Spectrum shaping method based on probability shaping

Abstract

The invention relates to the technical field of optical fiber communication, in particular to a frequency spectrum shaping method based on probability shaping, which combines probability shaping with a feedback filter in a signal generation stage and dynamically adjusts the output probability of the probability shaping, so that the frequency of a generated signal has a frequency spectrum narrowband sagging effect before filtering, and the frequency spectrum narrowband sagging effect is improved. And meanwhile, the output level of the signal is not influenced, and the cost is slight spectral efficiency reduction in the shaping stage. The method can be suitable for various modulation formats such as PAM, QAM and PSK under direct modulation and coherent modulation, and is compatible with a common static probability shaping technology of a coherent system. Through the steps, the probability screening of the output coding sequence is realized, the purpose of equivalent filtering is achieved, and the loss of signal quality is avoided.

Description

Technical Field

[0001] The present invention relates to the field of optical fiber communication technology, and more specifically, to a spectrum shaping method based on probability shaping. Background Art

[0002] In the field of single-carrier optical communications, due to the pursuit of spectrum efficiency in recent years, the optical bandwidth of analog signals has gradually decreased to near the baud rate (±1 / 2 baud rate range). Due to the lack of out-of-band signals, these signals are often difficult to estimate frequency deviation and recover clocks in the receiver. If a clock signal can be added within the signal bandwidth, such as at 1 / 4 baud rate frequency, it can greatly facilitate the demodulation of the receiver. However, directly superimposing the clock signal on the basis of the original signal will cause mutual interference between the two. On the one hand, filtering the residual clock signal will reduce the signal-to-noise ratio, and on the other hand, the incompletely separated signal will also affect the purity of the clock. Therefore, it is a prerequisite to use narrowband filtering and other methods to suppress the signal strength at the clock insertion frequency. However, traditional optical or DSP narrowband filtering also means additional signal quality loss. Summary of the invention

[0003] In order to overcome the defect of additional signal quality loss caused by filtering in the above-mentioned prior art, the present invention provides a spectrum shaping method based on probability shaping, which realizes the probability screening of the output coding sequence, achieves the purpose of equivalent filtering, and avoids the loss of signal quality.

[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0005] A spectrum shaping method based on probability shaping comprises the following steps:

[0006] Encode the input binary information bit by bit: For a sequence of length N, bit-by-bit encoding requires N cycles. During the encoding process, V is always guaranteed to be l ≤U l , U h ≤V h , where, for an equiprobable binary distribution, a binary sequence of length N is replaced by a unique segment of length 2 in a line segment of length 1. -N The line segment U represents that the lower end point of the line segment U is defined as U l , the top dead center is U h ; For binary distributions with unequal probabilities, each sequence is represented by a unique line segment V in a line segment of length 1. The length of the line segment V is proportional to the probability of the sequence appearing. The lower end point of the line segment V is defined as V l , top dead center is V h , the next split position is V m ;

[0007] The bit-by-bit encoding steps include:

[0008] S11. Update U for the kth bit of the current input l and U h , the length of U becomes half;

[0009] S12. For the current V l and V h , calculate the segmentation position V m , check whether V m ≤U l or U h ≤V m , if satisfied, the output code will be 0 or 1 accordingly, V l and V h Update accordingly. If V m ≤U l , then let V l =V m ;if U h ≤V m , then let V h =V m ;

[0010] S13. According to the current past output sequence S out , adjust V l and V h Each lower decision point V within the range 1,2… , so that the probability of each symbol appearing at the lower level increases with C k+1,n Adjust the size of C k+1,n Indicates that for each preset frequency point f n Calculate filtering results using feedback filters;

[0011] S14. Check [V l ,V h ] Is it in 2 -m-1 Length is step, length is 2 -m If so, U and V are uniformly scaled by remapping the corresponding line segment to [0,1]. During the algorithm implementation, since the line segment length decreases exponentially, the calculation accuracy cannot be guaranteed for the limited number of computer bits, so it is also necessary to uniformly scale U and V as a whole under appropriate circumstances. l ≤U l , U h ≤V h , so we only need to consider 0≤V after scaling l , V h ≤1. The scaling judgment method is: check [V l ,V h] is inside [0,1 / 2] or [1 / 4,3 / 4] or [1 / 2,1]. If so, scale U and V uniformly by remapping the corresponding line segment to [0,1]. This check step can also be expressed as: check [V l ,V h ] Is it in 2 -m-1 The length is step, length is 2 -m within the line segment.

[0012] Finishing process:

[0013] S21. After the input signal ends (the input binary signal is generally segmented, here it means after the last bit of a segment is processed according to the above process), calculate the segmentation position V m ,Right now Check whether U is satisfied l ≤V m ≤(U l +U h ) / 2 or (U l +U h ) / 2≤V m ≤U h , output 0 or 1 accordingly and update V, repeat the above steps S11 to S14, so that the final U l ≤V l ≤(U l +U h ) / 2≤V h ≤U j .

[0014] According to the above technical means, the present invention provides a spectrum shaping method based on probability shaping, which combines probability shaping with a feedback filter in the signal generation stage, and dynamically adjusts the output probability of probability shaping, so that the generated signal frequency has the effect of spectrum narrowband depression before filtering, and does not affect the output level of the signal at the cost of a slight decrease in spectrum efficiency in the shaping stage. This method can be applied to a variety of modulation formats such as PAM, QAM, PSK under direct modulation and coherent modulation, and is compatible with the static probability shaping technology commonly used in coherent systems. Through the above steps, the present invention realizes the probability screening of the output coding sequence, achieves the purpose of equivalent filtering, and avoids the loss of signal quality.

[0015] Furthermore, after each step S12 is executed, a feedback calculation loop is added to calculate the current output sequence S out (In the case of IQ modulation such as QAM, S out Can be a complex signal), adjust V l / h Each lower decision point V within the range 1,2… The steps can be described by two steps of calculation; including:

[0016] S131. For each preset frequency point f b , -1 / 2≤f n ≤1 / 2, use the feedback filter with exp decay to calculate the filtering result:

[0017]

[0018] In the formula, subscript k indicates the cycle in which the kth bit is output, subscript n indicates the frequency point number, and a n represents f n The attenuation coefficient corresponding to the frequency;

[0019] S132. Update the lower decision point V 1,2… , define P as V 1,2… The corresponding default probability distribution vector, b n is the frequency point f n The corresponding feedback modulation depth vector makes the probability of each symbol appearing at the lower level increase with C k+1,n To adjust the size:

[0020] V 1,2… =V l +(V h -V l )·(P+∑ n b n ·C k+1,n ).

[0021] Therefore, the probability screening of the output coding sequence is realized in the process of probability shaping, thereby achieving the purpose of equivalent filtering.

[0022] Furthermore, when U l ≤V l ≤(U l +U h ) / 2≤V h ≤U h , then line segment V can uniquely determine line segment U, that is, the codes of unequal probability corresponding to V are the encoding results of the input sequence.

[0023] Furthermore, in the bit-by-bit encoding process, steps S11 to S14 are looped multiple times, and in each loop, steps S12, S13, and S14 are executed once or more than once. The above S11 to S14 are one loop, and it should be noted that in the actual process, each loop of S12, S13, and S14 may be executed more than once, such as S11-S12-S12-S13-S14-S14.

[0024] The present invention provides a probability shaping single-carrier communication method for realizing clock simultaneous transmission, comprising the following steps:

[0025] Transmitter:

[0026] A11. The binary information executes the above-mentioned spectrum shaping method based on probability shaping to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and satisfying the preset probability distribution.

[0027] A12. The signal sequence is resampled, spectrum shaped, and clock inserted. The sampling rate is adjusted from the baud rate to the target sampling rate in the digital domain. At the same time, a digital clock signal is inserted at the suppression frequency during the resampling process, and the spectrum is shaped, such as raised cosine, root raised cosine, etc.

[0028] A13. The resampled signal sequence is converted from digital to analog to output an analog electrical signal;

[0029] A14. The modulator uses the laser as a carrier, modulates the electrical signal into the spectrum, and outputs a single carrier signal in the optical domain;

[0030] Receiver:

[0031] A21. The optical domain single carrier signal passes through the mixer and is downconverted to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiving end.

[0032] A22. The analog electrical signal is sampled into a digital electrical signal by an analog-to-digital converter;

[0033] A23. The digital electrical signal is resampled and the clock is extracted, usually by narrowband filtering, such as static filter, digital phase-locked loop and other structures, to separate the clock from the digital electrical signal and analyze the clock error information, including frequency and phase error;

[0034] A24. Sampling clock compensation for digital electrical signals;

[0035] A25. The digital electrical signal passes through the equalization module, such as CMA, LMS and other algorithms, to compensate for the filtering effects of the channel and the device;

[0036] A26. Digital electrical signals are used to compensate for the frequency deviation caused by the frequency difference between the transmitting and receiving lasers;

[0037] A27. The digital electrical signal is processed by carrier phase recovery to compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers;

[0038] A28. The digital electrical signal passes through the probability shaping decoding module to restore the binary information of the transmitter.

[0039] Further, in step A12, in order to ensure the quality of the transmitted signal and the receiving performance, the analog signal spectrum is usually slightly larger than 1 / 2 baud rate, so it is necessary to upsample the baud rate signal sequence and perform a raised cosine / root raised cosine roll-off filter at the frequency edge, which is therefore called (conventional) spectrum shaping. Since the signal sequence has been processed by the probability shaping module and has been suppressed at a specific frequency, it is necessary to insert a clock there, and this step is implemented before or after resampling as appropriate. The reason for the difference between inserting the clock before or after sampling is that the main purpose of inserting the clock is to facilitate synchronous sampling, but the process of determining synchronization inside the receiver can be implemented before or after resampling. Therefore, the module can be selectively combined, such as the traditional structure of resampling + spectrum shaping. For a single-carrier system with baud rate sampling, in order to ensure the demodulation performance, a clock is required to assist, and only a clock insertion module can be included (at this time, the optimal clock insertion frequency is ±1 / 4 baud rate).

[0040] Furthermore, in step A24, the sampling delay is corrected by a clock recovery algorithm, or the phase error of the clock is fed back to the sampling clock, and the sampling clock is locked in hardware.

[0041] Furthermore, in step A25, the frequency error information of the clock is compensated by an algorithm, or the frequency error is fed back to the laser with frequency control, and the laser frequency deviation is locked in hardware. Compensating the laser frequency deviation is equivalent to multiplying the complex signal at the receiving end by an e 2πj·Δft .

[0042] The present invention provides a probability shaping single-carrier communication system for realizing clock simultaneous transmission, comprising:

[0043] The transmitter includes:

[0044] Probabilistic shaping module: used to perform the above-mentioned spectrum shaping method based on probability shaping on binary information to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and meeting the preset probability distribution;

[0045] Resampling spectrum shaping and clock insertion module: used to resample, spectrum shape and clock insert the signal sequence, adjust the sampling rate from the baud rate to the target sampling rate in the digital domain, insert a digital clock signal at the suppression frequency during the resampling process, and then shape the spectrum.

[0046] Digital-to-analog conversion module: used to convert the resampled signal sequence into digital-to-analog signals and output analog electrical signals;

[0047] Modulator module: used to use the laser as a carrier, modulate the electrical signal into the spectrum, and output a single carrier signal in the optical domain;

[0048] The receiving end includes:

[0049] Mixer: used to down-convert the optical domain single carrier signal output by the transmitter to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiver;

[0050] Analog-to-digital conversion module: used to sample analog electrical signals into digital electrical signals through an analog-to-digital converter;

[0051] Resampling and clock extraction module: used to resample digital electrical signals and extract clocks, separate clocks from digital electrical signals, and analyze clock error information;

[0052] Clock recovery module: used to perform sampling clock compensation on the digital electrical signal processed by the resampling and clock extraction modules;

[0053] Equalization module: used to equalize the digital electrical signal processed by the clock recovery module and compensate for the filtering effect of the channel and the device;

[0054] Frequency deviation compensation module: used to compensate the frequency deviation caused by the frequency difference between the transmitting and receiving lasers of the digital electrical signal processed by the equalization module;

[0055] Carrier phase recovery module: used to perform carrier phase recovery processing on the digital electrical signal processed by the frequency offset compensation module, and compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers;

[0056] Probability shaping decoding module: used to restore the binary information of the transmitting end through the probability shaping decoding module of the digital electrical signal processed by the carrier phase recovery module.

[0057] Furthermore, in the resampled spectrum shaping and clock insertion module, in order to ensure the quality of the transmitted signal and the receiving performance, the analog signal spectrum is greater than 1 / 2 the baud rate, the baud rate signal sequence is upsampled, and a raised cosine / root raised cosine roll-off filter is performed at the frequency edge.

[0058] Compared with the prior art, the beneficial effects are as follows: a spectrum shaping method based on probability shaping proposed by the present invention combines probability shaping with a feedback filter in the signal generation stage, and dynamically adjusts the output probability of probability shaping, so that the generated signal frequency has the effect of spectrum narrowband depression before filtering, and does not affect the output level of the signal, at the cost of a slight decrease in spectrum efficiency in the shaping stage; the probability screening of the output coding sequence is realized, the purpose of equivalent filtering is achieved, and the loss of signal quality is avoided; this method can be applied to a variety of modulation formats such as PAM, QAM, PSK under direct modulation and coherent modulation, and is compatible with the static probability shaping technology commonly used in coherent systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a flow chart of a spectrum shaping method based on probability shaping of the present invention.

[0060] Figure 2 This is a diagram showing the effect of 1±1 / 4 baud rate spectrum shaping in an embodiment of the present invention (clock has been inserted).

[0061] Figure 3 This is a schematic diagram of the principle of probability shaping sequence mapping based on algebraic probability matching.

[0062] Figure 4 It is a schematic diagram of the process of Example 1.

[0063] Figure 5 This is a schematic diagram of the transmitting end structure of Example 3.

[0064] Figure 6 This is a schematic diagram of the receiving end structure of Example 3. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. The present invention is described in one of the embodiments in combination with the specific implementation methods. Among them, the drawings are only used for exemplary descriptions, and only schematic diagrams are shown, not physical drawings, and cannot be understood as limitations on the present invention; in order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0066] In the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right" and the like indicate an orientation or positional relationship based on the orientation or positional relationship shown in the drawings, it is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation. Therefore, the terms describing the positional relationship in the drawings are only used for exemplary explanations and cannot be understood as limiting the present invention. For ordinary technicians in the field, the specific meanings of the above terms can be understood according to specific circumstances. In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present invention, the descriptions of "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" can explicitly or implicitly include at least one of the features. In addition, the meaning of "and / or" appearing in the full text is to include three parallel schemes. Taking "A and / or B" as an example, it includes scheme A, or scheme B, or schemes that satisfy both A and B.

[0067] Embodiment 1:

[0068] In this embodiment, based on static probability shaping, an algebraic operation method called distribution matching (DM) is used. This method can realize encoding and decoding operations between any different distributions and different bases. The following takes the mapping of equal probability binary distribution (0.5, 0.5) to unequal probability binary distribution (0.7, 0.3) as an example to explain its working principle and encoding steps.

[0069] principle:

[0070] For an equiprobable binary distribution (usually the original information is this distribution), a binary sequence of length N can be replaced by a unique segment of length 2 in a line segment of length 1. -N The line segment U is represented by Figure 3 As shown in Figure 2, this process can be regarded as a step-by-step retrieval process. Define the lower end point of the line segment as U l , the top dead center is U h ;

[0071] For binary distributions with unequal probabilities, each sequence can also be represented by a unique line segment V in a line segment of length 1. The length of this line segment is proportional to the probability of the sequence appearing. The lower end point of this line segment is defined as V. l , top dead center is V h , the next split position is V m ;

[0072] When U is satisfied l ≤V l ≤(Ul +U h ) / 2≤V h ≤U h , it can be considered that at this time segment V can uniquely determine segment U, that is, the codes of unequal probability corresponding to V are the encoding results of the input sequence.

[0073] In this embodiment, if Figure 1 As shown in FIG. 1 , a spectrum shaping method based on probability shaping is provided. The overall method is divided into two main steps, namely, bit-by-bit encoding and a tailing process. For a sequence of length N, the bit-by-bit encoding needs to execute N cycles, and finally ends with a tailing process. Figure 4 Right half.

[0074] (1) Encode the input binary information bit by bit: For a sequence of length N, bit-by-bit encoding requires N cycles. During the encoding process, V l ≤U l , U h ≤V h ,;

[0075] The bit-by-bit encoding steps include:

[0076] S11. Update U for the kth bit of the current input l and U h , the length of U becomes half;

[0077] S12. For the current V l and V h , calculate the segmentation position V m , check whether V m ≤U l or U h ≤V m , if satisfied, the output code will be 0 or 1 accordingly, V l and V h Update accordingly. If V m ≤U l , then let V l =V m ;if U h ≤V m , then let V h =V m ;

[0078] S13. According to the current past output sequence S out , adjust V l and V h Each lower decision point V within the range 1,2… , so that the probability of each symbol appearing at the lower level increases with C k+1,n Adjust the size of C k+1,nIndicates that for each preset frequency point f n Calculate filtering results using feedback filters;

[0079] S14. Check [V l ,V h ] Is it in 2 -m-1 Length is step, length is 2 -m If so, U and V are uniformly scaled by remapping the corresponding line segment to [0,1]. During the algorithm implementation, since the line segment length decreases exponentially, the calculation accuracy cannot be guaranteed for the limited number of computer bits, so it is also necessary to uniformly scale U and V as a whole under appropriate circumstances. l ≤U l , U h ≤V h , so we only need to consider 0≤V after scaling l , V h ≤1. The scaling judgment method is: check [V l ,V h ] is inside [0,1 / 2] or [1 / 4,3 / 4] or [1 / 2,1]. If so, scale U and V uniformly by remapping the corresponding line segment to [0,1]. This check step can also be expressed as: check [V l ,V h ] Is it in 2 -m-1 The length is step, length is 2 -m within the line segment.

[0080] In the bit-by-bit encoding process, steps S11 to S14 are looped multiple times, and in each loop, steps S12, S13, and S14 are executed once or more than once. The above S11 to S14 are one loop. It should be noted that in the actual process, each loop of S12, S13, and S14 may be executed more than once, such as S11-S12-S12-S13-S14-S14.

[0081] (2) Closing process:

[0082] S21. After the input signal ends (the input binary signal is generally segmented, here it means after the last bit of a segment is processed according to the above process), calculate the segmentation position V m ,Right now Check whether U is satisfied l ≤V m ≤(U l +U h ) / 2 or (U l +U h ) / 2≤V m ≤U h, output 0 or 1 accordingly and update V, repeat the above steps S11 to S14, so that the final U l ≤V l ≤(U l +U h ) / 2≤V h ≤U h .

[0083] like Figure 4 As shown in the figure, it is a schematic diagram of the flow chart of this embodiment, in which the output is multi-base, V m Replaced by V 1,2… , the bit-by-bit encoding process is the same, and the ending process is not shown in the figure.

[0084] Specifically, in the present invention, after each step S12 is executed, a feedback calculation loop is added to calculate the current output sequence S out (In the case of IQ modulation such as QAM, S out Can be a complex signal), adjust V l / h Each lower decision point V within the range 1,2… The steps can be described by two steps of calculation; including:

[0085] S131. For each preset frequency point f n , -1 / 2≤f n ≤1 / 2, use the feedback filter with exp decay to calculate the filtering result:

[0086]

[0087] In the formula, subscript k indicates the cycle in which the kth bit is output, subscript n indicates the frequency point number, and a n represents f n The attenuation coefficient corresponding to the frequency;

[0088] S132. Update the lower decision point V 1,2… , define P as V 1,2… The corresponding default probability distribution vector, b n is the frequency point f n The corresponding feedback modulation depth vector makes the probability of each symbol appearing at the lower level increase with C k+1,n To adjust the size:

[0089] V 1,2… =V l +(V h -V l )·(P+∑ n b n ·C k+1,n ).

[0090] Therefore, the probability screening of the output coding sequence is realized in the process of probability shaping, thereby achieving the purpose of equivalent filtering.

[0091] In summary, the present invention provides a spectrum shaping method based on probability shaping, which combines probability shaping with a feedback filter in the signal generation stage and dynamically adjusts the output probability of probability shaping, so that the generated signal frequency has the effect of a narrow-band spectrum depression before filtering, while not affecting the output level of the signal. The cost is a slight decrease in spectrum efficiency in the shaping stage (depending on the coding parameters, it may be less than 5%). This method is applicable to a variety of modulation formats such as PAM, QAM, PSK under direct modulation and coherent modulation, and is compatible with the static probability shaping technology commonly used in coherent systems. Through the above steps, the present invention realizes the probability screening of the output coding sequence, achieves the purpose of equivalent filtering, and avoids the loss of signal quality. Typical shaping effects are as follows: Figure 2 shown.

[0092] Embodiment 2:

[0093] This embodiment provides a probability shaping single-carrier communication method for implementing clock simultaneous transmission, including the following steps:

[0094] Transmitter:

[0095] A11. The binary information executes the spectrum shaping method based on probability shaping described in Example 1 to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and satisfying a preset probability distribution.

[0096] A12. The signal sequence is resampled, spectrum shaped, and clock inserted. The sampling rate is adjusted from the baud rate to the target sampling rate in the digital domain. At the same time, a digital clock signal is inserted at the suppression frequency during the resampling process, and the spectrum is shaped, such as raised cosine, root raised cosine, etc.

[0097] In step A12, in order to ensure the quality of the transmitted signal and the receiving performance, the analog signal spectrum is usually slightly larger than 1 / 2 baud rate, so it is necessary to upsample the baud rate signal sequence and perform a raised cosine / root raised cosine roll-off filter at the frequency edge, which is therefore called (conventional) spectrum shaping. Since the signal sequence has been processed by the probability shaping module and has been suppressed at a specific frequency, it is necessary to insert a clock there, and this step is implemented before or after resampling depending on the situation. The reason for the difference between inserting the clock before or after sampling is that the main purpose of inserting the clock is to facilitate synchronous sampling, but the process of determining synchronization inside the receiver can be implemented before or after resampling. Therefore, the module can be selectively combined, such as the traditional structure of resampling + spectrum shaping. For a single-carrier system with baud rate sampling, in order to ensure the demodulation performance, a clock is required to assist, and only a clock insertion module can be included (at this time, the optimal clock insertion frequency is ±1 / 4 baud rate).

[0098] A13. The resampled signal sequence undergoes digital-to-analog conversion to output an analog electrical signal.

[0099] A14. The modulator uses the laser as a carrier, modulates the electrical signal into the spectrum, and outputs a single carrier signal in the optical domain.

[0100] Receiver:

[0101] A21. The optical domain single carrier signal passes through a mixer and is downconverted to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiving end.

[0102] A22. The analog electrical signal is sampled into a digital electrical signal by an analog-to-digital converter.

[0103] A23. The digital electrical signal is resampled and the clock is extracted, usually by narrowband filtering, such as static filter, digital phase-locked loop and other structures, to separate the clock from the digital electrical signal and analyze the clock error information, including frequency and phase errors.

[0104] A24. Perform sampling clock compensation on digital electrical signals; correct sampling delays through clock recovery algorithms, or feed back the phase error of the clock to the sampling clock to lock the sampling clock in hardware.

[0105] A25. The digital electrical signal passes through an equalization module, such as CMA, LMS and other algorithms, to compensate for the filtering effects of the channel and the device.

[0106] A26. The digital electrical signal is used to compensate for the frequency deviation caused by the frequency difference between the transmitting and receiving lasers. The frequency error information of the clock is compensated by an algorithm, or the frequency error is fed back to the laser with frequency control to lock the laser frequency deviation in hardware.

[0107] A27. The digital electrical signal is processed through carrier phase recovery to compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers.

[0108] A28. The digital electrical signal passes through the probability shaping decoding module to restore the binary information of the transmitter.

[0109] Traditional receiver structures generally perform double baud rate sampling at the resampling point to ensure that information is not lost due to the frequency deviation of the laser. However, for a baud rate single carrier system, since the frequency deviation can be compensated by hardware before sampling, baud rate downsampling can be performed here, and even baud rate sampling can be performed directly at the analog-to-digital conversion point without any resampling steps. This is also the advantage of the present invention in reducing complexity at the receiving end.

[0110] Embodiment 3:

[0111] This embodiment provides a probability shaping single-carrier communication system for implementing clock simultaneous transmission, including:

[0112] like Figure 5 As shown, the transmitter includes:

[0113] Probability shaping module: used to execute the spectrum shaping method based on probability shaping described in Example 1 on binary information to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and satisfying a preset probability distribution.

[0114] Resampling spectrum shaping and clock insertion module: used to resample, spectrum shape and clock insert the signal sequence, adjust the sampling rate from the baud rate to the target sampling rate in the digital domain, insert a digital clock signal at the suppression frequency during the resampling process, and then shape the spectrum.

[0115] In order to ensure the quality of the transmitted signal and the receiving performance, the spectrum of the analog signal is usually slightly larger than 1 / 2 of the baud rate. Therefore, it is necessary to upsample the baud rate signal sequence and perform a raised cosine / root raised cosine roll-off filter at the frequency edge. The latter is therefore called (conventional) spectrum shaping. Since the signal sequence has been processed by the probability shaping module and has been suppressed at a specific frequency, it is necessary to insert a clock there. This step is implemented before or after resampling depending on the situation. The reason for the difference between inserting the clock before or after sampling is that the main purpose of inserting the clock is to facilitate synchronous sampling, but the process of determining synchronization inside the receiver can be implemented either before or after resampling. Therefore, this module can be selectively combined, such as the traditional structure of resampling + spectrum shaping. For a single-carrier system with baud rate sampling, a clock is required to assist in order to ensure demodulation performance, and only a clock insertion module can be included (the optimal clock insertion frequency is ±1 / 4 of the baud rate at this time).

[0116] Digital-to-analog conversion module: used to convert the resampled signal sequence into digital-to-analog and output analog electrical signals.

[0117] Modulator module: used to use the laser as a carrier, modulate the electrical signal into the spectrum, and output a single carrier signal in the optical domain.

[0118] like Figure 6 As shown, the receiving end includes:

[0119] Mixer: It is used to down-convert the optical domain single carrier signal output by the transmitter to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiver.

[0120] Analog-to-digital conversion module: used to sample analog electrical signals into digital electrical signals through an analog-to-digital converter.

[0121] Resampling and clock extraction module: used to resample digital electrical signals and extract clocks, usually narrowband filtering, such as static filters, digital phase-locked loops and other structures, to separate the clock from the digital electrical signal and analyze the clock error information; including frequency and phase errors.

[0122] Clock recovery module: used to perform sampling clock compensation on the digital electrical signal after being processed by the resampling and clock extraction modules; this step can correct the sampling delay through the clock recovery algorithm module, and can also feed back the phase error of the clock to the sampling clock to lock the sampling clock at the hardware level.

[0123] Equalization module: used to perform equalization processing on the digital electrical signal processed by the clock recovery module, such as CMA, LMS and other algorithms, to compensate for the filtering effects of channels and devices.

[0124] Frequency deviation compensation module: It is used to compensate the frequency deviation caused by the frequency difference between the transmitting and receiving lasers of the digital electrical signal processed by the equalization module. This step can be compensated through the algorithm using the frequency error information of the clock, or the frequency error can be fed back to the laser with frequency control to lock the laser frequency deviation at the hardware level.

[0125] Carrier phase recovery module: used to perform carrier phase recovery on the digital electrical signal processed by the frequency offset compensation module to compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers.

[0126] Probability shaping decoding module: used to restore the binary information of the transmitting end through the probability shaping decoding module of the digital electrical signal processed by the carrier phase recovery module.

[0127] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0128] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A spectrum shaping method based on probability shaping, It is characterized in that The following steps are involved: Encode the input binary information bit by bit: For a sequence of length N, bit-by-bit encoding requires N cycles. During the encoding process, V is always guaranteed to be l ≤U l , U h ≤V h , where, for an equiprobable binary distribution, a binary sequence of length N is replaced by a unique segment of length 2 in a line segment of length 1. -N The line segment U represents that the lower end point of the line segment U is defined as U l , the top dead center is U h ; For binary distributions with unequal probabilities, each sequence is represented by a unique line segment V in a line segment of length 1. The length of the line segment V is proportional to the probability of the sequence appearing. The lower end point of the line segment V is defined as V l , top dead center is V h , the next split position is V m ; The bit-by-bit encoding steps include: S11. Update U for the kth bit of the current input l and U h , the length of U becomes half; S12. For the current V l and V h , calculate the segmentation position V m , check whether V m ≤U l or U h ≤V m , if satisfied, the output code will be 0 or 1 accordingly, V l and V h Update accordingly. If V m ≤U l , then let V l =V m ;if U h ≤V m , then let V h =V m ; S13. According to the current past output sequence S out , adjust V l and V h Each lower decision point V within the range 1，2... , so that the probability of each symbol appearing at the lower level increases with C k+1，n Adjust the size of C k+1，n Indicates that for each preset frequency point f n Calculate filtering results using feedback filters; S14. Check [V l , V h ] Is it in 2 -m-1 Length is step, length is 2 -m If so, U and V are uniformly scaled by remapping the corresponding line segment to [0, 1]; Finishing process: S21. After the input signal ends, calculate the segmentation position V m , check whether U is satisfied l ≤V m ≤(U l +U h ) / 2 or (U l +U h ) / 2≤V m ≤U h , output 0 or 1 accordingly and update V, repeat the above steps S11 to S14, so that the final U l ≤V l ≤(U l +U h ) / 2≤V h ≤U h .

2. The spectrum shaping method based on probability shaping according to claim 1, It is characterized in that The step S13 comprises: S131. For each preset frequency point f n , -1 / 2≤f n ≤1 / 2, use the feedback filter with exp decay to calculate the filtering result: In the formula, subscript k indicates the cycle in which the kth bit is output, subscript n indicates the frequency point number, and a n represents f n The attenuation coefficient corresponding to the frequency; S132. Update the lower decision point V 1，2… , define P as V 1，2… The corresponding default probability distribution vector, b n is the frequency point f n The corresponding feedback modulation depth vector makes the probability of each symbol appearing at the lower level increase with C k+1，n To adjust the size: V 1，2... =V l +(V h -V l )·(P+∑ n b n ·C k+1，n )。 3. The spectrum shaping method based on probability shaping according to claim 2, It is characterized in that When u is satisfied l ≤V l ≤(U l +U h ) / 2≤V h ≤U h , then line segment V can uniquely determine line segment U, that is, the codes of unequal probability corresponding to V are the encoding results of the input sequence.

4. The spectrum shaping method based on probability shaping according to claim 2, It is characterized in that In the bit-by-bit encoding process, steps S11 to S14 are looped multiple times, and in each loop, steps S12, S13, and S14 are executed once or more than once.

5. A probability shaping single-carrier communication method for clock simultaneous transmission, It is characterized in that The following steps are involved: Transmitter: A11. Binary information performs the spectrum shaping method based on probability shaping according to any one of claims 1 to 4 to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and satisfying a preset probability distribution; A12. The signal sequence is resampled, spectrum shaped, and clock inserted. The sampling rate is adjusted from the baud rate to the target sampling rate in the digital domain. A digital clock signal is inserted at the suppression frequency during the resampling process, and the spectrum is shaped. A13. The resampled signal sequence is converted from digital to analog to output an analog electrical signal; A14. The modulator uses the laser as a carrier, modulates the electrical signal into the spectrum, and outputs a single carrier signal in the optical domain; Receiver: A21. The optical domain single carrier signal passes through the mixer and is downconverted to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiving end. A22. The analog electrical signal is sampled into a digital electrical signal by an analog-to-digital converter; A23. The digital electrical signal is resampled and clock extracted to separate the clock from the digital electrical signal and analyze the clock error information; A24. Sampling clock compensation for digital electrical signals; A25. The digital electrical signal passes through the equalization module to compensate for the filtering effect of the channel and the device; A26. Digital electrical signals are used to compensate for the frequency deviation caused by the frequency difference between the transmitting and receiving lasers; A27. The digital electrical signal is processed by carrier phase recovery to compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers; A28. The digital electrical signal passes through the probability shaping decoding module to restore the binary information of the transmitter.

6. The probability shaping single-carrier communication method for realizing clock synchronization according to claim 5, It is characterized in that In step A12, in order to ensure the quality of the transmitted signal and the receiving performance, the analog signal spectrum is greater than 1 / 2 the baud rate, the baud rate signal sequence is upsampled, and a raised cosine / root raised cosine roll-off filter is performed at the frequency edge.

7. The probability shaping single-carrier communication method for realizing clock synchronization according to claim 5, It is characterized in that In step A24, the sampling delay is corrected by a clock recovery algorithm, or the phase error of the clock is fed back to the sampling clock, and the sampling clock is locked in hardware.

8. The probability shaping single-carrier communication method for realizing clock synchronization according to claim 5, It is characterized in that In step A25, compensation is performed using the frequency error information of the clock, or the frequency error is fed back to a laser with frequency control, and the laser frequency deviation is locked in hardware.

9. A probability shaping single-carrier communication system for clock simultaneous transmission. It is characterized in that include: The transmitter includes: Probabilistic shaping module: used to perform the spectrum shaping method based on probability shaping according to any one of claims 1 to 4 on binary information to form a PAM / PSK / QAM real / complex signal sequence with specific frequency suppression and satisfying a preset probability distribution; Resampling spectrum shaping and clock insertion module: used to resample, spectrum shape and clock insert the signal sequence, adjust the sampling rate from the baud rate to the target sampling rate in the digital domain, insert a digital clock signal at the suppression frequency during the resampling process, and then shape the spectrum. Digital-to-analog conversion module: used to convert the resampled signal sequence into digital-to-analog signals and output analog electrical signals; Modulator module: used to use the laser as a carrier, modulate the electrical signal into the spectrum, and output a single carrier signal in the optical domain; The receiving end includes: Mixer: used to down-convert the optical domain single carrier signal output by the transmitter to a baseband analog electrical signal under the equivalent multiplication effect of the laser at the receiver; Analog-to-digital conversion module: used to sample analog electrical signals into digital electrical signals through an analog-to-digital converter; Resampling and clock extraction module: used to resample digital electrical signals and extract clocks, separate clocks from digital electrical signals, and analyze clock error information; Clock recovery module: used to perform sampling clock compensation on the digital electrical signal processed by the resampling and clock extraction modules; Equalization module: used to equalize the digital electrical signal processed by the clock recovery module and compensate for the filtering effect of the channel and the device; Frequency deviation compensation module: used to compensate the frequency deviation caused by the frequency difference between the transmitting and receiving lasers of the digital electrical signal processed by the equalization module; Carrier phase recovery module: used to perform carrier phase recovery processing on the digital electrical signal processed by the frequency offset compensation module, and compensate for the complex plane rotation of the demodulated baseband signal caused by the phase difference between the transmitting and receiving lasers; Probability shaping decoding module: used to restore the binary information of the transmitting end through the probability shaping decoding module of the digital electrical signal processed by the carrier phase recovery module.

10. The probability shaping single-carrier communication system for realizing clock synchronization according to claim 9, It is characterized in that In the resampled spectrum shaping and clock insertion module, in order to ensure the quality of the transmitted signal and the receiving performance, the analog signal spectrum is greater than 1 / 2 the baud rate, the baud rate signal sequence is upsampled, and the raised cosine / root raised cosine roll-off filtering is performed at the frequency edge.

Images