  • the frequency components of the first sub-frequency domain signal at the nth frequency point and the frequency components at the n+th frequency points all exhibit a complex form including amplitude and phase.
  • L is at least a distance of half symbol rate from the spectral center frequency of the first sub-frequency domain signal
  • L is a fixed value
  • N fft is the number of points of the time-frequency transform
  • R s is the symbol rate
  • how many different values of M can be selected according to the accuracy requirement of the delay estimation, and if the accuracy of the delay estimation is high, Then M has a larger number of values, that is, M can take more values. If the accuracy of the delay estimation is low, the number of the values of the M is small, and those skilled in the art need to determine according to the actual situation, which is not limited in the embodiment of the present invention.
  • the calculation method of L 1 is the same as the first possible implementation of L, and will not be repeated.
  • the delay estimator 303 obtains the first signal component and the second component according to the frequency component of the first sub-frequency domain signal at the nth frequency point and the frequency component at the n+th frequency point.
  • the signal components can be implemented in various ways, which will be exemplified below.
  • L L 1
  • L L 1 + M
  • M is an integer.
  • the delay estimator 303 multiplies the frequency component of the first sub-frequency domain signal at the nth frequency point by the conjugate of the frequency component at the n+L frequency points, Can obtain the first signal component, ie
  • the delay estimator 303 multiplies the frequency component of the first sub-frequency domain signal at the n+L frequency points by the conjugate of the frequency component at the nth frequency point, thereby obtaining the second signal component, ie
  • phase angle of the first sub-frequency domain signal at the nth frequency point is The phase angle at the n+L frequency points is The complex conjugate is the real part, the imaginary part is inverted, that is, the phase is opposite.
  • conjugate multiplication of A XI in the expression (1) can be understood as the first sub-frequency domain signal at the nth Multiplying the amplitude at the frequency point and the n+L frequency points, and subtracting the phase of the first sub-frequency domain signal at the nth frequency point and the n+L frequency points to obtain an in-phase component
  • the phase angle is Wherein, the interval between the nth frequency point and the n+th frequency point is subtracted from the phase of the first sub-frequency domain signal at the nth frequency point and the n+th frequency point, and the interval between the nth frequency point and the n+th frequency point can be understood.
  • L is the distance between the center frequency of the spectrum of the first sub-frequency domain signal and the frequency point of the half-symbol rate on the negative spectrum.
  • the frequency point n shown in FIG. 4 is The center frequency of the spectrum of the first sub-frequency domain signal, that is, the coordinate 1 corresponding to the first sub-frequency domain signal.
  • the conjugate multiplication of B XI in the expression (2) can be understood as multiplying the amplitude of the first sub-frequency domain signal at the n+L frequency points and at the nth frequency point, and will be the first
  • the sub-frequency domain signal is subtracted from the phase at the n+th frequency point and the nth frequency point to obtain a first phase angle of the in-phase component, and the phase angle is
  • L between the n+L frequency points and the n-th frequency point That is, the distance between the center frequency point of the spectrum of the first sub-frequency domain signal and the frequency point at the half symbol rate position of the positive frequency spectrum.
  • the first signal component is the multiplication of the frequency component of the first sub-frequency domain signal at the n+th frequency point and the conjugate of the frequency component at the nth frequency point
  • the L between the n+th frequency point and the nth frequency point is the distance between the center frequency point of the spectrum of the first sub-frequency domain signal and the frequency point at the half symbol rate position of the positive frequency spectrum.
  • the second sub-signal component is obtained by multiplying the frequency component of the first sub-frequency domain signal at the nth frequency point by the conjugate of the frequency component at the n+L frequency points, then, the nth frequency point and the The L between the n+L frequency points is also the distance between the center frequency point of the spectrum of the first sub-frequency domain signal and the frequency point at the half symbol rate position on the negative frequency spectrum.
  • the calculation of the first signal component and the second signal component may be selected according to actual needs, as long as the interval between two frequency points for calculating the first signal component and the two frequency points for calculating the second signal component are guaranteed
  • the interval between the first sub-frequency domain signals may be separated by a symbol rate on the positive and negative spectrum.
  • the delay estimator 303 extracts the first phase angle of the in-phase component using the first signal component.
  • the first phase angle of the in-phase component can also be extracted using the second signal component And the phase angles extracted are only different in sign, which increases the probability of accurately extracting the phase angle, so as to improve the accuracy of the delay estimation.
  • the acquisition process of the third signal component and the fourth signal component of the second sub-frequency domain signal is the same as the first signal component and the second signal component, and will not be further described herein.
  • the delay estimator 303 simply conjugates a pair of frequency points, for example, the frequency components of the nth frequency point and the n+th frequency points.
  • the delay estimator 303 traverses n from 1 to N, that is, conjugates the frequency components of the N pairs of frequency points to obtain N first sub-signal components, and adds N first sub-signal components. Obtaining the first signal component and obtaining the N second sub-signal components, adding the N second sub-signal components to obtain the second signal component.
  • the frequency components of the N-pair frequency points are conjugate multiplied, and the N first sub-signal components are added and the N second sub-signal components are added, which is actually the phase difference of the N-pair frequency points.
  • the averaging is performed to improve the accuracy of extracting the first phase angle of the in-phase component according to the first signal component and the second signal component, thereby further improving the accuracy of the delay estimation.
  • the delay estimator 303 traverses n from 1 to N, may traverse in the order of n from small to large, or may traverse in the order of n from large to small, or may follow n from the middle.
  • the traversal is performed in the order of the two sides, or traversed in other manners, which is not limited in the embodiment of the present invention.
  • the first signal component and the second signal component are not repeated here.
  • L is equal to L 1 +M, and M can take different values.
  • the delay estimator 303 multiplies the frequency component of the first sub-frequency domain signal at the nth frequency point by the conjugate of the frequency component at the n+L 1 +2 frequency points to obtain the first signal component. And multiplying the frequency component at the n+L 1 frequency point by the conjugate of the frequency component at the n+2th frequency point to obtain the second signal component, so that the two first signal components are obtained in total And two second signal components.
  • the acquisition process of the third signal component and the fourth signal component of the second sub-frequency domain signal is the same as the first signal component and the second signal component, and will not be further described herein.
  • the delay estimator 303 adds the N first sub-signal components to obtain the first signal component and the N second sub-subscores when n is traversed.
  • the signal components are added to obtain a second signal component, which is for reducing the error of extracting the phase difference between different frequency points of the first sub-frequency domain signal, so as to improve the accuracy of extracting the first phase angle of the in-phase component, and further Increasing the accuracy of the delay estimation, and in the third implementation, when the M takes different values, a plurality of different first signal components, a plurality of different second signal components, and a plurality of different third signals can be obtained.
  • a component and a plurality of different fourth signal components thereby being able to utilize a plurality of different first signal components, a plurality of different second signal components, a plurality of different third signal components, and a plurality of different fourth signal components, Obtaining a plurality of different phase differences of the in-phase component and the quadrature component, and further processing the plurality of different phase differences to reduce the error of acquiring the phase difference between the in-phase component and the quadrature component, that is, from another dimension The accuracy of high latency estimation.
  • L and M are the same as the third implementation.
  • the delay estimator 303 traverses n from 1 to N, and the frequency component of the first sub-frequency domain signal at the nth frequency point and the n+L 1 +1 frequency point.
  • the conjugate of the frequency component is multiplied to obtain N first sub-signal components, and the N first sub-signal components are added to obtain a first signal component, and the frequency component at the n+L 1 frequency point is
  • the conjugate of the frequency components at the n+1th frequency point is multiplied to obtain N second sub-signal components, and the N second sub-signal components are added to obtain a second signal component.
  • the delay estimator 303 traverses n from 1 to N, and the frequency component of the first sub-frequency domain signal at the nth frequency point and the n+L 1 +2 frequency points. Multiplying the conjugate of the frequency components to obtain a first signal component, multiplying the frequency component at the n+L 1 frequency point by the conjugate of the frequency component at the n+2th frequency point, and obtaining N
  • the second sub-signal component adds the N second sub-signal components to obtain a second signal component, such that two first signal components and two second signal components are obtained in total.
  • the acquisition process of the third signal component and the fourth signal component of the second sub-frequency domain signal is the same as the first signal component and the second signal component, and will not be further described herein.
  • the delay estimator 303 adds the N first sub-signal components when traversing n, Obtaining a first signal component, adding N second sub-signal components to obtain a second signal component; adding N third sub-signal components, obtaining a third signal component, and adding N fourth sub-signal components Obtaining a fourth signal component, which reduces an error of extracting a phase difference between different frequency points of the first sub-frequency domain signal and extracting a phase difference between different frequency points of the second sub-frequency domain signal, thereby improving extraction The accuracy of the first phase angle of the in-phase component and the second phase angle of the quadrature component, thereby improving the accuracy of the delay estimation, and further, when the M takes different values, a plurality of different first signal components and a plurality of different a second signal component, a plurality of different third signal components, and a plurality of different fourth signal components, thereby being capable of utilizing a plurality of different first signal components, a plurality of different second signal components, and
  • the phase angle of the in-phase component can be obtained according to the first signal component and the second signal component, where the in-phase component is The phase angle is called the first phase angle and is expressed as For example, if the first signal component is The second signal component is Then the first phase angle is
  • the delay estimator 303 acquires a plurality of first signal components and a plurality of second signal components of the first sub-frequency domain signal, and further utilizes the plurality of first signal components and the plurality of second signal components.
  • a first signal component and a second signal component corresponding to the same M are obtained, and a plurality of first phase angles of the in-phase component are obtained.
  • the first signal component and the second signal component corresponding to the same M refer to the frequency component of the first sub-frequency domain signal at the nth frequency point and the n+L 1 when the first value is taken by M
  • the delay estimator 303 obtains the first signal component of the first sub-frequency domain signal. And obtaining a second signal component of the first sub-frequency domain signal.
  • the delay estimator 303 obtains the first signal component of the first sub-frequency domain signal. And obtaining a second signal component of the first sub-frequency domain signal. Then, with That is, the first signal component and the second signal component corresponding to the same M, with That is, the first signal component and the second signal component corresponding to the same M.
  • the delay estimator 303 thus utilizes the first signal component And second signal component Obtaining the first phase angle of the in-phase component Using the first signal component And second signal component Obtaining the first phase angle of the in-phase component
  • the phase angle of the orthogonal component can be obtained according to the third signal component and the fourth signal component, which will be orthogonalized herein.
  • the phase angle of the component is called the second phase angle and is expressed as
  • the delay estimator 303 acquires the plurality of third signal components and the plurality of fourth signal components of the second sub-frequency domain signal, and further utilizes the plurality of third signal components and the plurality of fourth signal components
  • the phase difference between the in-phase component and the quadrature component is obtained in the delay estimator 303. After that, you can also The delay T between the in-phase component and the quadrature component is calculated, where f is the frequency of the received optical signal. After the delay time T is calculated, the received optical signal is subsequently compensated by the delay T.
  • the delay compensator 304 is configured to delay compensation of the optical signal by using the phase difference.
  • the frequency domain signals FXI(N) and FXQ(N) may be phase-adjusted by using the obtained delay time T in the frequency domain to achieve the purpose of delay compensation for the received optical signal, or It is also possible to delay the received optical signal in the time domain, which is described below.
  • the delay compensator 304 may be a phase shifter, and the phase shifter passes the phase difference of the in-phase component and the quadrature component to the first sub-frequency domain signal FXI or the second sub-frequency.
  • the phase of the domain signal FXQ is adjusted to achieve delay compensation for the received optical signal.
  • the delay compensator 304 can also be a clock, and the time delay T calculated by the clock using the delay estimator can be used for the analog-to-digital converter in the analog signal processor 301.
  • the adjustment is extended to achieve the purpose of delay compensation for the received optical signal.
  • the delay compensator 304 uses the delay time T to delay-adjust the quantized signal x i or x q through the delay circuit to achieve delay compensation for the received optical signal, for example, by quantization.
  • the signal x i is taken as an example.
  • the input signal input to the delay circuit is x i (tT).
  • the output of the delay circuit is x i (t)
  • the output signal x i (t) is The signal after the quadrature delay compensation.
  • a delay estimation method is provided by an embodiment of the present invention.
  • the method can be implemented by a delay estimator, and the delay estimator is used as a part of an optical receiver.
  • the process of the method is described as follows:
  • Step 501 The delay estimator receives a frequency domain signal that is subjected to time-frequency transform processing on the received optical signal by the optical receiver.
  • Step 502 The delay estimator acquires, according to the frequency domain signal, a first sub-frequency domain signal of an in-phase component of the optical signal and a second sub-frequency domain signal of a quadrature component of the optical signal;
  • Step 503 The delay estimator calculates a phase difference between the in-phase component and the orthogonal component according to the first sub-frequency domain signal and the second sub-frequency domain signal, and the phase difference between the in-phase component and the quadrature component is used.
  • the optical receiver performs a delay estimation on the optical signal.
  • steps 501 to 503 have been introduced in the description of the delay estimator provided in the embodiment shown in FIG. 3, and details are not described herein again.
  • an embodiment of the present invention provides a delay estimator that includes a network interface 601 and a processor 602 that are connected to the same bus 600.
  • the processor 602 may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), and may be one or more integrated circuits for controlling program execution, and may be a baseband chip, etc. .
  • CPU central processing unit
  • ASIC application specific integrated circuit
  • the network interface 601 can be coupled to the processor 602 via bus 600 (e.g., as shown in FIG. 6) or to a processor via a dedicated connection line.
  • the delay estimator can also include a memory that can be coupled to processor 602 via bus 600.
  • the number of memories may be one or more, and the memory may be a read only memory (ROM), a random access memory (RAM), or a disk storage, and the like.
  • the code corresponding to the foregoing delay estimation method is solidified into the chip, so that the chip can perform the delay estimation method provided by the foregoing embodiment shown in FIG. 5 during operation, how to The processor 602 performs design programming and is well known to those skilled in the art, and details are not described herein again.
  • an embodiment of the present invention provides a delay estimator, which includes a receiving module 701, an obtaining module 702, and a calculating module 703.
  • the entity device corresponding to the obtaining module 702 and the calculating module 703 may be integrated into the processor 602 in FIG. 6, and the physical device corresponding to the receiving module 701 may be integrated in the network interface 601 in FIG.
  • Embodiments of the present invention provide an optical receiver in which a frequency domain signal is obtained by a delay estimator, and a first sub-frequency of an in-phase component of an optical signal received by an optical receiver is obtained according to a frequency domain signal. a second sub-frequency domain signal of a quadrature component of the domain signal and the received optical signal, and then calculating a phase difference between the in-phase component and the quadrature component according to the first sub-frequency domain signal and the second sub-frequency domain signal, so that the delay compensation The phase difference is used to compensate the received optical signal for delay compensation. It can be seen that the optical receiver first estimates the post-compensation, so there is no need to set a complex filter to blindly compensate the orthogonal delay. Thereby reducing the compensation complexity.

Landscapes

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

Abstract

一种光接收机及延时估计方法，该光接收机包括：模拟信号处理器，用于对接收的光信号进行光电转换处理，获得电信号；数字信号处理器，用于对电信号进行时频变换处理，获得频域信号；延时估计器，用于根据所述频域信号，获取所述光信号的同相分量的第一子频域信号和所述光信号的正交分量的第二子频域信号，以及根据所述第一子频域信号和所述第二子频域信号计算所述同相分量和所述正交分量的相位差；延时补偿器，用于利用所述相位差对所述光信号进行延时补偿。

Description

一种光接收机及延时估计方法 技术领域
本发明实施例涉及通信技术领域，尤其涉及一种光接收机及延时估计方法。
背景技术
随着数据通信业务的急速膨胀，对光纤传输系统的传输带宽和容量要求越来越高。目前，100G-波分复用(Wavelength Division Multiplexing，WDM)传输系统已经大规模商用，下一步单波传输速率将演进至单波400Gbps，甚至1Tbps，而高阶调制和高符号调制速率就可用来提升单波传输速率，但在相干检测系统中，高阶调制和高符号调制速率均对同相分量与正交分量之间(In-phase and quadrature，IQ)延时的要求较高，其中，若调制光信号表示为
Figure PCTCN2017081706-appb-000001
则I(t)＝Acos2πft典型地被称为同相分量，而Q(t)＝Asin2πft典型地被称为正交分量，其中fc为载波频率，f为基带信号频率。在光接收机端，对接收的调制光信号进行解调及光电转换处理，获得形式为电信号的调制光信号的同相分量I′及正交分量Q′，而同相分量I′和正交分量Q′在各自通道传输过程中产生的延迟，即为正交延时。
那么，为了降低正交延时对单波传输速率的影响，现有技术中采用4×4的多输入多输出(Multiple Input Multiple Out，MIMO)滤波器对正交延时进行盲补偿，即不需要预估正交延时大小，仅通过自适应FIR(Finite Impulse Response)滤波器4x4-MIMO进行延时补偿，通过构建代价函数，FIR滤波器可以自行收敛至补偿正交延时的最佳状态。但这种方法的补偿精度与滤波器复杂度相关，并且需要对每个调制符号进行实时更新，相比传统的2x2-MIMO滤波器复杂度要高。那么，为了降低补偿的复杂度，可以先对同相分量I′和正交分量Q′在各自通道传输过程中产生的延迟进行估计，利用得到的正交延时对同相分量I′或正交分量Q′的延时进行补偿。
而目前，现有技术中尚且没有提出较好地对正交延时进行估计的光接收机。
发明内容
本发明实施例提供一种光接收机及延时估计方法，用以提供一种新的光接收机。
第一方面，提供一种光接收机。在该光接收机中，模拟信号处理器用于对接收的光信号进行光电处理，获得电信号。然后由数字信号处理器对电信号进行时频变换处理，获得频域信号。之后，由延时估计器，根据频域信号获取光信号的同相分量的第一子频域信号和光信号的正交分量的第二子频域信号，以及根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差。最后，由延时补偿器利用相位差对光信号进行延时补偿。
本发明实施例中，由延时估计器获得频域信号，并根据频域信号获得光接收机接收的光信号的同相分量的第一子频域信号和接收的光信号的正交分量的第二子频域信号，进而根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差，使得延时补偿器利用该相位差对接收的光信号进行延时补偿，可以看到，这种光接收机中是先估计后补偿，因此无需设置复杂度较高的滤波器对正交延时进行盲补偿，从而降低了补偿复杂度。
在一个可能的设计中，所述延时估计器在根据所述第一子频域信号和所述第二子频域信号，计算所述同相分量和所述正交分量的相位差时，具体用于：根据所述第一子频域信号，获得第一信号分量和第二信号分量；其中，所述第一信号分量和所述第二信号分量用于提取所述同相分量的第一相位角；根据所述第二子频域信号，获得第三信号分量和第四信号分量；其中，所述第三信号分量和所述第四信号分量用于提取所述正交分量的第二相位角；根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差。
在本发明实施例中，利用第一信号分量和第二信号分量提取同相分量的第一相位角，利用第三信号分量和第四信号分量提取正交分量的第二相位角，能够降低因存在码间串扰而无法准确提取同相分量和正交分量的相位角的影响，从而可以提高提取同相分量和正交分量的相位角的精确度，进而可以提高延时估计的精确度。
在一个可能的设计中，所述延时估计器在根据所述第一子频域信号，获得第一信号分量和第二信号分量时，具体用于：根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量；所述延时估计器在根据所述第二子频域信号，获得第三信号分量和第四信号分量时，具体用于：根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量；其中，L至少为与所述第一子频域信号或所述第二子频域信号的频谱的中心频点间隔半个符号率的距离，n为大于零且小于N的整数，N等于所述时频变换的频点数减去L加1。
在本发明实施例中，L至少为与所述第一子频域信号或所述第二子频域信号的频谱的中心频点间隔半个符号率的距离，能够有效地区分第n个频点的频率分量和第n+L个频点的频率分量，从而可以有效地利用第一信号分量和第二信号分量提取同相分量的第一相位角，利用第三信号分量和第四信号分量提取正交分量的第二相位角。
在一个可能的设计中，
Figure PCTCN2017081706-appb-000002
其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率；所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量；将所述第一子频域信号在所述第n+L个频点处的频率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量。
本发明实施例中，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量，就是为将第一子频域信号在第n个频点处和第n+L个频点处的幅度相乘，及将第一子频域信号在第n个频点处和第n+L个频点处的相位相减，能够较为容易获得第n个频点和第n+L个频点处的相位差，以提取同相分量的第一相位角。
在一个可能的设计中，所述延时估计器将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量时，具体用于：将n从1遍历至N，获得N个第一子信号分量；将所述N个第一子信号分量相加，获得所述第一信号分量；所述延时估计器将所述第一子频域信号在所述第n+L个频点处的频 率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量时，具体用于：将n从1遍历至N，获得N个第二子信号分量；将所述N个第二子信号分量相加，获得所述第二信号分量。
在本发明实施例中，对N对频点的频率分量进行共轭相乘，并将获得N个第一子信号分量相加及N个第二子信号分量相加，实际就是对N对频点的相位差进行相加后求平均，以提高根据第一信号分量和第二信号分量提取同相分量的第一相位角的精确度，进一步可以提高延时估计的精确度。
在一个可能的设计中，L等于L1+M，
Figure PCTCN2017081706-appb-000003
其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：对于M的每个取值，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第一信号分量，共获得多个第一信号分量；对于M的每个取值，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第二信号分量，共获得多个第二信号分量；所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量时，具体用于：对于M的每个取值，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第三信号分量，共获得多个第三信号分量；对于M的每个取值，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第四信号分量，共获得多个第四信号分量。
在本发明实施例中，在M取不同值时，能够获得多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，进而能够利用多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，获得同相分量和正交分量的多个不同的相位差，进而对多个不同的相位差进行处理，以降低取获同相分量和正交分量之间的相位差的误差，从而提高延时估计的精确度。
在一个可能的设计中，L等于L1+M，
Figure PCTCN2017081706-appb-000004
其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N个第一子信号分量；将所述N个第一子信号分量相加，获得所述第一信号分量；所述第一取值为M的任意一个取值；对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第二子信号分量；将所述N个第二子信号分量相加，获得所述第二信号分量；对于M的多个取值，获得多个第一信号分量及多个第二信号分量；所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三 信号分量和所述第四信号分量时，具体用于：对于M的第一取值，n从1遍历至N，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N第三子信号分量，及，将所述N个第三子信号分量相加，获得所述第三信号分量；对于M的第一取值，n从1遍历至N，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第四子信号分量，及将所述N个第四子信号分量相加，获得所述第四信号分量；对于M的多个取值，获得多个第三信号分量及多个第四信号分量。
在本发明实施例中，延时估计器在对n进行遍历时，对N个第一子信号分量相加，获得第一信号分量，对N个第二子信号分量相加，获得第二信号分量；对N个第三子信号分量相加，获得第三信号分量，对N个第四子信号分量相加，获得第四信号分量，这降低了提取第一子频域信号不同频点之间的相位差及第二子频域信号不同频点之间的相位差的误差，从而可以提高提取同相分量的第一相位角及正交分量的第二相位角的准确度，进而提高延时估计精确度，进一步，M取不同值时，能够获得多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，进而能够利用多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，获得同相分量和正交分量的多个相位差，进而对多个不同的相位差进行处理，以降低取获同相分量和正交分量之间的相位差的误差，从另外一个维度进一步提高延时估计的精确度。
在一个可能的设计方式中，所述延时估计器根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差时，具体用于：根据所述多个第一信号分量和所述多个第二信号分量中对应相同M的第一信号分量和第二信号分量，获得所述同相分量的多个第一相位角，及，根据所述多个第三信号分量和所述多个第四信号分量中对应相同M的第三信号分量和第四信号分量，获得所述正交分量的多个第二相位角；将所述多个第一相位角与所述多个第二相位角中对应相同M的第一相位角和第二相位角相减，获得多个相位差；计算所述多个相位差的平均相位差，确定所述平均相位差为所述同相分量和所述正交分量的相位差。
在本发明实施例中，M取不同值时，获得多个同相分量和正交分量的相位差，然后计算多个相位差的平均相位差，将平均相位差作为正交分量和正交分量的相位差，也就是对获得的多个相位差通过滤波器进行滤波处理。对多个相位差进行滤波，以降低获取同相分量和正交分量之间的相位差的误差，进而提高延时估计的精确度。
在一个可能的设计方式中，所述延时补偿器利用所述相位差对所述光信号进行延时补偿时，具体用于：利用所述相位差对所述第一子频域信号或所述第二子频域信号的相位进行调整，以完成对所述光信号的延时补偿。
在本发明实施例中，延时补偿器利用相位差在频域对频域信号进行相位调整，由于是直接对频域信号的相位进行调整，补偿过程较为简单。
在一个可能的设计方式中，所述延时补偿器利用所述相位差对所述光信号进行延时补偿时，具体用于：利用由所述延时估计器根据所述相位差获得的延时，对所述模拟信号处理器中的模拟转换器的时延进行调整，以完成对所述光信号的延时补偿。
在本发明实施例中，延时补偿器还可以利用延时，在时域对模拟信号处理器中的模数转换器的时延进行调整，由于是直接对模拟转换器的时延进行调整，补偿过程较为简单。
第二方面，提供一种延时估计方法，该方法中包括的步骤由第一方面中光接收机的延时估计器执行。
第三方面，提供一种延时估计器，该延时估计器包括接收模块、获得模块及计算模块，延时估计器所包括的模块用于执行第二方面中所述的延时估计方法。
第四方面，提供一种延时估计器，在一个可能的设计中，该延时估计器的结构中包括处理器，所述处理器被配置为支持延时估计器执行第二方面中的延时估计方法中相应的功能。所述延时估计器还可以包括存储器，所述存储器与处理器耦合，用于保存延时估计器必要的程序指令和数据。
本发明实施例提供了一种光接收机，在该光接收机中，由延时估计器获得频域信号，并根据频域信号获得光接收机接收的光信号的同相分量的第一子频域信号和接收的光信号的正交分量的第二子频域信号，进而根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差，使得延时补偿器利用该相位差对接收的光信号进行延时补偿，可以看到，这种光接收机中是先估计后补偿，因此无需设置复杂度较高的滤波器对正交延时进行盲补偿，从而降低了补偿复杂度。
附图说明
图1为光网络系统的系统架构图；
图2为光网络系统中的光接收机的结构示意图；
图3为本发明实施例提供的一种光接收机的结构示意图；
图4为本发明实施例中L的示意图；
图5为本发明实施例提供的一种延时估计方法的流程图；
图6为本发明实施例提供的延时估计器的一种结构示意图；
图7为本发明实施例提供的延时估计器的一种结构示意图。
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，下面将结合附图对本发明实施例作进一步详细描述。
以下，对本发明实施例中的部分用语进行解释说明，以便于本领域技术人员理解。
(1)光线路终端(Optical Line Terminal，OLT)，是重要的局端设备，OLT与前端设备，例如汇聚层交换机，通过网线相连，以及，OLT通过单根光纤与用户端的分光器连接。OLT可以实现对用户端设备的控制、管理、测距等功能，其中，一种用户端设备例如为光网络单元(Optical Network Unit，ONU)。且OLT能够把接收到的电信号转化为光信号，是光电一体设备。
(2)光网络单元(Optical Network Unit，ONU)，分为有源光网络单元和无源光网络单元，ONU也可以为光电一体设备。
(3)同相分量和正交分量，若将调制的光信号表示为
Figure PCTCN2017081706-appb-000005
则I(t)＝Acos(2πft)典型地被称为同相分量，而Q(t)＝Asin(2πft)典型地被称为正交分量，其中fc为载波频率，f为基带信号频率，同相分量I(t)和正交分量Q(t)是理想地正交的。
(4)另外，本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B 这三种情况。另外，本文中字符“/”，如无特殊说明，一般表示前后关联对象是一种“或”的关系。且在本发明实施例的描述中，“第一”、“第二”等词汇，仅用于区分描述的目的，而不能理解为指示或暗示相对重要性，也不能理解为指示或暗示顺序。
为更好地介绍本发明实施例中的光接收机，下面首先介绍本发明实施例的一种应用场景，即光网络系统的系统架构，请参见图1。在图1中包括光发射机和光接收机，光发射机通过光纤将光信号无失真地传送到光接收机。其中，本发明实施例中的光网络系统可以为奈奎斯特系统，也可以为超奈奎斯特(Faster-Than-Nyquist，FTN)系统，在本发明实施例中不作限制。
为了更清楚地介绍本发明实施例中的光接收机，下面介绍图1所示的光网络系统中的光接收机，请参见图2。光接收机位于OLT和/或ONU中，在光接收机中包括模拟信号处理器和数字信号处理器：
模拟信号处理器包括如下几个部分：
90度混频器(90°Hybrid)，90度混频器与光接收机中的四个光电转换器耦合，那么，90度混频器和四个光电转换器之间存在四个信道，输出4信道信号分量。例如，同一波长的光，通过光发射机中的偏振片产生两路相互独立、正交的偏振光，将两路偏振光分别进行调制，得到两路偏振光信号，记为第一偏振光信号x和第二偏振光信号y。光发射机将第一偏振光信号x和第二偏振光信号y发送给光接收机。光接收机通过该90度混合接口接收第一偏振光信号x和第二偏振光信号y，输出与四信道信号分量相对应的四路光信号，该四信道信号分量为第一偏振光信号的同相分量xi、第一偏振光信号的正交分量xq、第二偏振光信号的同相分量yi以及第二偏振光信号的正交分量yq。
光电转换器(Optical to Electrical converter，O/E)，与90度混频器耦合，用于检测由90度混频器输出的四路光信号，并将四路光信号转化为四路电信号后输出。
模数转换器(Analog to Digital Converter，ADC)，与光电转换器耦合，用于将光电转化器输出的四路电信号转换为数字信号，此处，数字信号也是经量化的数字数据流xi、xq、yi和yq。
数字信号处理器，与模拟信号处理器中的模数转换器连接，用于对四个经量化的数字数据流xi、xq、yi和yq进行进一步处理。
现有技术中，例如，图2所示的光接收机，无法保证从90度混频器的输出到模数转换器的输入的四个信道的长度都相等，也就是无法保证从90度混频器的输出到光电转换器的输入的光纤的长度，以及从光电转换器的输出到模数转换器的输入的电缆长度都相等，从而导致光信号在传输过程中出现延时，从而对高阶调制和高符号调制速率造成影响。
鉴于此，本发明实施例提供一种光接收机，在光接收机接收由光发射机发送的偏振光信号之后，通过光接收机的90混频器、光电转换器、以及模数转换器进行处理，得到处理后的信号，其中，处理后的信号为时域信号，处理后的信号进入数字信号处理器，由数字信号处理器对处理后的信号进行快速傅里叶变换(Fast Fourier Transform，FFT)，即，将时域信号转化为频域信号，频域信号进入延时估计器，延时估计器根据频域信号，得到接收的光信号的同相分量的第一子频域信号和接收的光信号的正交分量的第二子频域信号，从而根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差，获得的相位差进入延时补偿器，延时补偿器利用获得相位差对接收的光信号进行延时补偿，即，本发明实施例提供了一种光接收机，通过光接收机中的延时估计器得到接收的光信号 的同相分量和正交分量的相位差，以使延时补偿器利用相位差对接收的光信号进行延时补偿，而无需设置复杂度较高的滤波器对正交延时进行盲补偿，从而降低了补偿复杂度。
下面结合附图介绍本发明实施例提供的技术方案，在下面的介绍过程中，以将本发明提供的技术方案应用在图1所示的应用场景中为例。
请参见图3，本发明实施例提供一种光接收机，该接收机能够应用于图1所示的应用场景中，该接收机包括：
模拟信号处理器301，用于对接收的光信号进行光电转换处理，获得电信号。
同一波长的光通过光发射机中的偏振片，产生两路相互独立、正交的偏振光，光发射机将两路偏振光分别进行调制，得到两路偏振光信号，记为第一偏振光信号x和第二偏振光信号y，光发射机将第一偏振光信号x和第二偏振光信号y发送给光接收机。因此，本发明实施例中，光接收机接收的光信号中可以包括两路偏振光信号，即，第一偏振光信号x和第二偏振光信号y，其中，第一偏振光信号x对应的信号表示式和第二偏振光信号y对应的信号表示式均为复数形式，且第一偏振光信号x和第二偏振光信号y的偏振方向垂直正交。
在下文中，将以上述两路偏振光信号中的一路偏振光信号为例介绍本发明实施例提供的一种光接收机。
在本发明实施例中，模拟信号处理器包括图2所示的90度混频器、光电转换器及模数转换器，其中，光接收机中的90度混频器接收光信号后，输出与两信道信号分量相对应的两路光信号，例如，接收的光信号以第一偏振光信号x为例，上述两信道信号分量分别为第一偏振光信号x的同相分量xi和正交分量xq，同相分量xi和正交分量xq进入光电转换器，经光电转换后输出两路电信号，进一步，经光电转换器输出的电信号经模数转换器处理，输出经量化的两路信号xi和xq。
数字信号处理器302，用于对电信号进行时频变换处理，获得频域信号。
那么数字信号处理器302接收由模拟信号处理器输出的电信号，此处的电信号就是指经量化的两路信号xi和xq，并对经量化的两路电信号进行FFT变换。
在数字信号处理器302中，在对接收到经量化的两路信号xi和xq进行FFT处理之前，先将经量化的两路信号xi和xq合并，得到复信号x＝xi+jxq，然后对复信号做FFT变换，获得频域信号FX(Nfft)，其中，复信号x＝xi+jxq为时域信号，Nfft表示FFT的频点数，例如，等于从0到4095的值；FX为时域复信号x的频域信号。
延时估计器303，用于根据频域信号，获取光信号的同相分量的第一子频域信号和光信号的正交分量的第二子频域信号，以及根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差。
在本发明实施例中，延时估计器303，具体可以为芯片或者是集成电路，接收由数字信号处理器302输出的频域信号。下面介绍延时估计器303根据频域信号获取第一子频域信号和第二子频域信号的方法。
在本发明实施例中，经数字信号处理器302进行FFT处理之前的时域信号x＝xi+jxq为复信号，经过FFT处理之后的频域信号仍为复信号，延时估计器303对频域信号进行数学变换即可得到接收的光信号的同相分量xi和接收的光信号的正交分量xq的频域表示。
在一种实施方式中，延时估计器303将频域信号的序列与该序列的对称序列的共轭相 加后除以2，得到同相分量xi的第一子频域信号，即
Figure PCTCN2017081706-appb-000006
延时估计器303将频域信号的序列与该序列的对称序列的共轭相减后除以2j，得到正交分量xq的第二子频域信号，即
Figure PCTCN2017081706-appb-000007
其中，FXI(Nfft)为时域信号x的同相分量xi的频域信号，FXQ(Nfft)为时域信号x的正交分量xq的频域信号。FXR(Nfft)＝[FX(1)，FX(end：-1:2)]，FX(end：-1:2)表示FX的序列中，坐标1的频率分量不变，对坐标2到坐标Nfft对应的频率分量进行倒置，FXR(Nfft)为FX的对称序列，其中，FX序列上的坐标1对应FX频谱的中心频点。
在本发明实施例中，在数字信号处理器302中，若直接对经量化的两路信号xi和xq做FFT变换，以获得时域信号xi的频域信号FXI和时域信号xq的频域信号FXQ，这样，在延时估计器303中则不用进行上述对频域信号进行数学变换的步骤。
在本发明实施例中，在延时估计器303获取第一子频域信号和第二子频域信号后，若要计算同相分量和正交分量的相位差，则要获取同相分量的第一相位角和正交分量的第二相位角，具体的，延时估计器303需要根据第一子频域信号，计算获得延时相关的两个信号分量，第一信号分量和第二信号分量，以及，根据第二子频域信号，计算获得延时相关的两个信号分量，第三信号分量和第四信号分量，以利用第一信号分量、第二信号分量、第三信号分量及第四信号分量获取同相分量和正交分量的相位差。
本发明实施例提供的光接收机能够应用于奈奎斯特系统，也能够应用于超奈奎斯特系统。其中，在应用于超奈奎斯特系统时，由于超奈奎斯特系统以超奈奎斯特码元速率传输，会引起码间串扰(Inter-Symbol Interference，ISI)。为了减少码间串扰对提取同相分量和正交分量的相位角的影响，在本发明实施例中，光接收机中的模拟信号处理器301对接收的光信号进行光电转换处理，获得电信号，电信号进入数字信号处理器302，由数字信号处理器302对电信号进行时频变换处理，即，对复信号x做FFT变换，得到x的频域信号FX，由延时估计器303利用频域信号FX获取接收的光信号的同相分量和正交分量的相位差，由于时频变换的频点数并不是固定的，频点数越多，也就越能更好地恢复出原始电信号，从而能更好的提取同相分量和正交分量的相位角，以提高延时估计的精确度。
在本发明实施例中，实际上延时估计器303还可以采用通过第一信号分量或第二信号分量提取同相分量的第一相位角的方式，而之所以前面介绍的是采用通过第一信号分量和第二信号分量提取同相分量的第一相位角的方式，是因为在超奈奎斯特系统中会因存在码间串扰而可能无法准确提取同相分量的第一相位角。
在延时估计器303采用通过第一信号分量或第二信号分量提取相位角的方式提取同相分量的第一相位角时，只需要利用第一子频域信号计算第一信号分量，无需同时计算第一信号分量和第二信号分量，所以，能够降低计算复杂度。
而如果本发明实施例提供的光接收机不应用在超奈奎斯特系统，或者不应用在与超奈奎斯特系统类似的系统，或者应用在能够消除码间串扰的系统，延时估计器303则可以采 用通过第一信号分量或第二信号分量提取相位角的方式，也可以采用通过第一信号分量和第二信号分量提取相位角的方式，可以根据情况选择。例如，当本发明实施例提供的光接收机应用在奈奎斯特系统中时，延时估计器303可以采用通过第一信号分量或第二信号分量提取同相分量的第一相位角的方式，或通过第一信号分量和第二信号分量提取同相分量的第一相位角的方式这两种方式中的任意一种来提取同相分量的第一相位角。
下面介绍延时估计器303获取信号分量的方法，在下面的介绍过程中，以获取第一信号分量和第二信号分量为例。相应的，对于第三信号分量和第四信号分量也可以采用下面介绍的方法来获取，不多赘述。
在一种实施方式中，延时估计器303可根据第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量获得第一信号分量和第二信号分量。
第一子频域信号在第n个频点处的频率分量和在第n+L个频点处的频率分量均表现为复数形式，包含幅度和相位。
在本发明实施例中，L至少为与第一子频域信号的频谱中心频点间隔半个符号率的距离，n为大于零且小于N的整数，N＝Nfft-L-1。
在该种实施方式中，L的第一种可能的实现方式，L为固定值，
Figure PCTCN2017081706-appb-000008
其中，Nfft为时频变换的点数，Rs为符号率，fs为采样率。若以2倍采样率为例，也就是fs＝2Rs，则可计算出，L＝0.75Nfft。或者，L为固定值L＝L1+M，
Figure PCTCN2017081706-appb-000009
M为大于零的整数。
在该种实施方式中，L的第二种可能实现方式，L＝L1+M，M可取不同的值，M的取值可以为[0，Nfft-n-L]，或者M也可以大于Nfft-n-L。当M的取值大于Nfft-n-L时，则FXI*(n+L1+M)＝FXI*(n+L1+M-Nfft)，即，对第n+L1+M个频点进行循环移位。在本发明实施例中，M究竟取多少种不同的值，即M的取值的数量，可根据对延时估计的精确度的要求进行选择，若对延时估计的精确度要求较高，则M的取值的数量较多，即可以将M取更多值。而若对延时估计的精确度要求较低，则M的取值的数量较少，本领域普通技术人员需要根据实际情况确定，在本发明实施例中不作限制。其中，L1的计算方式同L的第一种可能实现方式，不多赘述。
在本发明实施例中，延时估计器303根据第一子频域信号在第n个频点处的频率分量和在第n+L个频点处的频率分量获得第一信号分量和第二信号分量，可以有多种实现方式，以下将举例进行说明。
A、第一种实现方式：
L为固定值，即，L＝L1，或者L＝L1+M，M为整数。下面对第一种实现方式进行介绍。
在第一种实现方式中，延时估计器303将第一子频域信号在第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，就能获得第一信号分量，即
AXI＝FXI(n)·FXI*(n+L)       (1)
延时估计器303将第一子频域信号在第n+L个频点处的频率分量与在第n个频点处的频率分量的共轭相乘，就能获得第二信号分量，即
BXI＝FXI(n+L)·FXI*(n)      (2)
其中，若第一子频域信号在第n个频点处的相位角为
Figure PCTCN2017081706-appb-000010
在第n+L个频点处的相位角为
Figure PCTCN2017081706-appb-000011
复数取共轭则为实部不变，虚部取反，也就是相位相反，那么表达式(1)中AXI的共轭相乘，可理解为将第一子频域信号在第n个频点处和第n+L个频点处的幅度相乘，及将第一子频域信号在第n个频点处和第n+L个频点处的相位相减，获取同相分量的第一相位角，该相位角为
Figure PCTCN2017081706-appb-000012
其中，由将第一子频域信号在第n个频点处和第n+L个频点处的相位相减，可以理解第n个频点与第n+L个频点之间的间隔L就是与第一子频域信号的频谱的中心频点的与负频谱上间隔半个符号率位置处的频点间的距离，具体请参考图4，图4中所示的频点n为第一子频域信号的频谱的中心频点，也就是对应第一子频域信号的坐标1。
表达式(2)中BXI的共轭相乘，可理解为将第一子频域信号在第n+L个频点处和在第n个频点处的幅度相乘，及将第一子频域信号在第n+L个频点处和第n个频点处的相位相减，获取同相分量的第一相位角，该相位角为
Figure PCTCN2017081706-appb-000013
其中，由将第一子频域信号在第n+L个频点处和第n个频点处的相位相减，可以理解第n+L个频点与第n个频点之间的L就是第一子频域信号的频谱的中心频点与正频谱间隔半个符号率位置处的频点间的距离。
在第一种实现方式中，若第一信号分量为第一子频域信号在第n+L个频点处的频率分量与在第n个频点处的频率分量的共轭相乘，那么，第n+L个频点与第n个频点之间的L就是为第一子频域信号的频谱的中心频点与正频谱间隔半个符号率位置处的频点间的距离。第二子信号分量为第一子频域信号在第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，那么，第n个频点与第n+L个频点之间的L也就是第一子频域信号的频谱的中心频点的与负频谱上间隔半个符号率位置处的频点间的距离。
对于第一信号分量和第二信号分量的计算，可以根据实际需要选择，只要保证用于计算第一信号分量的两个频点之间的间隔和用于计算第二信号分量的两个频点之间的间隔为第一子频域信号在正负频谱上间隔一个符号率即可。
从上述举例可以看出，延时估计器303利用第一信号分量提取同相分量的第一相位角
Figure PCTCN2017081706-appb-000014
利用第二信号分量也能够提取同相分量的第一相位角
Figure PCTCN2017081706-appb-000015
且提取的相位角只有符号不同，从而增加了准确提取相位角的概率，以达到提高延时估计的精确度。
而对于第二子频域信号的第三信号分量和第四信号分量的获取过程，同第一信号分量和第二信号分量，在此，就不再一一赘述了。
B、第二种实现方式：
在第一种实现方式中，延时估计器303仅仅是将一对频点，例如第n个频点和第n+L个频点的频率分量进行共轭相乘。而在具体实现过程中，由于第一子频域信号在不同频点的相位角可能会有所不同，为了降低提取同相分量的第一相位角的误差，在第一种实现方式的基础上，延时估计器303将n从1遍历至N，即，将N对频点的频率分量进行共轭相乘，而获得N个第一子信号分量，将N个第一子信号分量相加，获得第一信号分量以及获得N个第二子信号分量，将N个第二子信号分量相加，获得第二信号分量。其中，对N对频点的频率分量进行共轭相乘，并将获得N个第一子信号分量相加及N个第二子信号分量相加，实际就是对N对频点的相位差进行相加后求平均，以提高根据第一信号分量和第二信号分量提取同相分量的第一相位角的精确度，进一步达到提高延时估计的精确度的目的。
在本发明实施例中，延时估计器303将n从1遍历至N，可以按照n从小到大的顺序进行遍历，也可以按照n从大到小的顺序进行遍历，也可以按照n从中间到两边的顺序进行遍历，或者按照其它方式进行遍历，在本发明实施例中不作限定。
而对于第二子频域信号的第三信号分量和第四信号分量的获取过程及能够达到的技术效果，同第一信号分量和第二信号分量，在此，就不再一一赘述了。
C、第三种实现方式：
L等于L1+M，M可取不同值。
下面以M取两种不同的值为例，例如，M＝1和M＝2。当M＝1时，延时估计器303将第一子频域信号在第n个频点处的频率分量与在第n+L1+1个频点处的频率分量的共轭相乘，获得第一信号分量，将在第n+L1个频点处的频率分量与在第n+1个频点处的频率分量的共轭相乘，获得第二信号分量；当M＝2时，延时估计器303将第一子频域信号在第n个频点处的频率分量与在第n+L1+2个频点处的频率分量的共轭相乘，获得第一信号分量，将在第n+L1个频点处的频率分量与在第n+2个频点处的频率分量的共轭相乘，获得第二信号分量，这样，共获得两个第一信号分量及两个第二信号分量。
而对于第二子频域信号的第三信号分量和第四信号分量的获取过程，同第一信号分量和第二信号分量，在此，就不再一一赘述了。
在本发明实施例中，在第二种实现方式中，延时估计器303在对n进行遍历时，将N个第一子信号分量相加，获得第一信号分量，对N个第二子信号分量相加，获得第二信号分量，这是为了降低提取第一子频域信号的不同频点的之间的相位差的误差，以提高提取同相分量的第一相位角的准确度，进而提高延时估计精确度，而在第三种实现方式中，在M取不同值时，能够获得多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，进而能够利用多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，获得同相分量和正交分量的多个不同的相位差，进而对多个不同的相位差进行处理，以降低取获同相分量和正交分量之间的相位差的误差，即从另外一个维度提高延时估计的精确度。
D、第四种实现方式：
L、M的取值同第三种实现方式。
继续以M取两种不同的值为例，例如，M＝1和M＝2。当M＝1时，延时估计器303将n从1遍历至N，将第一子频域信号在第n个频点处的频率分量与在第n+L1+1个频点处的频率分量的共轭相乘，获得N个第一子信号分量，将N个第一子信号分量相加，获得第一信号分量，将在第n+L1个频点处的频率分量与在第n+1个频点处的频率分量的共轭相乘，获得N个第二子信号分量，将N个第二子信号分量相加，获得第二信号分量。当M＝2时，延时估计器303将n从1遍历至N，将第一子频域信号在第n个频点处的频率分量与在第n+L1+2个频点处的频率分量的共轭相乘，获得第一信号分量，将在第n+L1个频点处的频率分量与在第n+2个频点处的频率分量的共轭相乘，获得N个第二子信号分量，将N个第二子信号分量相加，获得第二信号分量，这样，共获得两个第一信号分量及两个第二信号分量。
而对于第二子频域信号的第三信号分量和第四信号分量的获取过程，同第一信号分量和第二信号分量，在此，就不再一一赘述了。
在本发明实施例中，延时估计器303在对n进行遍历时，对N个第一子信号分量相加， 获得第一信号分量，对N个第二子信号分量相加，获得第二信号分量；对N个第三子信号分量相加，获得第三信号分量，对N个第四子信号分量相加，获得第四信号分量，这降低了提取第一子频域信号不同频点的之间的相位差及提取第二子频域信号的不同频点之间的相位差的误差，从而提高了提取同相分量的第一相位角及正交分量的第二相位角的准确度，进而提高延时估计精确度，进一步，M取不同值时，能够获得多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，进而能够利用多个不同的第一信号分量、多个不同的第二信号分量、多个不同的第三信号分量及多个不同的第四信号分量，获得同相分量和正交分量的多个相位差，进而对多个不同的相位差进行处理，以降低取获同相分量和正交分量之间的相位差的误差，从另外一个维度进一步提高延时估计的精确度。
在本发明实施例中，在延时估计器303获取第一信号分量和第二信号分量之后，根据第一信号分量和第二信号分量便可得到同相分量的相位角，本文中将同相分量的相位角称为第一相位角，将其表示为
Figure PCTCN2017081706-appb-000016
例如，若第一信号分量为
Figure PCTCN2017081706-appb-000017
第二信号分量为
Figure PCTCN2017081706-appb-000018
则第一相位角为
Figure PCTCN2017081706-appb-000019
而在M取值不同时，延时估计器303获取第一子频域信号的多个第一信号分量及多个第二信号分量，进而利用多个第一信号分量及多个第二信号分量中对应相同M的第一信号分量和第二信号分量，获得同相分量的多个第一相位角。其中，对应相同M的第一信号分量和第二信号分量，指的是M取第一值时，利用第一子频域信号在第n个频点处的频率分量和在第n+L1+M个频点处的频率分量获得信号分量。例如，继续以M取两种不同的值为例，M＝1和M＝2。当M＝1时，延时估计器303获得第一子频域信号的第一信号分量
Figure PCTCN2017081706-appb-000020
以及获得第一子频域信号的第二信号分量
Figure PCTCN2017081706-appb-000021
当M＝2时，延时估计器303获得第一子频域信号的第一信号分量
Figure PCTCN2017081706-appb-000022
以及获得第一子频域信号的第二信号分量
Figure PCTCN2017081706-appb-000023
那么，
Figure PCTCN2017081706-appb-000024
和
Figure PCTCN2017081706-appb-000025
就是对应相同M的第一信号分量和第二信号分量，
Figure PCTCN2017081706-appb-000026
和
Figure PCTCN2017081706-appb-000027
就是对应相同M的第一信号分量和第二信号分量。延时估计器303从而利用第一信号分量
Figure PCTCN2017081706-appb-000028
和第二信号分量
Figure PCTCN2017081706-appb-000029
获得同相分量的第一相位角
Figure PCTCN2017081706-appb-000030
利用第一信号分量
Figure PCTCN2017081706-appb-000031
和第二信号分量
Figure PCTCN2017081706-appb-000032
获得同相分量的第一相位角
Figure PCTCN2017081706-appb-000033
在本发明实施例中，在延时估计器303获取第三信号分量和第四信号分量之后，根据第三信号分量和第四信号分量便可得到正交分量的相位角，本文中将正交分量的相位角称为第二相位角，将其表示为
Figure PCTCN2017081706-appb-000034
在M取值不同时，延时估计器303获取第二子频域信号的多个第三信号分量和多个第四信号分量，进而利用多个第三信号分量及多个第四信号分量中对应相同M的第三信号分量和第四信号分量获得正交分量的多个第二相位角。例如，继续以M取两种不同的值为例，M＝1和M＝2。当M＝1时，延时估计器303获得正交分量的第二相位角
Figure PCTCN2017081706-appb-000035
在M＝2时，延时估计器303获得正交分量的第二相位角
Figure PCTCN2017081706-appb-000036
在本发明实施例中，在延时估计器303获取多个第一信号分量、多个第二信号分量、多个第三信号分量及多个第四信号分量后，相应的，也就获得多个第一相位角和多个第二相位角，根据多个第一相位角和多个第二相位角中的对应相同M的第一相位角和第二相位角，获得多个相位差，继续以M取两种不同的值为例，例如，M＝1和M＝2。当M＝1时，延时估计器303获得同相分量和正交分量的相位差为
Figure PCTCN2017081706-appb-000037
当M＝2时，延时估计器303 获得同相分量和正交分量的相位差为
Figure PCTCN2017081706-appb-000038
然后计算两个相位差的平均相位差，将平均相位差作为正交分量和正交分量的相位差，也就是对获得的多个相位差通过滤波器进行滤波处理。对多个相位差进行滤波，以降低获取同相分量和正交分量之间的相位差的误差，进而提高延时估计的精确度。
在本发明实施例中，在延时估计器303获取同相分量和正交分量的相位差
Figure PCTCN2017081706-appb-000039
之后，还可以根据
Figure PCTCN2017081706-appb-000040
计算出同相分量和正交分量之间的延时T，其中，f为接收的光信号的频率。在计算出延时T之后，便后续通过延时T对接收的光信号进行补偿。
延时补偿器304，用于利用相位差对光信号进行延时补偿。
在本发明实施例中，可以在频域利用得到的延时T对频域信号FXI(N)和FXQ(N)进行相位调整，以达到对接收的光信号进行延时补偿的目的，或者，也可以在时域对接收的光信号进行延时补偿，下面则分别进行介绍。
在本发明实施例中，第一种实现方式，延时补偿器304可以为移相器，移相器通过同相分量和正交分量的相位差对第一子频域信号FXI或第二子频域信号FXQ的相位进行调整，以达到对接收的光信号进行延时补偿的目的。
在本发明实施例中，第二种实现方式，延时补偿器304也可以为时钟，时钟利用延时估计器计算出的时延T，对模拟信号处理器301中的模数转换器的时延进行调整，以达到对接收的光信号进行延时补偿的目的。或者，延时补偿器304利用延时T，通过延时电路对经量化的信号xi或xq进行延时调整，以达到对接收的光信号进行延时补偿的目的，例如，以经量化的信号xi为例，输入到延时电路的输入信号为xi(t-T)，经过延时电路的调整后，从延时电路输出xi(t)，输出的信号xi(t)就是经过正交延时补偿后的信号。
请参见图5，本发明实施例提供的一种延时估计方法，该方法能够通过延时估计器实现，延时估计器作为光接收机的一部分，该方法的流程描述如下：
步骤501：延时估计器接收光接收机对接收的光信号进行时频变换处理后的频域信号；
步骤502：延时估计器根据频域信号，获取光信号的同相分量的第一子频域信号和光信号的正交分量的第二子频域信号；
步骤503：延时估计器根据所述第一子频域信号和第二子频域信号，计算所述同相分量和所述正交分量的相位差，同相分量和正交分量的相位差用于光接收机对光信号进行延时估计。
在本发明实施例中，步骤501-步骤503所提供的方法已在对图3所示实施例中提供的延时估计器的描述中介绍，这里不再赘述。
请参见图6，本发明实施例提供一种延时估计器，该延时估计器包括连接到同一总线600的网络接口601、处理器602。
其中，处理器602可以是中央处理器(CPU)，或特定应用集成电路(Application Specific Integrated Circuit，ASIC)，可以是一个或多个用于控制程序执行的集成电路，可以是基带芯片，等等。
网络接口601可以通过总线600与处理器602相连(例如图6所示)，或者通过专门的连接线分别与处理器连接。
延时估计器还可以包括存储器，存储器可以通过总线600与处理器602连接。存储器的数量可以是一个或多个，存储器可以是只读存储器(Read only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)或磁盘存储器，等等。
通过对处理器602进行设计编程，将前述的延时估计方法所对应的代码固化到芯片内，从而使芯片在运行时能够执行前述图5所示的实施例提供的延时估计方法，如何对处理器602进行设计编程为本领域技术人员公知的技术，这里不再赘述。
请参见图7，本发明实施例提供一种延时估计器，该延时估计器包括接收模块701、获取模块702及计算模块703。
在实际应用中，获取模块702及计算模块703对应的实体装置可以集成在图6中的处理器602中，接收模块701对应的实体装置可以集成在图6中的网络接口601中。
本发明实施例提供了一种光接收机，在该光接收机中，由延时估计器获得频域信号，并根据频域信号获得光接收机接收的光信号的同相分量的第一子频域信号和接收的光信号的正交分量的第二子频域信号，进而根据第一子频域信号和第二子频域信号，计算同相分量和正交分量的相位差，使得延时补偿器利用该相位差对接收的光信号进行延时补偿，可以看到，这种光接收机中是先估计后补偿，因此无需设置复杂度较高的滤波器对正交延时进行盲补偿，从而降低了补偿复杂度。
以上实施例仅用以对本发明实施例的技术方案进行详细介绍，但以上实施例的说明只是用于帮助理解本发明实施例的方法及其核心思想，不应该理解为对本申请的限制。本领域技术人员在本发明实施例揭露的技术范围内，可轻易想到的变化或替换，都应涵盖在本发明实施例的保护范围之内。

Claims (19)

  1. 一种光接收机，其特征在于，包括：
    模拟信号处理器，用于对接收的光信号进行光电转换处理，获得电信号；
    数字信号处理器，用于对所述电信号进行时频变换处理，获得频域信号；
    延时估计器，用于根据所述频域信号获取所述光信号的同相分量的第一子频域信号和所述光信号的正交分量的第二子频域信号，以及根据所述第一子频域信号和所述第二子频域信号，计算所述同相分量和所述正交分量的相位差；
    延时补偿器，用于利用所述相位差对所述光信号进行延时补偿。
  2. 如权利要求1所述的光接收机，其特征在于，所述延时估计器在根据所述第一子频域信号和所述第二子频域信号，计算所述同相分量和所述正交分量的相位差时，具体用于：
    根据所述第一子频域信号，获得第一信号分量和第二信号分量；其中，所述第一信号分量和所述第二信号分量用于提取所述同相分量的第一相位角；
    根据所述第二子频域信号，获得第三信号分量和第四信号分量；其中，所述第三信号分量和所述第四信号分量用于提取所述正交分量的第二相位角；
    根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差。
  3. 如权利要求2所述的光接收机，其特征在于，所述延时估计器在根据所述第一子频域信号，获得第一信号分量和第二信号分量时，具体用于：
    根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量；
    所述延时估计器在根据所述第二子频域信号，获得第三信号分量和第四信号分量时，具体用于：
    根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量；
    其中，L至少为与所述第一子频域信号的频谱的中心频点间隔半个符号率的距离或所述第二子频域信号的频谱的中心频点间隔半个符号率的距离，n为大于零且小于N的整数，N等于所述时频变换的频点数减去L加1。
  4. 如权利要求3所述的光接收机，其特征在于，
    Figure PCTCN2017081706-appb-100001
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：
    将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量；
    将所述第一子频域信号在所述第n+L个频点处的频率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量。
  5. 如权利要求4所述的光接收机，其特征在于，所述延时估计器将所述第一子频域 信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量时，具体用于：
    将n从1遍历至N，获得N个第一子信号分量；
    将所述N个第一子信号分量相加，获得所述第一信号分量；
    所述延时估计器将所述第一子频域信号在所述第n+L个频点处的频率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量时，具体用于：
    将n从1遍历至N，获得N个第二子信号分量；
    将所述N个第二子信号分量相加，获得所述第二信号分量。
  6. 如权利要求3所述的光接收机，其特征在于，L等于L1+M，
    Figure PCTCN2017081706-appb-100002
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：
    对于M的每个取值，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第一信号分量，共获得多个第一信号分量；
    对于M的每个取值，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第二信号分量，共获得多个第二信号分量；
    所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量时，具体用于：
    对于M的每个取值，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第三信号分量，共获得多个第三信号分量；
    对于M的每个取值，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第四信号分量，共获得多个第四信号分量。
  7. 如权利要求3所述的光接收机，其特征在于，L等于L1+M，
    Figure PCTCN2017081706-appb-100003
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量时，具体用于：
    对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N个第一子信号分量；将所述N个第一子信号分量相加，获得所述第一信号分量；所述第一取值为M的任意一个取值；
    对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第二子信号分量；将所述N个第二子信号分量相加，获得所述第二信号分量；
    对于M的多个取值，获得多个第一信号分量及多个第二信号分量；
    所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量时，具体用于：
    对于M的第一取值，n从1遍历至N，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N第三子信号分量，及，将所述N个第三子信号分量相加，获得所述第三信号分量；
    对于M的第一取值，n从1遍历至N，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第四子信号分量，及将所述N个第四子信号分量相加，获得所述第四信号分量；
    对于M的多个取值，获得多个第三信号分量及多个第四信号分量。
  8. 如权利要求6或7所述的光接收机，其特征在于，所述延时估计器根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差时，具体用于：
    根据所述多个第一信号分量和所述多个第二信号分量中对应相同M的第一信号分量和第二信号分量，获得所述同相分量的多个第一相位角，及，根据所述多个第三信号分量和所述多个第四信号分量中对应相同M的第三信号分量和第四信号分量，获得所述正交分量的多个第二相位角；
    将所述多个第一相位角与所述多个第二相位角中对应相同M的第一相位角和第二相位角相减，获得多个相位差；
    计算所述多个相位差的平均相位差，确定所述平均相位差为所述同相分量和所述正交分量的相位差。
  9. 如权利要求1所述的光接收机，其特征在于，所述延时补偿器利用所述相位差对所述光信号进行延时补偿时，具体用于：
    利用所述相位差对所述第一子频域信号或所述第二子频域信号的相位进行调整，以完成对所述光信号的延时补偿。
  10. 如权利要求1所述的光接收机，其特征在于，所述延时补偿器利用所述相位差对所述光信号进行延时补偿时，具体用于：
    利用由所述延时估计器根据所述相位差获得的延时，对所述模拟信号处理器中的模数转换器的时延进行调整，以完成对所述光信号的延时补偿。
  11. 一种延时估计方法，其特征在于，包括：
    延时估计器接收光接收机对接收的光信号进行时频变换处理后的频域信号；
    所述延时估计器根据所述频域信号，获取所述光信号的同相分量的第一子频域信号和所述光信号的正交分量的第二子频域信号；
    所述延时估计器根据所述第一子频域信号和所述第二子频域信号，计算所述同相分量和所述正交分量的相位差，所述同相分量和所述正交分量的相位差用于所述光接收机对所述光信号进行延时估计。
  12. 如权利要求11所述的方法，其特征在于，所述延时估计器根据所述第一子频域信号和所述第二子频域信号，计算所述同相分量和所述正交分量的相位差，包括：
    所述延时估计器根据所述第一子频域信号，获得第一信号分量和第二信号分量；其中，所述第一信号分量和所述第二信号分量用于提取所述同相分量的第一相位角；
    所述延时估计器根据所述第二子频域信号，获得第三信号分量和第四信号分量；其中，所述第三信号分量和所述第四信号分量用于提取所述正交分量的第二相位角；
    所述延时估计器根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差。
  13. 如权利要求12所述的方法，其特征在于，所述延时估计器根据所述第一子频域信号，获得第一信号分量和第二信号分量，包括：
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量；
    所述延时估计器根据所述第二子频域信号，获得第三信号分量和第四信号分量，包括：
    所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量；
    其中，其中，L至少为与所述第一子频域信号的频谱的中心频点间隔半个符号率的距离或所述第二子频域信号的频谱的中心频点间隔半个符号率的距离，n为大于零且小于N的整数，N等于所述时频变换的频点数减去L加1。
  14. 如权利要求13所述的方法，其特征在于，
    Figure PCTCN2017081706-appb-100004
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量，包括：
    所述延时估计器将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量；
    所述延时估计器将所述第一子频域信号在所述第n+L个频点处的频率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量。
  15. 如权利要求14所述的方法，其特征在于，所述延时估计器将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L个频点处的频率分量的共轭相乘，获得所述第一信号分量，包括：
    所述延时估计器将n从1遍历至N，获得N个第一子信号分量；
    所述延时估计器将所述N个第一子信号分量相加，获得所述第一信号分量；
    所述延时估计器将所述第一子频域信号在所述第n+L个频点处的频率分量与在所述第n个频点处的频率分量的共轭相乘，获得所述第二信号分量，包括：
    所述延时估计器将n从1遍历至N，获得N个第二子信号分量；
    所述延时估计器将所述N个第二子信号分量相加，获得所述第二信号分量。
  16. 如权利要求13所述的方法，其特征在于，L等于L1+M，
    Figure PCTCN2017081706-appb-100005
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量，包括：
    所述延时估计器对于M的每个取值，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第一信号分量，共获得多个第一信号分量；
    所述延时估计器对于M的每个取值，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第二信号分量，共获得多个第二信号分量；
    所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量，包括：
    所述延时估计器对于M的每个取值，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得所述第三信号分量，共获得多个第三信号分量；
    所述延时估计器对于M的每个取值，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得所述第四信号分量，共获得多个第四信号分量。
  17. 如权利要求13所述的方法，其特征在于，L等于L1+M，
    Figure PCTCN2017081706-appb-100006
    其中，Nfft为所述时频变换的点数、Rs为符号率、fs为采样率，M为大于零的整数；
    所述延时估计器根据所述第一子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第一信号分量和所述第二信号分量，包括：
    所述延时估计器对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N个第一子信号分量；将所述N个第一子信号分量相加，获得所述第一信号分量；所述第一取值为M的任意一个取值；
    所述延时估计器对于M的第一取值，将n从1遍历至N，将所述第一子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第二子信号分量；将所述N个第二子信号分量相加，获得所述第二信号分量；
    所述延时估计器对于M的多个取值，获得多个第一信号分量及多个第二信号分量；
    所述延时估计器根据所述第二子频域信号在第n个频点处的频率分量及在第n+L个频点处的频率分量，获得所述第三信号分量和所述第四信号分量，包括：
    所述延时估计器对于M的第一取值，n从1遍历至N，将所述第二子频域信号在所述第n个频点处的频率分量与在第n+L1+M个频点处的频率分量的共轭相乘，获得N个第三子信号分量，及，将所述N个第三子信号分量相加，获得所述第三信号分量；
    所述延时估计器对于M的第一取值，n从1遍历至N，将所述第二子频域信号在第n+L1个频点处的频率分量与在第n+M个频点处的频率分量的共轭相乘，获得N个第四子信号分量，及将所述N个第四子信号分量相加，获得所述第四信号分量；
    对于M的多个取值，获得多个第三信号分量及多个第四信号分量。
  18. 如权利要求16或17所述的方法，其特征在于，所述延时估计器根据所述第一信号分量、所述第二信号分量、所述第三信号分量及所述第四信号分量，计算所述同相分量和所述正交分量的相位差，包括：
    所述延时估计器根据所述多个第一信号分量和所述多个第二信号分量中对应相同M的第一信号分量和第二信号分量，获得所述同相分量的多个第一相位角，及，根据所述多个第三信号分量和所述多个第四信号分量中对应相同M的第三信号分量和第四信号分量，获得所述正交分量的多个第二相位角；
    所述延时估计器根据所述多个第一相位角及所述多个第二相位角，计算所述同相分量和所述正交分量的相位差；
    所述延时估计器将所述多个第一相位角与所述多个第二相位角中对应相同M的第一相位角和第二相位角相减，获得多个相位差；
    所述延时估计器计算所述多个相位差的平均相位差，确定所述平均相位差为所述同相分量和所述正交分量的相位差。
  19. 如权利要求18所述的方法，其特征在于，所述方法还包括：
    所述延时估计器根据所述相位差，获得所述同相分量和所述正交分量之间的延时，所述延时用于所述光接收机对所述光信号进行延时补偿。
PCT/CN2017/081706 2017-04-24 2017-04-24 一种光接收机及延时估计方法 WO2018195722A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201780082891.1A CN110168967B (zh) 2017-04-24 2017-04-24 一种光接收机及延时估计方法
PCT/CN2017/081706 WO2018195722A1 (zh) 2017-04-24 2017-04-24 一种光接收机及延时估计方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2017/081706 WO2018195722A1 (zh) 2017-04-24 2017-04-24 一种光接收机及延时估计方法

Publications (1)

Publication Number Publication Date
WO2018195722A1 true WO2018195722A1 (zh) 2018-11-01

Family

ID=63917881

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/081706 WO2018195722A1 (zh) 2017-04-24 2017-04-24 一种光接收机及延时估计方法

Country Status (2)

Country Link
CN (1) CN110168967B (zh)
WO (1) WO2018195722A1 (zh)

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1959590A2 (en) * 2007-02-16 2008-08-20 Fujitsu Ltd. Analog-to-digital conversion controller, optical receiving device, optical receiving method, and waveform-distortion compensating device
CN101610114A (zh) * 2008-06-19 2009-12-23 富士通株式会社 光接收装置和数字接收电路

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8503545B2 (en) * 2006-08-31 2013-08-06 Advanced Micro Devices, Inc. I/Q imbalance compensation
CN101154958A (zh) * 2006-09-29 2008-04-02 安国国际科技股份有限公司 调整同相/正交讯号的方法
CN201690437U (zh) * 2010-05-13 2010-12-29 高宪民 一种精密鉴相器
KR101127467B1 (ko) * 2010-10-13 2012-07-12 지씨티 세미컨덕터 인코포레이티드 Ｉｑ 부정합을 추정 및 보상하는 수신기
RU2557012C2 (ru) * 2011-02-17 2015-07-20 Хуавэй Текнолоджиз Ко., Лтд. Модуль оценивания расфазировки, модуль компенсации расфазировки и когерентный приемник
US10050744B2 (en) * 2012-03-16 2018-08-14 Analog Devices, Inc. Real-time I/Q imbalance correction for wide-band RF receiver
KR20140077734A (ko) * 2012-12-14 2014-06-24 한국전자통신연구원 광 수신기에 대한 iq 불균형 측정 장치 및 iq 불균형 측정 방법
CN104980377B (zh) * 2014-04-03 2018-11-16 华为技术有限公司 一种iq不平衡的估计和校正的设备、系统及方法

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1959590A2 (en) * 2007-02-16 2008-08-20 Fujitsu Ltd. Analog-to-digital conversion controller, optical receiving device, optical receiving method, and waveform-distortion compensating device
CN101610114A (zh) * 2008-06-19 2009-12-23 富士通株式会社 光接收装置和数字接收电路

Also Published As

Publication number Publication date
CN110168967A (zh) 2019-08-23
CN110168967B (zh) 2020-12-22

Similar Documents

Publication Publication Date Title
RU2557012C2 (ru) Модуль оценивания расфазировки, модуль компенсации расфазировки и когерентный приемник
JP6405833B2 (ja) 信号処理装置及び信号処理方法
CN101610114B (zh) 光接收装置和数字接收电路
US8374512B2 (en) Frequency offset monitoring device and optical coherent receiver
US8781333B2 (en) Clock recovery apparatus
CN102231648B (zh) 基于单载波和多载波的混合传输系统
EP2583424B1 (en) Method for phase and oscillator frequency estimation
JP6794885B2 (ja) 周波数オフセット推定装置、チャネル間隔推定装置、方法及びシステム
JP2013048453A (ja) 周波数オフセット補償装置及び方法並びに光コヒーレント受信機
US20150372764A1 (en) Optical receiver having an equalization filter with an integrated signal re-sampler
CN104780037A (zh) 一种时钟恢复方法、装置及系统
EP3048746A1 (en) Method and device for estimation of chromatic dispersion in optical coherent communication
CN105612700B (zh) 用于表征光接收信号的色度色散的装置
CN110266380B (zh) 一种采用单探测器的光场重建与时频同步系统与方法
JP6310091B2 (ja) 色度分散測定方法、装置およびデジタルコヒーレント受信機
US11777723B2 (en) Post-reception synchronization in a continuous variable quantum key distribution (CV-QKD) system
CN106788734B (zh) 一种采用无数据辅助频偏估计算法的光ofdm系统
KR20160060677A (ko) 신호를 송신/수신하기 위한 방법, 이에 대응하는 디바이스와 시스템
US10505641B2 (en) Clock recovery for band-limited optical channels
WO2014060031A1 (en) Method and apparatus for estimating channel coefficients of a mimo communications channel
WO2018195722A1 (zh) 一种光接收机及延时估计方法
WO2020108632A1 (zh) 色散估计方法、装置、接收机及存储介质
Beppu et al. Mode-multiplexed 16QAM transmission over 60-km coupled four-core fibres using real-time MIMO-DSP with high-accuracy frequency offset estimation
JP7482128B2 (ja) 分散推定方法、装置、受信機及び記憶媒体
EP2976852B1 (en) Timing recovery apparatus

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: 17907546

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: 17907546

Country of ref document: EP

Kind code of ref document: A1