Local oscillator frequency offset compensation in a coherent optical signal receiver

Cross-reference of related applications

This application is a continuation of US Patent Application No. 12 / 718,177 (Filing Date: March 5, 2010), which is incorporated herein by reference in its entirety. Priority of continuation of application No. 61 / 159,011 (Filing date: March 10, 2009) and US patent application No. 12 / 718,124 (Filing date: March 5, 2010), and provisional US patent It claims the priority of Application No. 61 / 159,018 (Filing Date: March 10, 2009).

The present application relates to optical information communication, and specifically, to correction of a local oscillator frequency offset of a coherent optical signal receiver.

Data can be sent to remote locations using signals. For example, in an optical communication system, by modulating data to one or more optical wavelengths, it is possible to generate an optical modulation signal that can be transmitted through an optical waveguide such as an optical fiber. A modulation scheme that can be used in optical communication systems in which data is transmitted by modulating the phase of the optical wavelength so that the phase or phase transition of the optical wavelength represents a symbol that encodes one or more bits. There is transmodulation. For example, in a two-phase phase shift keying (BPSK) scheme, two phases can be used to represent one bit per symbol. In the orthogonal phase shift keying (QPSK) method, it is possible to code 2 bits per symbol using 4 phases. Other phase shift keying methods include variants of phase shift keying and differential phase shift keying, such as differential phase shift keying (DPSK) and zero return DPSK (RZ-DPSK). ..

In order to receive the data, it is necessary to detect and demodulate the signal. For example, in a phase-modulated optical communication system, an optical modulation signal can be detected by coherent detection by a coherent optical receiver, which is advantageous in terms of sensitivity as compared with receivers of other detection methods. In such a system, demodulated data can be obtained by processing the received signal by digital signal processing (DSP). Digital signal processing of the received signal improves processing speed and flexibility and can perform a variety of functions, including estimating the carrier phase of the received signal and detecting data based on the estimated carrier phase.

<p num="0005"> A local oscillator signal mixed with the received signal is used for coherent detection. Unfortunately, the local oscillator output of a photodetector system can range from MHz to GHz and drift over time due to a variety of factors. In order to demodulate the optical signal accurately, it is desirable to correct the frequency offset between the received signal and the local oscillator signal. In order for the frequency offset correction method to be reliable, it must be able to cover a wide range of frequency offsets and handle expected frequency drifts.</p>

<p num="0006"> According to one aspect of the present disclosure, there is provided a system for correcting a frequency offset between a local oscillator signal and a data signal in a coherent receiver. This system obtains an estimated value of the low-speed drift frequency offset by executing the feed-forward frequency offset correction function, corrects the low-speed drift frequency offset based on the obtained estimated value, and executes the judgment feedback frequency offset correction function. A digital signal processing device (DSP) configured to obtain an estimated value of the high-speed drift frequency offset in response to feedback from the bit determination function and correct the high-speed drift frequency offset based on the obtained estimated value. have.</p><p num="0007"> According to another aspect of the present disclosure, an optical communication system is provided. In this optical communication system, at least one is a phase shift keying (PSK) signal having data modulated according to a phase shift keying format, and a plurality of optical signals having different wavelengths corresponding to each are transmitted to an optical information path. Has a transmitting terminal. This optical communication system further has a receiving terminal connected to an optical information path and receiving at least one of a plurality of optical signals. The receiving terminal has an optical signal receiver that receives a PSK signal, mixes the received PSK signal with a local oscillator signal, and generates at least one electrical signal representing the PSK signal. In addition, this receiving terminal receives an electric signal and executes a feed / forward frequency offset correction function to obtain an estimated value of the low-speed drift frequency offset between the received signal and the local oscillator signal, and obtains the estimated value. Based on this, the low-speed drift frequency offset is corrected and the judgment feedback frequency offset correction function is executed to obtain an estimated value of the high-speed drift frequency offset between the received signal and the local oscillator signal in response to the feedback from the bit judgment function. It has a digital signal processor (DSP) configured to correct the fast drift frequency offset based on the calculated and estimated values.</p><p num="0008"> According to yet another aspect of the present disclosure, there is provided a method of demodulating a received optical signal having data modulated according to a phase shift keying format. This method involves mixing the local oscillator signal with the optical signal, converting the optical signal into at least one electrical signal that represents the optical signal, and performing a feed-forward frequency offset correction function to the received signal. By obtaining an estimated value of the low-speed drift frequency offset between the local oscillator signal and correcting the low-speed drift frequency offset of the electric signal based on the obtained estimated value, and executing the determination feedback carrier phase estimation function. By executing the step of obtaining the estimated value of the operating carrier phase of the optical signal according to the feedback from the electric signal and the bit determination function and the determination feedback frequency offset correction function, the received signal is received according to the feedback from the bit determination function. By obtaining an estimated value of the high-speed drift frequency offset between the local oscillation signal and correcting the high-speed drift frequency offset in the estimated value of the operating carrier phase based on the obtained estimated value, and executing the bit determination function. It has a step of determining a data value from an electric signal according to an estimated value of the operating carrier phase and generating an output representing the data modulated into an optical signal.</p>

The following detailed description should be read in conjunction with the following drawings, where similar elements are indicated by similar reference numerals.<figref num="1">A block diagram of an exemplary embodiment of a system conforming to the present disclosure.</figref><figref num="2">A block diagram of an exemplary embodiment of a receiver conforming to the present disclosure.</figref><figref num="3">A block diagram of an exemplary embodiment of a receiver incorporating a dual stage frequency offset correction feature conforming to the present disclosure.</figref><figref num="4">A block diagram of an exemplary embodiment of an optical communication system incorporating a dual stage frequency offset correction function conforming to the present disclosure.</figref><figref num="5">The flowchart which shows the 1st stage of the exemplary dual stage frequency offset correction method which fits this disclosure.</figref><figref num="6">The figure which plotted the cumulative phase change between consecutive symbols with respect to a sample number (bit index) in the 1st stage of the dual stage frequency offset correction method conforming to this disclosure.</figref><figref num="7">The figure which plotted the cumulative phase change amount between the continuous symbol and the sample number (bit index) in the output of the 1st stage of the dual stage frequency offset correction method conforming to this disclosure.</figref><figref num="8">A flowchart illustrating a second stage of an exemplary dual stage frequency offset correction method conforming to the present disclosure.</figref>

Roughly speaking, a system conforming to this disclosure performs dual stage local oscillator (LO) frequency offset correction (FOC). In the first stage, the feedforward FOC function corrects the frequency offset that drifts relatively slowly. In the second stage, the judgment feedback FOC function corrects the frequency offset that drifts at a relatively high speed. The feed-forward frequency offset correction is performed by the feed-forward carrier phase estimation function, and the judgment feedback frequency offset correction is performed by the judgment feedback carrier phase estimation function. FIG. 1 is a simplified block diagram of an exemplary embodiment of a WDM transmission system 100 conforming to the present disclosure. This transmission system transmits a plurality of optical channels from the transmitting terminal 104 to one or more remote receiving terminals 106 via the optical information path 102. This exemplary system 100 may be a long-range submarine system configured to transmit channels from transmitters separated by 5,000 km or more to receivers. In an exemplary embodiment, an optical system is taken as an example to provide a valid description for a long-range WDM optical system, but the broad concept described herein is another communication system that transmits and receives different types of signals. It is also possible in.

For ease of explanation, it will be readily apparent to those skilled in the art that System 100 is a significant simplification of the two-point system. For example, the transmitting terminal 104 and the receiving terminal 106 may of course be configured as transceivers each performing a transmitting and receiving function. However, for the sake of brevity, the present specification illustrates and describes each terminal as performing only a transmitting or receiving function. Systems and methods conforming to this disclosure can be incorporated into various network components and network configurations. The illustrated exemplary embodiments are for illustration purposes only and are not meant to be limiting.

In the illustrated exemplary embodiment, each of the plurality of transmitters TX1, TX2, --------, TXN has the corresponding input ports 108-1, 108-2, --------, 108N, respectively. Receives a data signal and sets the received data signal to the corresponding wavelength λ₁, Λ₂, -----Λ_NSend by. For example, with a PSK modulation format such as DBPSK, DQPSK, RZ-DPSK, RZ-DQPSK, one or more of the transmitters TX1, TX2, --------, TXN to modulate the data to the corresponding wavelength. Can be configured. Of course, the transmitter is greatly simplified for the sake of explanation. It is well known to those skilled in the art that each transmitter can have electrical and optical components configured to have the desired amplitude and transmit data signals at the corresponding wavelengths according to the desired modulation scheme.

The transmission wavelength, that is, the transmission channel, is transmitted by each of the plurality of paths 110-1, 110-2, -----, 110-N. The N data channels are coupled to the overall signal in the optical information path 102 by a multiplexer or coupler 112. The optical information path 102 can include an optical fiber waveguide, an optical amplifier, an optical filter, a dispersion correction module, and other active and passive components. The overall signal is received by one or more remote receiving terminals 106. Wavelength λ₁, Λ₂, -----Λ_NThe transmission channel of is separated by the demultiplexer 114 into the corresponding receivers RX1, RX2, -----, and the corresponding paths 116-1, 116-2, -----, 116-N connected to the RXN. To. Receivers RX1, RX2, -----, to demodulate the transmit signal and output the output data signal to the corresponding output paths 118-1, 118-2, 118-3, -----, 118-N, One or more of the RXNs can be configured. As used herein, the term "connection" means any connection, connection, etc. in which a signal carried by one system element is transmitted to the "connected" element. Such "connected" devices or signals and devices do not necessarily have to be directly connected to each other and may be separated by intermediate components or devices that manipulate or modify such signals. FIG. 2 is a simplified block diagram of an exemplary receiver 200 conforming to the present disclosure. In the illustrated exemplary embodiment 200, the coherent receiver configuration 202 that receives the input signal of the path 116-N and the digital signal processing that processes the output of the coherent receiver and outputs the output data signal to the path 118-N. It has a (DSP) circuit 204. Data is the carrier wavelength λ of the optical input signal according to the PSK modulation format_N Is modulated into. The coherent receiver 202 converts the received optical input signal into one or more digital signals and supplies them as inputs to the DSP circuit 204. DSP circuit 204 has a carrier wavelength λ_NThe data representing the modulated data is demodulated from the digital signal and output as an output data stream to path 118-N.

The coherent receiver 202 can be configured in various ways. The receiver of the illustrated exemplary embodiment is a deflecting beam splitter (PBS) 206, first and second 90 degree optical hybrids 208, 210, local oscillator (LO) 212, equilibrium detectors 214, 216, 218. , 220, and analog-to-digital (A / D) converters 222, 224, 226, 228. The operation of these components in a coherent optical signal receiver will be briefly described below. Normally, input optical signals with different deflections are separated into different paths by PBS206. Each deflected light is connected to the corresponding 90 degree optical hybrids 208, 210. Each 90 degree optical hybrid mixes the corresponding input signal with four orthogonal LO signals in complex space. Each optical hybrid then supplies four mixed signals to two pairs of equilibrium detectors 214, 216, 218, 220. The output of the equilibrium detector is converted into a digital signal by the A / D converters 222, 224, 226 and 228.

The output of the A / D converter is supplied to the DSP circuit 204 as an input. DSPs typically involve signal processing by a special purpose processor that executes one or more application-specific integrated circuits (ASICS) and / or specific instruction sequences, eg, directly and / or under the control of software instructions. In the illustrated exemplary embodiment, the DSP circuit 204 has preprocessing function 230, local oscillator (LO) frequency offset correction function 232, carrier phase estimation (CPE) function 234, bit determination function 236, and any training sequence. -Has a bit error rate test function 238. These functions can be performed in various configurations using any combination of hardware, software, and / or firmware. Although each function is shown to be separate, one integrated circuit or processor, or a combination of integrated circuits and / or processors, can perform any one or more functions. It is also possible to have integrated circuits and / or processors that perform DSP functions share all or part of the illustrated functions. The DSP pre-processing function 230 can have various optical signal detection functions performed by different types of DSP-based coherent detection receivers. The pre-processing function 230 can have, for example, waveform recovery and alignment functions, deterministic distortion correction functions, clock recovery functions, synchronous data resampling functions, polarization tracking and polarization mode dispersion (PMD) correction functions. Since the data of a PSK-modulated signal is usually encoded in the phase of the optical carrier signal, demodulation of the PSK-modulated signal in a DSP-based receiver involves estimating and tracking the carrier phase. The carrier phase estimation function 234 can be configured to estimate and track the optical carrier signal phase required for demodulation, and can be a dual-stage carrier phase estimation function. The estimated value of the carrier phase by the carrier phase estimation function is a bit determination function that determines the data value or bit value represented by the carrier phase in the signal to be modulated and reduces the influence of signal distortion that depends on the data pattern such as phase distortion. Connected to 236. Therefore, the output of the bit determination function is the carrier frequency λ._N It represents the modulated data and can be connected to path 118-N. Any training sequence bit error rate test function 238 can be configured to perform a bit error rate (BER) test on the training sequence for motion training of the carrier phase estimation function.

The LO frequency offset correction function 232 can be configured to track and correct the frequency offset between the received signal and the LO signal. FIG. 3 is a block diagram of an exemplary embodiment of a receiver 300 incorporating a dual stage LO frequency offset correction function 232 conforming to the present disclosure. The illustrated exemplary embodiment comprises a coherent receiver 202 that receives a data-modulated optical signal. The dual stage LO frequency offset correction function 232 has a feed-forward FOC function 302 that performs low-speed drift local oscillator frequency offset correction and a judgment feedback FOC function 304 that performs high-speed drift local oscillator frequency offset correction. Fast drift frequency offset correction corrects frequency offsets that change at relatively high speeds (eg at GHz speeds), and low speed drift frequency offset correction corrects frequency offsets that change at relatively low speeds (eg at MHz speeds). To do. When estimating the carrier phase in demodulating a received signal, the local oscillator frequency offset is indicated by the cumulative phase change. An example of the low-speed drift local oscillator frequency offset is generally represented by the slope of the phase curve of FIG. 6, and an example of the high-speed drift local oscillator frequency offset is represented by the slope of the phase change of FIG.

Usually, the estimated value of the local oscillator frequency offset can be determined from, for example, the carrier phase estimated value by the CPE function 234. The feedforward FOC function 302 can be implemented as part of any feedforward CPE function used to estimate the carrier phase of the optical signal, eg, CPE function 234 of FIG. The decision feedback FOC function 304 can be implemented as part of any decision feedback CPE function used to estimate the carrier phase of the optical signal, eg, the CPE function of FIG. The DSP circuit can be shared for the processing of the feed-forward FOC function and the feed-forward CPE function, and / or the DSP circuit can be shared for the processing of the judgment feedback FOC function and the judgment feedback CPE.

The feedforward CPE function capable of supplying an estimate of the carrier phase used for the feedforward FOC function has, for example, a costas loop function, an average phase function, and an M-th power method. An example of a coherent receiver configuration using the M-th power method for feed-forward CPE is incorporated herein by reference in Ly-Gagon et al., "Coherent Detection of Optical Quadrature Phase". -Shift Keying Signals With Carrier Phase Optimization ", Optical Wave Technology Journal, Vol.24, No.1, pp.12-21 (January 2006) and R.Noe's" PLL-Free Synchronous QPSK Polarization Multiplex / Diversity Receiver Concept " With Digital I &amp; Q Baseband "Processing", IEEE Photonics Letter, Vol.17, No.4.pp.887-889 (April 2005). In addition, the explanation and comparison of the M-th power method and the judgment feedback CPE scheme used for DSP-based coherent receivers are cited here, and the teaching contents are fully incorporated in the present specification. Described in "Comparison of Two Carrier Phase Optimization Schemes in Optical Coherent Detection Systems" by Cai and Alexei N. Pilipetskii, Optical Fiber Communications and NFOEC, 2007, pages 1-3 (March 2007).

The judgment feedback CPE function can obtain an estimated value of the carrier phase used by the judgment feedback FOC function. The output of the judgment feedback FOC function 304 can be determined according to the feedback 306 from the bit judgment function 236. In operation, the feedforward FOC function of the first stage performs initial frequency offset correction that corrects the low-speed drift local oscillator frequency offset, and the judgment accuracy that can operate the judgment feedback FOC function of the second stage is established. To. The decision feedback FOC function corrects the fast-changing local oscillator frequency offset. FIG. 4 is a simplified block diagram of an exemplary embodiment of an optical communication system 400 that incorporates dual stage frequency offset correction functions 302 and 304 in combination with dual stage carrier phase estimation functions 404 and 406 that conform to the present disclosure. Is. For convenience, the illustrated system has only one coherent receiver 202 that receives only one wavelength. However, it goes without saying that this system can be configured as a WDM system consisting of a demultiplexer and multiple receivers that receive multiple wavelengths. The feedforward FOC function 302 and the feedforward CPE function 404 are shown as separate functional blocks, and the judgment feedback FOC function 304 and the judgment feedback CPE function 406 are shown as separate functional blocks. As you can, you can perform FOC functions along with CPE functions. In the illustrated exemplary embodiment, transmitter 402 is configured to selectively transmit user data in mode of operation and training sequence in training mode, eg, pseudo-random bit sequence (PRBS). can do. The transmitter output is connected to the optical transmission path 102 and transmitted to the coherent receiver 202. The coherent receiver 202 mixes the received optical signal with a local oscillator signal to generate one or more electrical signals that represent user data. Next, the feedforward FOC function 302, the feedforward CPE, the determination feedback FOC304, and the determination feedback CPE406 perform frequency offset correction and carrier phase estimation according to the received electric signal. The feedforward CPE function 404 allows the feedforward FOC function 302 to supply the initial CPE used to determine the low speed drift local oscillator frequency offset amount.

The decision feedback CPE function 406 can supply the operating CPE and / or the initial CPE according to the received signal. When determining the operating CPE, the decision feedback FOC function 304 obtains an estimate of the fast drift local oscillator frequency offset and corrects the estimated fast drift frequency offset. The operation CPE output from the judgment feedback CPE function 406 is connected to the bit judgment function 236. The bit determination function uses the operating CPE to determine the data value or bit value represented by the carrier phase of the modulated signal and generate an output representing the data modulated to the carrier wavelength. Further, the bit determination function supplies the feedback 410 for performing the determination feedback CPE together with the determination feedback frequency offset correction to the determination feedback CPE function 406 and / or the determination feedback FOC function 304.

In one embodiment, the feedforward CPE function 404 can be operated in training mode. In training mode, transmitter 402 can transmit a training sequence. The CPE set by the feed-forward CPE function 404 can be connected to the training sequence bit error rate test function 408, and the training sequence bit error rate test function 408 can synchronize the training sequence using the CPE. It is determined whether or not it is. For example, in one embodiment, phase ambiguity from the feedforward CPE function is achieved by rotating through possible phase rotation values in the signal and synchronizing the training sequence using the training sequence tester 238. Can be removed. In this embodiment, the feedforward CPE function 404 does not supply the initial CPE until the training sequence bit error rate test function 408 synchronizes the training sequence with the CPE from the feedforward CPE function. When the training mode ends, the system enters operating mode and transmitter 402 can transmit user data to the transmission line. In the operating mode, the initial CPE is not supplied to the decision feedback CPE function 406. However, by running the training mode on a regular basis, the carrier phase estimation can be maintained correctly.

The initial stage feedforward FOC function 302 supplies an initial FOC that corrects the low-speed drift local oscillator frequency offset, and the determination accuracy is established so that the determination feedback FOC function 304 and the determination feedback CPE function 406 can be operated. The decision feedback FOC function 304 corrects the fast-changing local oscillator frequency offset. Further, when the judgment feedback CPE function alone is not practical because the judgment accuracy is insufficient, the judgment accuracy that can operate the judgment feedback CPE function is established by the initial CPE set by the feedforward CPE.

Figures 5 and 8 show a dual stage frequency offset correction process that fits this disclosure. The illustrated flowchart for explaining many embodiments includes a specific step sequence. However, the step sequence is merely an example of how the general functions described herein are performed. Also, each step sequence need not be executed in the order shown, unless otherwise specified.

FIG. 5 is a flow chart showing a first stage frequency offset estimation process 500 capable of performing low speed drift frequency offset correction according to the present disclosure. Typically, the first stage frequency offset estimation process 500 uses a feed-forward CPE process to determine the carrier phase estimate used to determine the frequency offset estimate. In particular, the illustrated exemplary embodiment uses the M-th power scheme in the feedforward CPE process. The M-th power method is a PSK-formatted signal with an M value (ie, M = 2 for BPSK, M = 4, etc. for QPSK) by averaging the N adjacent symbols whose signal sample is M-th power. Estimate the phase of the current symbol of. However, any system or method conforming to this disclosure is not limited to using a particular feedforward CPE process for the feedforward FOC process.

In the exemplary embodiment shown in FIG. 5, S_iIs the i-th signal sample, L_iIs the i-th local oscillator sample and N is the total number of processed symbol samples. First, steps 502, 504, and 506 calculate the current phase from the received sample, for example using the M-th power CPE calculation. In step 508, for example, the phase change (dphase) between consecutive symbols is calculated by subtracting the previous phase of the previous symbol from the current phase. To correct the phase jump, the calculated phase change is compared to -π / M and π / M in steps 510 and 512, and the corrected phase change is determined in steps 516 and 518. Next, in step 514, the estimated value of the carrier phase is determined based on, for example, the carrier phase obtained by adding an appropriate phase change to the pre-signal sample. In step 520, the previous phase is set to the current phase and the process is repeated for the number of symbol samples to process. Next, in step 522, the local oscillator frequency offset estimate is determined, for example, by determining the slope of the phase change of all processed symbol samples. The correction of this estimated frequency offset is S_i × L_iThis can be achieved by subtracting the phase change due to frequency offset from the sample.

The resulting phase estimate over a period of time has the characteristics shown in FIG. 6, ie the plot 600 of cumulative phase changes between consecutive symbols for the sample number (bit index). There is. The constant gradient of the cumulative phase shown in FIG. 6 indicates a constant local oscillator frequency offset from the received signal.

FIG. 7 is a diagram showing a plot 700 of the cumulative phase change between consecutive symbols with respect to the sample number (bit index) after the frequency offset correction of the first stage. Plot 700 in FIG. 7 shows a frequency offset with a small cumulative phase change but a relatively fast change. The fast changing frequency offset is corrected by a second stage FOC process performed by the decision feedback FOC function conforming to the present disclosure.

FIG. 8 is a flowchart showing a second stage frequency offset correction process 800 capable of performing high speed drift frequency offset correction according to the present disclosure. Usually, the second stage frequency offset uses the decision feedback CPE to determine the carrier phase estimate used to determine the frequency offset estimate. In particular, the estimated value of the frequency offset, that is, the slope of the phase change, can evenly divide the averaging window of the judgment feedback CPE into two and determine the slope of the phase change of the entire averaging window.

In the exemplary embodiment shown in FIG. 8, φ_kIs the kth (signal x local oscillator) phase, φ_dk Is the kth judgment data phase, and N is the average length of CPE. First, in step 802, the determination feedback CPE is determined based on the average phase of the averaging window, and in step 804, the average phase of the first and second halves of the CPE averaging window is calculated, respectively. If it is determined in step 806 that frequency offset correction is necessary, the amount of frequency offset in the averaging window is determined and corrected from the calculated gradient of the phase change between the two average phases in step 808. In step 810, a determination is made to determine the data value or bit value represented by the carrier phase of the signal to be modulated from the estimated value of the carrier phase, and the subsequent estimation of the carrier phase is performed by determination feedback.

Although the principle of the present invention has been described, this description is merely an example for those skilled in the art and is not construed as limiting the scope of the present invention. In addition to the exemplary embodiments illustrated and described herein, other embodiments are possible within the scope of the invention. Modifications and substitutions by those skilled in the art are to be construed as included in the scope of the invention, and the scope of the invention is limited only by the following claims.

100 WDM transmission system 102 Optical information path 104 Sending terminal 106 Receiving terminal 108 input port 110 route 112 Multiplexer / Coupler 114 Demultiplexer 116 route 118 Output path TX transmitter RX receiver