A method for sending and receiving OTFS signals

Abstract

The present invention discloses an OTFS signal sending and receiving method, which belongs to the field of wireless communication technology. At the transmitting end, the present invention converts the delay-Doppler domain signal into an OTFS time domain signal; uses a phase predistortion sequence to scramble the delay-Doppler domain signal; obtains the signal with the minimum in-band distortion through an iterative method, changes the position of the pilot symbol, and transmits the OTFS time domain signal. At the receiving end, the time domain received signal is converted into a delay-Doppler domain signal; the phase distortion information of the pilot symbol is estimated based on the energy distribution of the pilot area; pilot-assisted channel estimation is performed, and phase correction is performed on the channel estimation result; message transmission detection is performed on the delay-Doppler domain signal, the pilot area is restored to the default position, and the constellation mapping is solved. The present invention can effectively reduce the peak-to-average power ratio of the OTFS signal, can more accurately recover the transmitted data, and improve the transmission efficiency of the communication.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to an OTFS signal sending and receiving method. Background Art

[0002] In high-speed mobile scenarios, the high latency and high Doppler shift characteristics of wireless channels pose significant challenges to wireless communication technologies. Orthogonal frequency division multiplexing (OFDM) technology experiences a sharp decline in system performance when affected by high Doppler shift. In recent years, orthogonal time-frequency space (OTFS) modulation has been shown to offer superior performance over OFDM in high-speed mobile channels.

[0003] Unlike OFDM, which processes signals in the time-frequency domain, OTFS transforms signals into the delay-Doppler domain for processing, which can obtain delay diversity and Doppler diversity, giving OTFS superior performance in time-varying channels.

[0004] Currently, there are three types of OTFS peak-to-average ratio reduction solutions:

[0005] The first is to perform precoding on the signal matrix. The precoding process requires the construction of a precoding matrix, but matrix multiplication usually leads to high complexity.

[0006] The second approach is the Partial Transmit Sequence (PTS) and Selective Mapping (SLM) scheme. This scheme requires an exhaustive search for the optimal phase factor or phase sequence, resulting in high complexity. Furthermore, the transmission of side information related to the phase factor or phase sequence requires additional resources or bandwidth.

[0007] The third method is to limit the amplitude of the time-domain signal. This method reduces the peak-to-average ratio by changing the amplitude of the time-domain signal, but the nonlinear distortion it brings will seriously affect the system performance. Summary of the Invention

[0008] In view of this, the present invention proposes an OTFS signal sending and receiving method, which solves the problems of high time complexity, additional resource occupation and severe nonlinear distortion in current peak-to-average ratio reduction schemes.

[0009] The technical solution adopted in the present invention is:

[0010] An OTFS signal sending method, applied to a signal sending end, comprises the following steps:

[0011] Step T1: In the delay-Doppler domain, the constellation-mapped symbols, pilot symbols, and guard intervals are mapped to a delay grid number of , the Doppler grid number is The two-dimensional resource grid is converted into OTFS time domain signal through inverse symplectic Fourier transform and Heisenberg transform;

[0012] Step T2, calculating the peak-to-average ratio of the latest OTFS time-domain signal. If the peak-to-average ratio exceeds the peak-to-average ratio threshold, executing step T3; otherwise, executing step T7;

[0013] Step T3, generating multiple phase predistortion sequences;

[0014] Step T4: Using each phase predistortion sequence, scramble the delay-Doppler domain signal corresponding to the latest OTFS time domain signal; convert the scrambled delay-Doppler domain signal with the minimum distortion into the OTFS time domain signal, and update the latest OTFS time domain signal;

[0015] Step T5: Determine whether the maximum amplitude of the most recently obtained OTFS time domain signal exceeds the amplitude threshold. If so, convert all the scrambled delay-Doppler domain signals in step T4 into OTFS time domain signals, and select the OTFS time domain signal with the smallest amplitude at the index corresponding to the scrambled position as the most recently obtained OTFS time domain signal, and execute step T7. Otherwise, execute step T6.

[0016] Step T6, determine whether the maximum number of iterations has been reached. If so, proceed to step T7, otherwise return to step T2;

[0017] Step T7: Determine whether the pilot symbol of the most recently obtained OTFS time-domain signal is predistorted. If so, transform the position of the pilot symbol in the delay-Doppler domain by two-dimensional cyclic shift, and then convert the processed delay-Doppler domain signal into an OTFS time-domain signal. Otherwise, directly execute step T8.

[0018] Step T8: transmitting the final OTFS time domain signal.

[0019] Furthermore, the pilot symbols are located within a pre-set pilot area, the distance between the pilot areas on the delay axis is not less than the maximum delay spread range, and the distance between the pilot areas on the Doppler axis is not less than 2 times the maximum Doppler frequency shift spread range; the phase distortion information of the pilot area and the pilot symbol is in a one-to-one correspondence. If the pilot symbol is not distorted, its position remains unchanged.

[0020] Furthermore, the specific method of step T3 is:

[0021] Step T301: Generate a binary pseudo-random sequence with pseudo-random properties , the sequence is an m-sequence, an M-sequence or a gold sequence;

[0022] Step T302: Obtain the maximum allowable rotation angle of the constellation point based on the constellation mapping method and its correct judgment area. ; Select multiple less than Angle θ;

[0023] Step T303, according to each binary pseudo-random sequence For each combination of angles θ, calculate the corresponding phase predistortion sequence z:

[0024] ,

[0025] Where j is the imaginary unit.

[0026] Furthermore, in step T4, the delay-Doppler domain signal corresponding to the latest OTFS time domain signal is scrambled in the following manner:

[0027] Step T401: for the latest OTFS time domain signal, find the index corresponding to its maximum amplitude;

[0028] Step T402: Based on the index obtained in step T401, a one-dimensional vector extending along the Doppler axis with the same delay index is found in the delay-Doppler domain signal corresponding to the most recently obtained OTFS time-domain signal. This one-dimensional vector is the delay-Doppler domain vector that causes the peak to appear.

[0029] In step T403, a Hadamard product operation is performed on the phase predistortion sequence and the delay-Doppler domain vector causing the peak to appear to achieve scrambling. The processed delay-Doppler domain signal is the scrambled delay-Doppler domain signal.

[0030] Furthermore, the amplitude threshold in step T5 is:

[0031] ,

[0032] in, is the amplitude threshold, represents the average power of the OTFS signal, is the peak-to-average ratio threshold.

[0033] An OTFS signal receiving method, applied to a signal receiving end, for receiving a signal sent by the OTFS signal sending method, comprising the following steps:

[0034] Step R1, converting the time domain received signal into a delay-Doppler domain signal through Wigner transform and sigmoid Fourier transform;

[0035] Step R2, estimating the phase distortion information of the pilot symbol based on the energy distribution of the delay-Doppler domain signal in each pilot area;

[0036] Step R3, performing pilot-assisted channel estimation using the pilot symbols, and then performing phase correction on the channel estimation result based on the phase distortion information of the pilot symbols;

[0037] Step R4: perform message passing detection on the delay-Doppler domain signal based on the phase-corrected channel information, then perform a two-dimensional cyclic shift on the detected signal matrix, restore the pilot area to the default position, and de-constellation map.

[0038] Furthermore, the specific method of step R2 is:

[0039] Step R201, calculating the variance of the amplitude of the delay-Doppler domain signal in each pilot area, and selecting the area with the largest variance as the estimated pilot area;

[0040] Step R202: Estimate the phase distortion information of the pilot symbol based on the one-to-one correspondence between the pilot region and the phase distortion information of the pilot symbol.

[0041] Furthermore, in step R3, the specific method of performing phase correction on the channel estimation result is:

[0042] ,

[0043] in, Indicates the first channel estimation result based on pilot assistance channel coefficients, is the estimated phase of the pilot rotation caused by the peak-to-average ratio reduction, Indicates the first channel coefficients.

[0044] The beneficial effects of the present invention are:

[0045] 1. The signal transmission method of the present invention proposes a Peak-to-Average Power Ratio (PAPR) reduction algorithm. Through iterative execution, the PAPR is gradually brought closer to the system's preset threshold while minimizing the introduced in-band distortion. This method addresses the problems of current PAPR reduction schemes, such as high time complexity, the need for additional frequency resources, and the resulting severe nonlinear distortion.

[0046] 2. The signal receiving method of the present invention can further eliminate the signal distortion caused by the PAPR reduction at the transmitting end by eliminating the signal coefficient distortion, and restore more accurate transmission data.

[0047] 3. The present invention improves the transmission efficiency of communication, is easy to implement, and has strong practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 Peak-to-average ratio performance curve based on the present invention under QPSK modulation mode provided for simulation experiments;

[0049] Figure 2 Peak-to-average ratio performance curve based on the present invention under 8-APSK modulation mode provided for simulation experiments;

[0050] Figure 3 Peak-to-average ratio performance comparison curves based on exhaustive search algorithm and random forest algorithm provided for simulation experiments;

[0051] Figure 4 The bit error rate curve based on the present invention and the pilot interference elimination algorithm under the QPSK modulation mode provided for the simulation experiment;

[0052] Figure 5 The bit error rate curve based on the present invention and the pilot interference elimination algorithm under the 8-APSK modulation mode provided for the simulation experiment;

[0053] Figure 6 The iteration number distribution curve of the present invention is provided for simulation experiments. DETAILED DESCRIPTION

[0054] The present invention is further described in detail below with reference to the accompanying drawings.

[0055] An OTFS signal sending method, applied to a signal sending end, comprises the following steps:

[0056] Step T1: In the delay-Doppler domain, the constellation-mapped symbols, pilot symbols, and guard intervals are mapped to a delay grid number of , the Doppler grid number is The two-dimensional resource grid is converted into OTFS time domain signal through inverse symplectic Fourier transform and Heisenberg transform;

[0057] The delay-Doppler domain signal includes a data region and a pilot region. The pilot region includes an embedded pilot and a guard interval region with surrounding values ​​of 0.

[0058] The pilot symbols are located within pre-set pilot areas. The distance between pilot areas on the delay axis is no less than the maximum delay spread range, and the distance between pilot areas on the Doppler axis is no less than twice the maximum Doppler frequency shift spread range. The phase distortion information of the pilot areas and pilot symbols is in a one-to-one correspondence. If the pilot symbol is not distorted, its position remains unchanged and is the default position.

[0059] The receiving end can obtain the pilot signal distortion information by detecting the pilot signal position, and then recover the pilot signal distortion.

[0060] Step T2, calculating the peak-to-average ratio of the latest OTFS time-domain signal. If the peak-to-average ratio exceeds the peak-to-average ratio threshold, executing step T3; otherwise, executing step T7;

[0061] Step T3: Generate multiple phase predistortion sequences; the specific method is:

[0062] Step T301: Generate a binary pseudo-random sequence with values ​​of 0 and 1. , the sequence can be an m-sequence, an M-sequence or a gold sequence;

[0063] Step T302: Obtain the maximum allowable rotation angle of the constellation point based on the constellation mapping method and its correct judgment area. , at the maximum permissible rotation angle In the range, select multiple less than Angle θ;

[0064] Step T303, according to each binary pseudo-random sequence For each combination of angles θ, calculate the corresponding phase predistortion sequence z:

[0065] ,

[0066] Where j is the imaginary unit.

[0067] The phase predistortion sequence used in this step has the following characteristics:

[0068] The phase predistortion sequence is a one-dimensional phase sequence extended along the Doppler domain with an amplitude of 1. The average power of the signal remains unchanged after scrambling.

[0069] It is based on a pseudo-random sequence and has the properties of a pseudo-random sequence;

[0070] The resulting predistortion manifests itself as random deflection of the signal constellation points to the left and right, with an adjustable deflection angle.

[0071] Step T4: Using each phase predistortion sequence, scramble the delay-Doppler domain signal corresponding to the most recently obtained OTFS time domain signal; convert the scrambled delay-Doppler domain signal with the minimum distortion into an OTFS time domain signal, and update the most recently obtained OTFS time domain signal; wherein, the delay-Doppler domain signal corresponding to the most recently obtained OTFS time domain signal is scrambled in the following manner:

[0072] Step T401: for the latest OTFS time domain signal, find the index corresponding to its maximum amplitude;

[0073] Step T402: Based on the index obtained in step T401, a one-dimensional vector extending along the Doppler axis with the same delay index is found in the delay-Doppler domain signal corresponding to the most recently obtained OTFS time-domain signal. This one-dimensional vector is the delay-Doppler domain vector that causes the peak to appear.

[0074] In the In the iterations, the time domain signal after i-1 iterations The maximum amplitude index of and the corresponding Doppler vector delay index The relationship can be expressed as:

[0075] ,

[0076] in, Express Take the remainder.

[0077] In step T403, a Hadamard product operation is performed on the phase predistortion sequence and the delay-Doppler domain vector causing the peak to appear to achieve scrambling. The processed delay-Doppler domain signal is the scrambled delay-Doppler domain signal.

[0078] In this process, the optimal phase predistortion sequence needs to be obtained through exhaustive search or pre-trained random forest. This method is a conventional technical means in this field and will not be described here.

[0079] Step T5: determine whether the maximum amplitude of the latest OTFS time domain signal exceeds the amplitude threshold. If it exceeds, all the scrambled delay-Doppler domain signals in step T4 are converted into OTFS time domain signals, and the OTFS time domain signal with the smallest amplitude at the index corresponding to the scrambled position is selected as the latest OTFS time domain signal, and step T7 is executed; otherwise, step T6 is executed;

[0080] Among them, the amplitude threshold for:

[0081] ,

[0082] Where, represents the average power of the OTFS signal, is the peak-to-average ratio threshold.

[0083] Through steps T4 and T5, the introduced in-band distortion can be minimized while meeting the peak-to-average ratio requirement. The in-band distortion is described by the error amplitude vector of the signal before and after predistortion.

[0084] Step T6, determine whether the maximum number of iterations has been reached. If so, proceed to step T7, otherwise return to step T2;

[0085] Step T7: Determine whether the pilot symbol of the most recently obtained OTFS time-domain signal is predistorted. If so, transform the position of the pilot symbol in the delay-Doppler domain by two-dimensional cyclic shift to convey its distortion information. Then, convert the processed delay-Doppler domain signal into an OTFS time-domain signal. Otherwise, directly execute step T8.

[0086] Step T8: transmitting the final OTFS time domain signal.

[0087] This method adopts an innovative PAPR reduction algorithm to make the PAPR iteratively approach the system preset threshold while minimizing the introduced in-band distortion. It can effectively reduce the PAPR of the OTFS signal, solve the problems of high time complexity, the need to occupy additional frequency band resources, and severe nonlinear distortion in the current PAPR reduction scheme, and improve the signal transmission efficiency.

[0088] An OTFS signal receiving method, applied to a signal receiving end, for receiving a signal sent by the above-mentioned OTFS signal sending method, comprises the following steps:

[0089] Step R1, converting the time domain received signal into a delay-Doppler domain signal through Wigner transform and sigmoid Fourier transform;

[0090] Step R2: Estimate the phase distortion information of the pilot symbol based on the energy distribution of the delay-Doppler domain signal in each pilot area. The specific method is:

[0091] Step R201, calculating the variance of the amplitude of the delay-Doppler domain signal in each pilot area, and selecting the area with the largest variance as the estimated pilot area;

[0092] Step R202: Estimate the phase distortion information of the pilot symbol based on the one-to-one correspondence between the pilot region and the phase distortion information of the pilot symbol.

[0093] For the Preset pilot area , , the variance of its signal amplitude Expressed as:

[0094] ,

[0095] in, Indicates the maximum delay extension range, Indicates the maximum Doppler shift extension range, Indicates the The Doppler shift index in the pilot area, Indicates the The delay index in the pilot area, , represents the mathematical expectation, Represents the received signal in the time-Doppler domain.

[0096] Step R3: perform pilot-assisted channel estimation using the pilot symbols, and then perform phase correction on the channel estimation result based on the phase distortion information of the pilot symbols:

[0097] ,

[0098] in, Indicates the first channel estimation result based on pilot assistance channel coefficients, is the estimated phase of the pilot rotation caused by the peak-to-average ratio reduction, Indicates the first channel coefficients.

[0099] Step R4: perform message passing detection on the delay-Doppler domain signal based on the corrected channel information, then perform a two-dimensional cyclic shift on the detected signal matrix, restore the pilot area to the default position, and de-constellation map.

[0100] The following example illustrates the specific iterative process at the sending end:

[0101] use represents the OTFS time domain signal, represents the OTFS delay-Doppler domain signal, Indicates the system preset peak-to-average ratio threshold (in dB). represents the maximum number of iterations, represents the average power of the signal, Indicates the amplitude threshold related to the preset peak-to-average ratio threshold, represents the phase predistortion sequence.

[0102] In the In the iteration, first calculate the time domain signal after i-1 iterations If the threshold is not exceeded , the signal can be sent directly; otherwise, it enters the iterative process, which is as follows:

[0103] S1, find the time domain signal The maximum amplitude and the corresponding index , and its corresponding Doppler vector in the delay-Doppler domain signal. The delay index of the Doppler vector is It can be obtained by the following formula:

[0104] ,

[0105] in, It means taking the remainder of M.

[0106] The searched Doppler vector can be expressed as ,in Indicates that it has passed The delay-Doppler domain signal of the iteration.

[0107] S2, the Doppler vector found With the appropriate phase predistortion sequence Multiply so that the time domain index found The signal amplitude at Not greater than an amplitude threshold related to the peak-to-average ratio threshold The process can be expressed as follows:

[0108] ,

[0109] in, represents the Hadamard product.

[0110] S3, if no suitable phase predistortion sequence is found, select The minimum phase predistortion sequence is selected as the optimal phase predistortion sequence. If there are multiple suitable phase predistortion sequences, the predistortion sequence that minimizes the in-band distortion of the introduced signal is selected as the optimal phase predistortion sequence. In-band distortion can be described by the error amplitude vector (EVM):

[0111] ,

[0112] in, express norm. At this point, the optimization problem of the predistorted sequence can be expressed as:

[0113] ,

[0114] Iterate the above process until the signal peak-to-average ratio does not exceed the preset peak-to-average ratio threshold, or the maximum number of iterations is reached, or all phase predistortion sequences are tried and still not satisfied. .

[0115] In the above process, the phase pseudo-random sequence can be expressed as:

[0116] ,

[0117] in, represents a binary pseudo-random sequence with values ​​of 0 and 1, represents the deflection angle of the signal constellation point caused by the predistortion sequence, Indicates the maximum allowed deflection angle; The constructed phase predistortion sequence is a one-dimensional phase sequence extended along the Doppler domain, with an amplitude of 1 and a length equal to the number of Doppler domain grids. For all phase predistortion sequences The collection composed of.

[0118] The optimal phase predistortion sequence can be obtained through exhaustive search or through pre-trained random forest. The real and imaginary parts of can be regarded as features and used as input of random forest to obtain the optimal phase predistortion. exist The index in can be regarded as the classification result, which is the output of random forest. The training process of random forest is as follows:

[0119] (1) Use Bootstrap sampling to select a feature subset for each decision tree;

[0120] (2) When the nodes of each tree need to be split, some features are randomly selected from the feature subset;

[0121] (3) Use the CART (Classification And Regression Tree) algorithm to train the decision tree;

[0122] (4) Repeat the above process until all decision trees are constructed.

[0123] The advantage of using pre-trained random forests is that pre-training can be completed offline, which can avoid the exhaustive search process for the optimal phase pre-distortion sequence during the peak-to-average ratio reduction process and significantly reduce the time complexity.

[0124] In the subsequent processing of the transmitting end, for the pilot signal with phase distortion, the position of the pilot signal region can be changed by cyclic shift to transmit the distortion information of the pilot signal, which means that the phase distortion of the pilot signal and the pilot signal position have a one-to-one correspondence. Indicates the allowed pilot positions, then:

[0125] Indicates the allowed pilot areas. , Indicates the number of allowed pilot regions.

[0126] The one-to-one correspondence between the pilot phase distortion angle and the pilot position is:

[0127] ,

[0128] At the transmitting end, the pilot area is arranged in a default area :

[0129] .

[0130] in, represents the Doppler shift index in the default pilot area, represents the delay index in the default pilot area, .

[0131] If the phase distortion of the pilot signal after the peak-to-average ratio is reduced, , then the pilot area is shifted two-dimensionally to the area corresponding to the mobile terminal and the pilot distortion. This process can be expressed as follows:

[0132] ,

[0133] in, represents the delay-Doppler domain signal after reducing the peak-to-average ratio, represents the delay-Doppler domain signal after two-dimensional cyclic shift. Indicates that the matrix is ​​circularly shifted row by row Step, circular shift by column step.

[0134] At the receiving end, the location of the pilot signal is detected by calculating the energy distribution characteristics of all possible pilot areas in the received signal, and then the distortion information of the pilot signal is obtained, which is used to restore the pilot signal. Since the energy distribution of the data area and the pilot area is different, the following formula is used The energy characteristics represented as the Detection indicators of allowed pilot areas:

[0135] ,

[0136] in, represents the received signal in the time-Doppler domain, represents the mathematical expectation.

[0137] It should be noted that the received signal is a received transmit signal transmitted through a channel, and the transmit signal is a signal that has undergone peak-to-average ratio reduction and cyclic shift.

[0138] At this time, the detected pilot area It can be expressed as follows:

[0139] ,

[0140] ,

[0141] Among them, arg max means making b i Take the maximum value of i, which is .

[0142] The estimated pilot phase distortion angle is .

[0143] According to the above algorithm, The following formula should be satisfied:

[0144] ,

[0145] in, Indicates rounding down.

[0146] After channel estimation using embedded pilots, the estimated channel coefficients need to be multiplied by , to eliminate the additional phase deflection caused by the pilot signal distortion:

[0147] ,

[0148] in, Indicates the first channel estimation result based on pilot assistance channel coefficients, is the estimated phase of the pilot rotation caused by the peak-to-average ratio reduction, Indicates the first channel coefficients.

[0149] Finally, the data after signal detection is restored to its original position through cyclic shift to obtain the delay-Doppler domain signal at the original position. This process can be expressed as follows:

[0150] ,

[0151] in, represents the delay-Doppler domain signal after signal detection, It is the delay-Doppler domain signal that is restored to its original position after cyclic shift. Indicates that the matrix is ​​circularly shifted row by row Step, circular shift by column step.

[0152] In order to better illustrate the beneficial effects of the present invention, the following simulation experiments were conducted:

[0153] set up , maximum number of iterations , number of channels , the channel coefficient conforms to the complex Gaussian distribution: , the pilot power is 30dB, the system preset peak-to-average ratio threshold Take 8.5dB and 9.5dB respectively, the maximum rotation angle allowed Take separately 、 and The modulation modes are QPSK and 8-APSK. Pseudo-random sequence Select all 6th-order m-sequences and append a 0 or 1 to the end of each sequence, with the polarity opposite to the value of the last element in the sequence. During random forest pre-training, set the number of classification trees to 100 and the minimum number of leaves to 5.

[0154] The performance of this method is compared with that of the modified iterative clipping filtering (m-ICF) scheme. In the m-ICF scheme, the filter coefficients in the pilot region are Set to 0.6, the filter coefficient of the data area Set to 0.8, the clipping rate Set it to 3.0 and the number of iterations to 5.

[0155] Simulation Experiment 1

[0156] See also Figure 1 and Figure 2 , it can be seen that for the same peak-to-average ratio threshold , the value of the complementary cumulative distribution function (CCDF) curve changes with For QPSK and 8-APSK modulation modes, when and When , this method can almost completely reduce the peak-to-average ratio of the OTFS signal to below the preset threshold.

[0157] For QPSK modulation, compared with the OTFS system without peak-to-average ratio reduction, At the level of The peak-to-average gain is about 3.6dB. The peak-to-average gain is about 2.6dB. At the level of The peak-to-average gain is about 1.9dB; The peak-to-average gain is about 0.9dB.

[0158] For 8-APSK modulation, compared with the OTFS system without peak-to-average ratio reduction, At the level of The peak-to-average gain is about 3.6dB. The peak-to-average gain is about 2.8dB. At the level of The peak-to-average gain is about 1.8dB; The peak-to-average gain is approximately 1.0dB.

[0159] Simulation Experiment 2

[0160] See also Figure 3 ,It can be seen that when the simulation parameters are the same, the exhaustive search algorithm and the random forest algorithm have almost the same performance.,This means that the high time complexity caused by the exhaustive search can be reduced,by offline training of the random forest.

[0161] Simulation Experiment 3

[0162] See also Figure 4 and Figure 5 It can be seen that when the pilot distortion is not eliminated, the system performance drops sharply; when the pilot distortion elimination scheme provided by the present invention is used, the loss of the system error performance is very small and is better than the system performance of the m-ICF scheme.

[0163] Simulation Experiment 4

[0164] See also Figure 6 , it can be seen that when When , the number of iterations is concentrated in the interval; when When , the number of iterations is concentrated in the interval. And the number of iterations increases as the system preset peak-to-average ratio threshold decreases.

[0165] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0166] Furthermore, those skilled in the art will appreciate that, in combination with the information disclosed herein, they may implement each specific technical feature in different ways. Without exceeding the scope of the claims of the present invention, these different implementations are all within the scope of protection of the present invention.

Claims

1. An OTFS signal sending method, applied to a signal sending end, characterized in that: The following steps are involved: Step T1: In the delay-Doppler domain, the constellation-mapped symbols, pilot symbols, and guard intervals are mapped to a delay grid number of , the Doppler grid number is The two-dimensional resource grid is converted into OTFS time domain signal through inverse symplectic Fourier transform and Heisenberg transform; Step T2, calculating the peak-to-average ratio of the latest OTFS time-domain signal. If the peak-to-average ratio exceeds the peak-to-average ratio threshold, executing step T3; otherwise, executing step T7; Step T3, generating multiple phase predistortion sequences; Step T4: Using each phase predistortion sequence, scramble the delay-Doppler domain signal corresponding to the latest OTFS time domain signal; convert the scrambled delay-Doppler domain signal with the minimum distortion into the OTFS time domain signal, and update the latest OTFS time domain signal; Step T5: Determine whether the maximum amplitude of the most recently obtained OTFS time domain signal exceeds the amplitude threshold. If so, convert all the scrambled delay-Doppler domain signals in step T4 into OTFS time domain signals, and select the OTFS time domain signal with the smallest amplitude at the index corresponding to the scrambled position as the most recently obtained OTFS time domain signal, and execute step T7. Otherwise, execute step T6; Step T6, determine whether the maximum number of iterations has been reached. If so, proceed to step T7, otherwise return to step T2; Step T7: Determine whether the pilot symbol of the most recently obtained OTFS time-domain signal is predistorted. If so, transform the position of the pilot symbol in the delay-Doppler domain by two-dimensional cyclic shift, and then convert the processed delay-Doppler domain signal into an OTFS time-domain signal. Otherwise, directly execute step T8. Step T8: transmitting the final OTFS time domain signal.

2. The OTFS signal sending method according to claim 1, wherein: The pilot symbols are located within the pre-set pilot area. The distance between the pilot areas on the delay axis is not less than the maximum delay spread range, and the distance between the pilot areas on the Doppler axis is not less than 2 times the maximum Doppler shift spread range. There is a one-to-one correspondence between the pilot region and the phase distortion information of the pilot symbol. If the pilot symbol is not distorted, its position remains unchanged.

3. The OTFS signal sending method according to claim 1, wherein: The specific method of step T3 is: Step T301: Generate a binary pseudo-random sequence with pseudo-random properties , the sequence is an m-sequence, an M-sequence or a gold sequence; Step T302: Obtain the maximum allowable rotation angle of the constellation point based on the constellation mapping method and its correct judgment area. ; Select multiple less than Angle θ; Step T303, according to each binary pseudo-random sequence For each combination of angles θ, calculate the corresponding phase predistortion sequence z: ， Where j is the imaginary unit.

4. The OTFS signal sending method according to claim 1, wherein: In step T4, the delay-Doppler domain signal corresponding to the latest OTFS time domain signal is scrambled in the following manner: Step T401: for the latest OTFS time domain signal, find the index corresponding to its maximum amplitude; Step T402: Based on the index obtained in step T401, a one-dimensional vector extending along the Doppler axis with the same delay index is found in the delay-Doppler domain signal corresponding to the most recently obtained OTFS time-domain signal. This one-dimensional vector is the delay-Doppler domain vector that causes the peak to appear. In step T403, a Hadamard product operation is performed on the phase predistortion sequence and the delay-Doppler domain vector causing the peak to appear to achieve scrambling. The processed delay-Doppler domain signal is the scrambled delay-Doppler domain signal.

5. The OTFS signal sending method according to claim 1, wherein: The amplitude threshold in step T5 is: ， in, is the amplitude threshold, represents the average power of the OTFS signal, is the peak-to-average ratio threshold.

6. An OTFS signal receiving method, applied to a signal receiving end, for receiving a signal sent by any OTFS signal sending method according to any one of claims 1 to 5, characterized in that: The following steps are involved: Step R1, converting the time domain received signal into a delay-Doppler domain signal through Wigner transform and sigmoid Fourier transform; Step R2, estimating the phase distortion information of the pilot symbol based on the energy distribution of the delay-Doppler domain signal in each pilot area; Step R3, performing pilot-assisted channel estimation using the pilot symbols, and then performing phase correction on the channel estimation result based on the phase distortion information of the pilot symbols; Step R4: perform message passing detection on the delay-Doppler domain signal based on the phase-corrected channel information, then perform a two-dimensional cyclic shift on the detected signal matrix, restore the pilot area to the default position, and de-constellation map.

7. The OTFS signal receiving method according to claim 6, wherein: The specific method of step R2 is: Step R201, calculating the variance of the amplitude of the delay-Doppler domain signal in each pilot area, and selecting the area with the largest variance as the estimated pilot area; Step R202: Estimate the phase distortion information of the pilot symbol based on the one-to-one correspondence between the pilot region and the phase distortion information of the pilot symbol.

8. The OTFS signal receiving method according to claim 6, wherein: In step R3, the specific method of performing phase correction on the channel estimation result is: ， in, Indicates the first channel estimation result based on pilot assistance channel coefficients, is the estimated phase of the pilot rotation caused by the peak-to-average ratio reduction, Indicates the first channel coefficients.

Images