WO2014205893A1

WO2014205893A1 - Receiving end equalization method and system based on frequency domain channel estimation

Info

Publication number: WO2014205893A1
Application number: PCT/CN2013/081341
Authority: WO
Inventors: 张帆; 郑桢楠; 陈章渊
Original assignee: 北京大学
Priority date: 2013-06-27
Filing date: 2013-08-13
Publication date: 2014-12-31
Also published as: CN103338171A; CN103338171B

Abstract

The present invention relates to a receiving end equalization method and system based on frequency domain channel estimation. The method comprises: first, inserting a training sequence with a flat spectrum characteristic into a transmission end signal; then, transforming the received training sequence into a frequency domain at a receiving end with a fast Fourier transform; dividing by a corresponding ideal frequency spectrum to obtain a transfer function of a channel, and then taking the reciprocal of the transfer function and transforming same into a time domain with an inverse fast Fourier transform, so as to obtain a tap coefficient of a time domain filter; and after this, filtering a data signal in the time domain with the tap coefficient of the time domain filter, so as to realize signal equalization. The system contains various modules for realizing the corresponding functions. The present invention combines the characteristics of time domain equalization, without inserting a cyclic prefix/suffix causing an overhead, and simple channel estimation of frequency domain equalization, and can effectively conduct phase noise compensation in the case where chromatic dispersion of an optical fibre is small or is compensated by means of other algorithms, and the algorithm complexity is small, which can save calculation time.

Description

Receiver equalization method and system based on frequency domain channel estimation

The present invention relates to the field of coherent optical communication transmission, and relates to a receiving end equalization method based on frequency domain channel estimation, and a system for implementing the method. Background technique

Coherent communication based on digital signal processing technology is an important solution for long-distance optical fiber communication transmission systems. How to overcome the communication channel damage is an important issue to be solved. The impairment of the communication channel reduces the signal-to-noise ratio of the signal and introduces inter-symbol interference, resulting in the generation of errors. In fiber-optic communication systems, the equalization algorithm estimates the channel and compensates for linear distortion of the channel to attenuate or eliminate inter-symbol interference.

At present, there are two main types of equalization algorithms, one is time domain equalization (TDE) and the other is frequency domain equalization (FDE). Both types of algorithms have the same equalization effect.

1) Time Domain Equalization (TDE) algorithm. The algorithm performs channel impulse response estimation based on the training sequence or signal-based constellation characteristics, and compensates the signal in the time domain. The channel estimation algorithm based on the training sequence is more complicated. The channel estimation algorithm based on the characteristics of the signal constellation depends on the modulation format. When the modulation format changes, the algorithm also needs to change, and the algorithm effect is different. Due to the dispersion in the fiber, when the dispersion is large, the equalization algorithm requires more equalizer taps, resulting in higher algorithm complexity.

2) Frequency Domain Equalization (FDE) algorithm. The algorithm performs estimation of the channel transfer function based on the training sequence and compensates the signal in the frequency domain. The channel estimation algorithm is simple, and when the dispersion is large, the algorithm complexity is still low. However, the algorithm requires a cyclic prefix/suffix to be inserted into the signal, which causes a certain signal rate overhead. Summary of the invention

The invention provides a simple and effective receiving end equalization method based on frequency domain channel estimation, and a system for realizing the method, which can effectively perform phase noise compensation in the case where the fiber dispersion is small or the dispersion is compensated by other algorithms.

To achieve the above object, the present invention adopts the following technical solutions:

An equalization method based on frequency domain channel estimation includes the following steps:

Step 1: Insert a training sequence with flat spectral characteristics into the transmitter signal;

The second step: transforming the received training sequence into the frequency domain by using a fast Fourier transform (FFT) at the receiving end, The ideal spectrum corresponding to the training sequence (that is, the spectrum of the theoretical sequence) is divided to obtain the transfer function of the channel, and then the reciprocal of the transfer function is taken and transformed into the time domain by the inverse fast Fourier transform (IFFT) to obtain the time domain. Filter tap coefficient; Step 3: Use the obtained time domain filter tap coefficients to filter the data signal in the time domain to achieve signal equalization. Preferably, the training sequence is an M sequence or a chu sequence.

Further, the transmitting end adds a cyclic prefix/suffix to the training sequence to eliminate inter-code crosstalk; the receiving end first removes the cyclic prefix/suffix in the received training sequence, and then performs the fast Fourier Leaf transformation.

Further, the receiving end first performs front-end data processing on the signal received by the transmitting end, including dispersion coarse compensation, carrier frequency recovery, receive matching filtering, and digital synchronization; and performs back-end data processing after step 3), including carrier Phase recovery.

Further, the signal at the transmitting end is frame-transmitted, and the frame structure of each frame includes two polarization directions, each of which includes a training sequence and a data signal, and the training sequence is composed of a pair of parts, labeled as ^ and ^ ² , alternately arranged in time, and:

Wherein t _x and ^ are M sequences or chu sequences, and "0" represents a sequence having a length equal to 0 in length and a chu sequence. Furthermore, the selection of the training sequence length is determined by the channel carrier frequency offset and the degree of phase drift; the selection of the number of training sequences is determined by the spontaneous radiated noise intensity in the channel; the selection of the data signal length is determined by the drift of the channel transfer function. Decide.

An equalization system based on frequency domain channel estimation for implementing the above method, comprising a transmitting end and a receiving end,

The transmitting end includes:

a training sequence insertion module, configured to insert a training sequence with flat spectral characteristics in a signal of the transmitting end; a cyclic prefix/suffix adding module, and the training sequence inserting module is connected, and is used to add a cyclic prefix/suffix to the training sequence to eliminate Crosstalk between codes;

a transmit filtering module, configured to connect the cyclic prefix/suffix join module, configured to filter a signal after adding a cyclic prefix/suffix and send the signal to a communication channel;

The receiving end includes:

a cyclic prefix/suffix removal module for removing a cyclic prefix/suffix in a training sequence;

An FFT module, configured to connect the cyclic prefix/suffix removal module, to perform a fast Fourier transform on the training sequence with the cyclic prefix/suffix removed, and transform the frequency to the frequency domain; a channel transfer function estimating module, connected to the FFT module, configured to divide the ideal spectrum corresponding to the training sequence to obtain a transfer function of the channel;

An IFFT module, connected to the channel transfer function estimating module, configured to perform an inverse fast Fourier transform (IFFT) to transform a signal into a time domain, to obtain a time domain filter tap coefficient;

An FIR filter is coupled to the IFFT module for filtering the data signal in the time domain using the obtained time domain filter tap coefficients to achieve signal equalization.

Further, it also includes:

a front-end data processing module, configured to preprocess the signal received from the transmitting end, and then send the signal to the cyclic prefix/suffix removal module;

A back-end data processing module is coupled to the FIR filter for carrier phase recovery.

Further, the preprocessing performed by the front end data processing module includes: dispersion coarse compensation, carrier frequency recovery, receive matching filtering, and digital synchronization.

Further, a decision module is further included, and the back end data processing module is connected to recover the received signal information into binary data.

The method of the present invention combines the time-domain equalization non-insertion cyclic prefix/suffix to form a simple channel estimation of overhead and frequency domain equalization in the case where the dispersion of the optical fiber is small or the dispersion is compensated by other algorithms. This method is less complex and saves computation time. In theory, the equalization method of the present invention operates on a data signal equivalent to a frequency domain equalization algorithm, and thus the equalization effect is the same. DRAWINGS

1 is a flow chart of a method for equalization based on frequency domain channel estimation according to an embodiment of the present invention.

2 is a schematic structural diagram of a signal frame according to an embodiment of the present invention.

FIG. 3 is a schematic structural diagram of a frequency domain channel estimation based equalization system according to an embodiment of the present invention.

4 is a schematic diagram showing the experimental results of a 1.76 Tb/s signal 720 km transmission according to an embodiment of the present invention. detailed description

The present invention will be further described in detail below by way of specific embodiments and the accompanying drawings.

In the present invention, the relationship between the transmitted signal and the received signal is as follows: C(k) = -— = 丽 → s _tr (n) = r _tr {n) * c(l)

H (k) R _tr (k) ^{tr tr} where is the transmitted training sequence, (") is the received training sequence, their corresponding frequency domain form is the sum of the frequency domain equalizer tap, is the receiver estimation The channel transfer function, c(l), is the tap coefficient of the time domain finite impulse response (FIR) filter.

The implementation of the technical solution will be specifically described below with reference to the algorithm flow chart 1 of the present embodiment. The portion shown by the dotted line on the right in Fig. 1 is the main content of the solution of the present invention.

Step 1: Insert several training sequences at the front end of the signal at the transmitting end. Take the chu sequence as an example. The chu sequence is:

C(n) = e

2, 』 where N is the length of the chu sequence. The chu sequence has the characteristics of a flat spectrum, and a sequence of M sequences having similar characteristics can also be selected in this step, and the implementation using the M sequence is the same as that using the chu sequence.

At the same time, since the receiving end needs to transform the training sequence into the frequency domain for operation, a cyclic prefix/suffix is added to the training sequence at the transmitting end to eliminate crosstalk between codes. The signal is then filtered and transmitted to the receiving end.

Step 2: At the receiving end:

The training sequence received at the receiving end is distorted by the channel and is noisy. The receiving end first performs front-end data processing on it, including dispersion coarse compensation, carrier frequency recovery, receive matching filtering, digital synchronization, and the like.

Then, the cyclic prefix/suffix of the training sequence is removed, and the received training sequence is transformed into the frequency domain by Fast Fourier Transform (FFT), and divided by the spectrum of the ideal training sequence to obtain a channel transfer function, and the reciprocal thereof is used. IFFT fast inverse Fourier transform (IFFT) transform to the time domain, resulting in the time domain filter tap coefficients: s _tr (k)

c(l) = IDFT[C(k)] = IDFT

R _tr (k) where IDFT is the inverse discrete Fourier transform, and IFFT is a fast implementation method.

Then, using the resulting time domain filter (FIR filter) tap coefficients (shown in FIRtaps in Figure 1), the data signal is filtered in the time domain, that is, the data signal is linearly convolved with the filter tap, which is equalization.

Finally, back-end data processing (carrier phase recovery) and decision making are performed.

The above method requires signal framing transmission, and Figure 2 shows the frame structure of one frame of the signal, including two polarization directions. Each polarization direction includes a training sequence portion and a data signal portion. The training sequence consists of pairs of parts, labeled ^ and ί ₂ , alternated in time, and:

Wherein t _x and ^ are M sequences or chu sequences, and "0" represents a sequence having a length equal to 0 in length and a chu sequence. The choice of the length of the M-sequence (or chu sequence) during implementation is determined by the channel carrier frequency offset and the degree of phase drift, which is required to be short enough to ensure that the carrier frequency offset and phase are stable over a pair of training sequences.

The choice of the number of M-sequences (or chu sequences) during the implementation is determined by the intensity of the spontaneous radiated noise in the channel, which is needed to ensure that the error caused by the spontaneous radiated noise on the channel estimation at the receiving end is small.

The choice of the length of the data signal during implementation is determined by the degree of drift of the channel transfer function, which is required to be short enough to ensure that the transfer function of the channel does not undergo large drifts within the frame.

FIG. 3 is a schematic diagram of the composition of an equalization system based on frequency domain channel estimation corresponding to the above method, including a transmitting end and a receiving end.

The transmitting end includes: a training sequence insertion module, configured to insert a training sequence with flat spectral characteristics in the transmitting end signal; a cyclic prefix/suffix adding module, and the training sequence insertion module is connected, and the cyclic prefix is added to the training sequence / suffix to eliminate inter-code crosstalk; a transmit filter module, connected to the cyclic prefix / suffix join module, for filtering the signal after adding the cyclic prefix / suffix and transmitting to the receiving end;

The receiving end includes: a front end data processing module, configured to preprocess the signal received from the transmitting end, including dispersion coarse compensation, carrier frequency recovery, receive matching filtering, digital synchronization, etc., and then sent to the cyclic prefix/suffix removal a cyclic prefix/suffix removal module for removing a cyclic prefix/suffix in the training sequence; an FFT module, connecting the cyclic prefix/suffix removal module, for performing a training sequence with a cyclic prefix/suffix removed a fast Fourier transform, which is transformed into a frequency domain; a channel transfer function estimation module, connected to the FFT module, configured to divide the ideal spectrum corresponding to the training sequence to obtain a transfer function of the channel; an IFFT module, connecting the channel transfer a function estimation module, configured to perform an inverse fast Fourier transform (IFFT) to transform a signal into a time domain, to obtain a time domain filter tap coefficient; and an FIR filter connected to the IFFT module for utilizing the obtained time domain filter The tap coefficient filters the data signal in the time domain to achieve signal equalization; the back-end data processing module, The FIR filter is connected for carrier phase recovery; and the decision module is connected to the backend data processing module for restoring the received signal information into binary data.

Figure 4 shows the results of the 1.76 Tb/s transmission experiment. The signal is the transmission distance of 720 km, and the horizontal axis is the fiber input power of the signal. The common algorithm of the equalization method of the present invention and the time domain equalization algorithm (the norm algorithm (CMA) plus the cascade multi-mode algorithm (CMMA)) For comparison, the results show that the equalization algorithm of the present invention is superior to the commonly used time domain equalization algorithm because the noise of the channel has a large influence on the filter tap estimation of the time domain equalization algorithm.

The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to be limiting, and those skilled in the art can make modifications or equivalents to the embodiments of the present invention without departing from the spirit and scope of the present invention. The scope of protection shall be as stated in the claims.

Claims

claims

1. An equalization method based on frequency domain channel estimation, including the following steps:

1) Insert a training sequence with flat spectrum characteristics into the transmitter signal;

2) Use fast Fourier transform at the receiving end to transform the received training sequence into the frequency domain, then divide the ideal spectrum corresponding to the training sequence to obtain the transfer function of the channel, take the reciprocal of the transfer function and use fast The inverse Fourier transform is transformed to the time domain to obtain the time domain filter tap coefficients;

3) Use the time domain filter tap coefficients to filter the data signal in the time domain to achieve signal equalization.

2. The method according to claim 1, characterized in that: the training sequence is an M sequence or a chu sequence.

3. The method of claim 1, characterized in that: the transmitting end adds a cyclic prefix/suffix to the training sequence to eliminate inter-code crosstalk; the receiving end first removes cyclic prefixes/suffixes from the received training sequence. Cyclic prefix/suffix, then perform the Fast Fourier Transform.

4. The method according to claim 1, characterized in that: the receiving end first performs front-end data processing on the signal received by the transmitting end, including coarse dispersion compensation, carrier frequency recovery, receiving matching filtering and digital synchronization; and Step 3) is followed by back-end data processing, including carrier phase recovery.

5. The method of claim 1, characterized in that: the signal at the transmitting end is transmitted in frames, the frame structure of each frame includes two polarization directions, each polarization direction includes a training sequence and a data signal, and the training sequence is composed of Made up of pairs of parts, labeled ^ and ^, alternating in time, and:

Among them, t _x and ^ are M sequences or chu sequences, and "0" represents a sequence with a value of 0 that is the same length as the chu sequence.

6. The method according to claim 5, characterized in that: the selection of the length of the training sequence is determined by the channel carrier frequency offset and the degree of phase drift; the selection of the number of training sequences is determined by the intensity of spontaneous emission noise in the channel; the data signal The choice of length is determined by the degree of drift in the channel transfer function.

7. An equalization system based on frequency domain channel estimation, including a transmitter and a receiver, characterized by:

The transmitter includes:

A training sequence insertion module, used to insert a training sequence with flat spectrum characteristics into the transmitter signal; a cyclic prefix/suffix adding module, connected to the training sequence insertion module, used to add a cyclic prefix/suffix to the training sequence to eliminate inter-code crosstalk; A transmit filtering module, connected to the cyclic prefix/suffix adding module, used to filter the signal after adding the cyclic prefix/suffix and send it to the communication channel;

The receiving end includes:

Cyclic prefix/suffix removal module, used to remove cyclic prefix/suffix in the training sequence;

The FFT module is connected to the cyclic prefix/suffix removal module and is used to perform a fast Fourier transform on the training sequence with the cyclic prefix/suffix removed and transform it into the frequency domain;

The channel transfer function estimation module is connected to the FFT module and is used to divide the ideal spectrum corresponding to the training sequence to obtain the transfer function of the channel;

The IFFT module is connected to the channel transfer function estimation module and is used to perform an inverse fast Fourier transform (IFFT) to transform the signal into the time domain to obtain the time domain filter tap coefficients;

The FIR filter is connected to the IFFT module and is used to filter the data signal in the time domain using the obtained time domain filter tap coefficients to achieve signal equalization.

8. The system of claim 7, further comprising:

The front-end data processing module is used to preprocess the signal received from the sending end and then send it to the cyclic prefix/suffix removal module;

The back-end data processing module is connected to the FIR filter for carrier phase recovery.

9. The system of claim 8, wherein the preprocessing includes: coarse dispersion compensation, carrier frequency recovery, receiving matched filtering, and digital synchronization.

10. The system according to claim 8, further comprising a decision module connected to the back-end data processing module and used to restore the received signal information into binary data.