OTFS signal sending and receiving method of pilot frequency with low additional peak-to-average ratio

Abstract

The invention discloses an OTFS signal sending and receiving method of a pilot frequency with a low additional peak-to-average ratio, and relates to the technical field of ground-air communication. The method comprises the following steps: a transmitting end maps a data signal and a pilot signal to different two-dimensional grids in a time delay-Doppler domain; through parameter transformation, transformation of a data signal and a pilot signal is converted into OTFS modulation under unified parameters, time domain superposition of the pilot signal and the data signal is converted into time delay-Doppler domain superposition, and a sending signal is obtained. A pilot signal is separated by a receiving end through parameter transformation; performing channel coarse estimation and data signal coarse detection by using the pilot signal; performing fine estimation on the number of channel paths by using the pilot signals after interference elimination; and finally, iteratively carrying out channel time delay, Doppler frequency shift, channel coefficient fine estimation and signal fine detection based on fine estimation channel information. The method has the characteristics of low peak-to-average ratio, low signal generation complexity and high channel estimation precision.

Description

Technical Field

[0001] The present invention relates to the field of ground-air communication technology, and in particular to a method for transmitting and receiving OTFS signals with low additional peak-to-average power ratio pilots, which can realize the design of low peak-to-average power ratio (PAPR) time-domain superimposed pilots for orthogonal time-frequency space (OTFS) modulation technology under fast-varying multipath channels and high-precision channel estimation. Background Art

[0002] In communication scenarios between high-speed and high-dynamic aerial targets such as unmanned aerial vehicles and low-earth orbit satellites and ground targets, due to the influence of Doppler frequency and frequency variation on the communication channel, it exhibits fast time-varying characteristics. On the other hand, due to ground-air low-elevation reflection and platform self-occlusion and other reasons, the multipath characteristics of the channel are significant. The fast time-varying and multipath propagation characteristics of the channel greatly affect the ground-air communication performance. Especially when the future ground-air communication frequency band and bandwidth continue to increase, the fast time-varying and multipath characteristics of the channel will become more significant, and the impact on communication performance will be further increased. Under the conditions of ground-air fast time-varying multipath channels, traditional modulation and demodulation technologies such as orthogonal frequency division multiplexing (OFDM) will face great challenges and even cannot be applied.

[0003] Orthogonal time-frequency space (OTFS) modulation technology has significant performance advantages in high-speed mobile scenarios. With its signal processing ability in the time-delay-Doppler domain, it greatly improves the communication reliability under fast-varying multipath channel conditions. OTFS mainly uses two-dimensional Fourier transform in the time-frequency domain to construct a new time-delay-Doppler domain, and remodulates the signals in the time-frequency domain of the OFDM system in the time-delay-Doppler domain, so as to counteract the destruction of orthogonality between OFDM subcarriers caused by high frequency bands and high-speed movement.

[0004] However, the performance of the OTFS system is closely related to the channel estimation performance. The most common existing OTFS channel estimation method is embedded pilot channel estimation, which places a guard interval between the pilot and the data to make the pilot unaffected by the data, thus achieving excellent performance. However, due to the influence of the guard interval, the spectrum utilization efficiency of the system is reduced. To improve the spectrum utilization efficiency of the system, an embedded pilot structure without a guard interval is proposed, and the performance is improved through iteration and interference cancellation between the data and the pilot. However, due to the mutual influence between the data and the pilot, its performance deteriorates severely compared with the embedded pilot.

[0005] In addition, a pilot method with all pilots superimposed in the time delay and Doppler domains has been proposed in the prior art. However, it assumes that the time delay and Doppler frequency shift remain unchanged in several frames, and it requires additional frames using embedded pilots to estimate the time delay and Doppler frequency shift. To further improve the performance, a structure for sparsifying pilots has also been proposed in the prior art, which improves the channel estimation performance by reducing the interference between pilots, but still requires using additional frames with embedded pilots to estimate the time delay and Doppler frequency shift. It should be noted that neither of the above two channel estimation methods considers the impact of pilots on the peak-to-average power ratio. Summary of the Invention

[0006] In view of this, the present invention proposes a method for transmitting and receiving OTFS signals with low additional peak-to-average power ratio pilots. This method can solve the problems of signal peak-to-average power ratio deterioration and large data interference in channel estimation in the existing OTFS signal superimposed pilot method, and provides a practical technical approach for the engineering implementation of OTFS and the improvement of communication performance.

[0007] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] An OTFS signal transmission method with low additional peak-to-average power ratio pilots, applied to a signal transmitter, includes the following steps:

[0009] Step T1, in the time delay-Doppler domain, map the data signal onto a two-dimensional grid with a time delay parameter of M and a Doppler parameter of N, and map the low additional peak-to-average power ratio pilot signal onto a two-dimensional grid with a time delay parameter of M1 and a Doppler parameter of N1 to form a pilot matrix; where MN = M1N1, M = kM1, and the pilot matrix is a matrix composed of multiple embedded pilots and guard intervals. The pilot signals are continuous along the time delay axis and have guard intervals on the Doppler axis;

[0010] Step T2, in the time delay-Doppler domain, through parameter transformation, convert the inverse symplectic Fourier transform and Heisenberg transform with different parameters for the data signal and the pilot signal respectively to OTFS modulation under unified parameters, and convert the time domain superposition of the pilot signal and the data signal to time delay-Doppler domain superposition to obtain the transmitted signal and transmit it.

[0011] Further, in the pilot matrix, the energy of each pilot is the same, and the length of the guard interval is 2 times the maximum Doppler frequency shift expansion.

[0012] Further, in step T2, the inverse symplectic Fourier transform and Heisenberg transform with different parameters for the data signal and the pilot signal respectively are converted to OTFS modulation under unified parameters through a parameter transformation matrix; the parameter transformation matrix is:

[0013]

[0014] Among them, denotes rounding downwards, (i) k denotes the remainder of i divided by k, P represents the row-column permutation interleaving matrix, (·) T denotes the transpose of a matrix; denotes a diagonal matrix with parameter as the main diagonal elements, i = 0, 1,..., MN - 1; I N denotes the N-order identity matrix, F k denotes the one-dimensional k-order DFT matrix, (·) H denotes the conjugate transpose of a matrix, denotes the Kronecker product.

[0015] A method for receiving OTFS signals with low additional peak-to-average ratio pilots, which is applied to the signal receiving end and is used to obtain the information in the signals transmitted by the above-mentioned OTFS signal transmission method, includes the following steps:

[0016] Step R1, for the received signal, convert the Wigner transform and the symplectic Fourier transform with different parameters into OTFS demodulation under the same parameter, and separate the pilot signal through parameter transformation;

[0017] Step R2, use the pilot signal to perform a rough channel estimation based on a threshold to obtain the roughly estimated channel information; according to the roughly estimated channel information, perform a rough detection of the data signal;

[0018] Step R3, according to the roughly detected data signal, eliminate the interference of the pilot signal;

[0019] Step R4, use the pilot signal after interference cancellation to perform a fine estimation of the number of signal path channels;

[0020] Step R5, according to the finely estimated number of signal path channels, iteratively perform fine estimations of the channel delay, Doppler frequency shift, channel coefficient, and fine detection of the signal based on the finely estimated channel information.

[0021] Furthermore, the specific method of step R1 is:

[0022] Step R101, perform the Wigner transform and the symplectic Fourier transform with parameters M and N on the received time-domain signal to obtain the time-delay - Doppler domain signal;

[0023] Step R102, perform parameter transformation on the time-delay - Doppler domain signal to obtain a pilot matrix with parameters M1 and N1, which is the pilot signal.

[0024] Further, in step R2, parameter transformation is performed according to the roughly estimated channel information, interference cancellation is performed on the received signal with parameters M and N to obtain a roughly estimated received data matrix, and a rough detection of the data signal is performed using the message passing algorithm.

[0025] Further, in step R3, parameter transformation is performed using the roughly detected data signal, and interference cancellation is performed on the pilot matrix with parameters M1 and N1.

[0026] Further, the specific manner of step R4 is as follows:

[0027] Step R401: Using the pilot signal after interference cancellation, increase the number of channel paths to obtain new channel information, calculate the channel matrix according to the new channel information, and perform interference cancellation and signal detection;

[0028] Step R402: According to the new channel matrix and the signal detection result, calculate the 2-norm of the difference between it and the received signal, and select the channel information corresponding to the result with the smallest 2-norm as the updated channel information;

[0029] Step R403: Iteratively execute step R401 and step R402 until the number of channel paths no longer increases. At this time, the obtained number of channel paths is the accurately estimated number of channel paths.

[0030] Further, the specific manner of step R5 is as follows:

[0031] Step R501: According to the accurately estimated number of channel paths and its channel information, perform interference cancellation and signal detection on the data signal;

[0032] Step R502: According to the signal detection result, perform interference cancellation on the pilot signal, and then perform threshold-based channel estimation; at this time, the number of channel paths is fixed, and only the channel coefficient, time delay, and Doppler frequency shift are re-estimated;

[0033] Step R503: Iteratively execute step R501 and step R502 until the preset number of iterations is reached. The value range of the preset number of iterations is 2 to 5.

[0034] The present invention has the following advantages compared with the prior art:

[0035] (1) The low peak-to-average ratio time-domain superimposed pilot signal structure designed by the present invention disperses the single high-power embedded pilot separated from the data into multiple lower-power superimposed pilots, effectively reducing the peak-to-average ratio of the signal, thereby reducing system nonlinear distortion and improving channel estimation accuracy.

[0036] (2) The OTFS time-domain superimposed pilot of the present invention adopts a low-complexity implementation method, enabling pilots with different modulation parameters and data signals to use the same modulator, simplifying the modulator structure and complexity, and facilitating engineering implementation.

[0037] (3) The iterative channel estimation and signal detection method based on superimposed pilots proposed by the present invention has good channel estimation accuracy and demodulation performance, achieving almost the same BER performance as ideal channel estimation, and slightly superior to existing embedded pilot estimation methods. Description of the Drawings

[0038] Figure 1 It is a schematic diagram of the distribution method of OTFS data and pilots in the time-delay Doppler domain;

[0039] Figure 2 It is a schematic diagram of the principle of the low-complexity implementation method in the OTFS system;

[0040] Figure 3 It is the PAPR curve of different pilot structures when the pilot power is 20 dB higher than the data power provided by the simulation experiment of the present invention;

[0041] Figure 4 It is the BER rate curve of different pilot structures when the pilot power is 20 dB higher than the data power provided by the simulation experiment of the present invention;

[0042] Figure 5 It is the PAPR curve of different pilot structures when the pilot power is 26 dB higher than the data power provided by the simulation experiment of the present invention;

[0043] Figure 6 It is the BER rate curve of different pilot structures when the pilot power is 26 dB higher than the data power provided by the simulation experiment of the present invention. Detailed Embodiment

[0044] The present invention will be described in detail below with reference to the accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0045] An OTFS signal transmission method with a low additional peak-to-average power ratio pilot, applied to a signal transmitter, includes the following steps:

[0046] Step T1, in the time-delay - Doppler domain, map the data signal onto a two-dimensional grid with a time-delay parameter of M and a Doppler parameter of N, and map the low additional peak-to-average power ratio pilot signal onto a two-dimensional grid with a time-delay parameter of M1 and a Doppler parameter of N1 to form a pilot matrix; where MN = M1N1, M = kM1, and the pilot matrix is a matrix composed of multiple embedded pilots and guard intervals. The pilot signals are continuous along the time-delay axis and leave guard intervals on the Doppler axis to prevent interference with each other;

[0047] Figure 1 The figure shows a low peak-to-average ratio time-domain superposition pilot structure, where data and pilots use OTFS modulation with different parameters and are superposed in the time domain.

[0048] This step splits the traditional single embedded pilot into multiple pilots, dispersing the energy of the pilots as much as possible in the time domain to avoid excessive power differences between the pilot signal and the data. In addition, the pilots are continuous along the time-delay axis and a guard interval is left on the Doppler axis so that they do not interfere with each other.

[0049] Dispersing a single high-power embedded pilot separated from the data into multiple lower-power pilots superposed with the data can effectively reduce the peak-to-average ratio of the time-domain signal. The pilots are placed sequentially along the time-delay axis, and a guard interval is left on the Doppler axis so that they do not interfere with each other. The guard interval is set to 2 times the maximum Doppler frequency shift expansion. The energy of each pilot is the same, showing a small number of high-power signals in the time-delay-Doppler domain with parameters N1 and M1; and showing multiple low-power signals in the time-delay-Doppler domain with parameters N and M.

[0050] The superposition of the pilot signal and the data signal in the time domain can be simplified to superposition in the time-delay-Doppler domain, which can reduce the complexity of the OTFS modulator.

[0051] Step T2, as Figure 2 shown, in the time-delay-Doppler domain, through parameter transformation, the inverse symplectic Fourier transform and the Heisenberg transform performed on the data signal and the pilot signal respectively with different parameters are converted to OTFS modulation under unified parameters, and the time-domain superposition of the pilot signal and the data signal is converted to time-delay-Doppler domain superposition to obtain the transmitted signal and transmit it.

[0052] For OTFS modulation with parameters N1, M1 and N, M, the modulation parameters can be unified through a parameter transformation matrix, thereby simplifying the structure of the OTFS modulator. Among them, the parameter transformation matrix can be expressed as:

[0053]

[0054] Among them, represents rounding down, (i) k represents the remainder of i divided by k, P represents the row-column permutation interleaving matrix, (·) T represents the transpose of the matrix. represents a diagonal matrix with parameters as the main diagonal elements, i = 0, 1,..., MN - 1. I N represents the N-order identity matrix, F k represents the one-dimensional k-order DFT matrix, (·) H represents the conjugate transpose of the matrix, represents the Kronecker product.

[0055] In a specific embodiment, according to the system sampling rate and Doppler resolution, the quantization values of the maximum time delay and the maximum Doppler shift in the time-delay-Doppler grid are l max and k max . Referring to Figure 1 , in the time-delay-Doppler domain, the pilot signal is placed on a two-dimensional grid with a time delay number of M1 and a Doppler number of N1, where M1 represents the number of subcarriers corresponding to pilot modulation, and N1 represents the number of available time slots corresponding to pilot modulation; the data signal is placed on a two-dimensional grid with a time delay number of M and a Doppler number of N, where M represents the number of subcarriers corresponding to pilot modulation, and N represents the number of available time slots corresponding to pilot modulation. And MN = M1N1, M1 ≥ l max +1.

[0056] Exemplarily, and respectively represent the vectorized time-domain and time-delay-Doppler domain data signals, where represents the complex domain; x d = vec(X d ), vec(·) represents the column vectorization operation, X d = {X d [l,k] ∈ A, l = 0, 1,..., M - 1, k = 0, 1,..., N - 1}, represents the signal matrix in the time-delay-Doppler domain, represents the set of constellation points; and respectively represent the vectorized time-domain and time-delay Doppler domain pilot signals, x p = vec(X p ), represents the pilot matrix in the time-delay-Doppler domain, and can be expressed as:

[0057]

[0058] where, x p1 ,..., x pN represent multiple pilot signals placed in the time-delay-Doppler domain, and pN represents the total number of pilot signals.

[0059] Referring to Figure 1 , the pilots are placed sequentially along the time-delay axis, and a guard interval is left on the Doppler axis to prevent interference with each other. Since the range of the Doppler shift is [-k max , k max , the length of the guard interval is set to 2k max , and the energy of each pilot is the same. In the time-delay-Doppler domain with parameters N1 and M1, it appears as a small number of high-power signals.

[0060] Use OTFS modulation with parameters N1 and M1 for pilots and OTFS modulation with parameters N and M for data. Then, the pilots and data in the time domain can be expressed as:

[0061]

[0062] where I M represents the M - order identity matrix, F N represents the one - dimensional N - order DFT matrix, (·) H represents the conjugate transpose of the matrix, represents the Kronecker product. Then, the transmitted signal s in the time domain can be expressed as:

[0063] s = s d + s p

[0064] Different from the traditional OTFS system with embedded pilots, this method disperses a single high - power embedded pilot separated from the data into multiple lower - power pilots superimposed on the data, which can effectively reduce the peak - to - average power ratio of the system.

[0065] This method enables pilots and data signals with different modulation parameters to use the same modulator, simplifying the structure and complexity of the modulator.

[0066] In the parameter transformation, to keep MN = M1N1, assume N1 equals kN and M equals kM1. According to the above representation of the OTFS time - domain pilot signal, s p can be further decomposed as:

[0067]

[0068] where, represents rounding down to negative infinity, (i) k represents the remainder of i divided by k, P represents the interleaving matrix, (·) T represents the transpose of the matrix. diag represents the diagonal matrix with parameters as the main diagonal elements, i = 0, 1,..., MN - 1. The interleaving matrix P in the present invention represents a row - column permutation interleaving matrix, and the N - order row - column permutation interleaving matrix can be expressed as:

[0069] P = [e σ(1) , e σ(2) ,..., e σ(N)

[0070] where, e σ ​Denote a column vector with only one element being 1 and the rest being 0. σ is a permutation of {1, 2,..., N} used to define the row-column interleaving pattern, and the row-column interleaving pattern defined in the present invention is row-column permutation.

[0071] According to the properties of the Kronecker product We get:

[0072]

[0073] where Then the time-domain pilot signal can be re-expressed as:

[0074]

[0075] where x p′ is the pilot signal after parameter transformation, and its matrix dimension is M×N. According to the above derivation, for OTFS modulation with different parameters, the transformation between two parameters can be completed through matrix A. Therefore, pilots and data signals with different modulation parameters can use the same modulator, simplifying the structure and complexity of the modulator. Then the time-domain transmitted signal can be expressed as:

[0076]

[0077] In addition, referring to Figure 2 , due to the influence of the transformation matrix A, a small number of high-power pilot signals in the delay-Doppler domain under parameters N1 and M1 will be spread into a large number of low-power pilot signals in the delay-Doppler domain under parameters N and M, and there is still no interference between these pilot signals. At this time, the power difference between the data signal and the pilot signal is smaller than the power difference under different parameter modulations, which is beneficial to the actual hardware implementation.

[0078] An OTFS signal receiving method with a low additional peak-to-average ratio pilot is applied to the signal receiving end to obtain the information in the signal transmitted by the above OTFS signal transmitting method. The pilot signal and the data signal can be separated by a parameter transformation matrix at the receiving end, reducing the complexity of the OTFS demodulator. This method includes the following steps:

[0079] Step R1, for the received signal, first perform the Wigner transform and the symplectic Fourier transform with different parameters on the time-domain signal to obtain the delay-Doppler domain signal; the specific method is:

[0080] Step R101, perform the Wigner transform and the symplectic Fourier transform with parameters M and N on the received time-domain signal to obtain the delay-Doppler domain signal;

[0081] Step R102, perform parameter transformation on the delay-Doppler domain signal to obtain the pilot matrix with parameters M1 and N1, which is the pilot signal.

[0082] Step R2: Use the pilot signal to perform threshold-based rough channel estimation to obtain the rough estimated channel information; perform parameter transformation on the time-delay Doppler domain signal according to the rough estimated channel information to obtain a pilot signal matrix with parameters M1 and N1. Perform interference cancellation on the received signal with parameters M and N to obtain a rough estimated received data matrix, and use the message passing algorithm to perform rough detection of the data signal.

[0083] Step R3: Perform parameter transformation on the data signal detected roughly, and perform interference cancellation on the pilot matrix with parameters M1 and N1.

[0084] Step R4: Use the pilot signal after interference cancellation to perform fine estimation of the number of channel paths; the specific method is as follows:

[0085] Step R401: Use the pilot matrix after interference cancellation to increase the number of channel paths to obtain new channel information, calculate the channel matrix according to the new channel information, and perform interference cancellation and signal detection;

[0086] Step R402: According to the new channel matrix and the signal detection result, calculate the 2-norm of the difference between it and the received signal, and select the channel information corresponding to the result with the smallest 2-norm as the updated channel information;

[0087] Step R403: Iteratively execute Step R401 and Step R402 until the number of channel paths no longer increases. At this time, the obtained number of channel paths is the finely estimated number of channel paths.

[0088] Step R5: According to the finely estimated number of channel paths, iteratively perform fine estimation of the channel time delay, Doppler frequency shift, channel coefficient, and fine detection of the signal based on the finely estimated channel information; the specific method is as follows:

[0089] Step R501: According to the finely estimated number of channel paths and its channel information, perform interference cancellation and signal detection on the data signal;

[0090] Step R502: According to the signal detection result, perform interference cancellation on the pilot signal, and then perform threshold-based channel estimation; at this time, the number of channel paths is fixed, and only the channel coefficient, time delay, and Doppler frequency shift are re-estimated;

[0091] Step R503: Iteratively execute Step R501 and Step R502 until the preset number of iterations is reached, or the difference between the results of two channel estimations is less than the threshold. The value range of the preset number of iterations is 2 to 5.

[0092] This method adopts a pilot arrangement method that modulates data and pilots separately, reducing the impact of pilots on the Peak-to-Average Power Ratio (PAPR). Based on this kind of pilot, an iterative channel estimation and signal detection method is adopted to achieve a Bit-to-Error Rate (BER) performance almost the same as that of ideal channel estimation. Under the condition of using superimposed pilots to ensure high frequency utilization, this method reduces the impact of pilots on PAPR and improves the channel estimation performance by optimizing the pilot structure.

[0093] The following is a specific example of a receiving method:

[0094] This receiving method includes three stages: rough estimation of the channel and data, fine estimation of the number of signal paths of the channel, and accurate estimation of the channel coefficients and accurate detection of the signal.

[0095] In the first stage, rough estimation of the channel and data.

[0096] ① At the receiving end, perform OTFS demodulation and parameter transformation on the received signal to obtain the pilot signal y for channel estimation p :

[0097]

[0098] where r represents the received signal in the time domain, represents the channel matrix in the time domain, z represents the complex additive white Gaussian noise, represents the interference of the data signal on the pilot, represents the equivalent noise in the time-Doppler domain.

[0099] Assume that the pilot signals at all positions form a pilot vector at the receiving end after experiencing a time delay l and a Doppler shift k given by the following formula:

[0100]

[0101] where, represents the position information of the pilot at the transmitting end, i = 1,..., N p , N p represents the number of pilots placed at the transmitting end, x pp represents the pilot vector at the transmitting end, h (k,l) represents the channel coefficient corresponding to the actual signal path. The vector β (k,l) can be expressed as:

[0102]

[0103] The vector I is defined as:

[0104]

[0105] Noise vector z DD is defined as:

[0106]

[0107] ② Perform threshold-based rough channel estimation.

[0108] When the channel has a path with corresponding delay l and Doppler shift k, h (k,l) 10. The idea of rough estimation of the channel coefficient is: by using to eliminate the phase influence β of the pilot signal (k,l) and then using all pilots for threshold-based channel estimation. If the signal exceeds the threshold, it indicates that there is a pilot signal at that position. The threshold-based channel estimation can be expressed as:

[0109] There is a channel at [k, l]

[0110] There is no channel at [k, l]

[0111] where k = -k max ,..., k max , l = 0,... l max .

[0112] Through the above formula, we can obtain the estimated delay and Doppler shift as well as the roughly estimated number of channel paths where the corresponding channel coefficient is:

[0113]

[0114] Construct it into a vector:

[0115]

[0116] According to the estimated channel information, the roughly estimated time-domain - Doppler-domain channel can be obtained:

[0117]

[0118] where Π is the forward cyclic shift permutation matrix:

[0119]

[0120] Δ is a diagonal matrix:

[0121]

[0122] ③Coarse detection of the signal is performed using the result of coarse channel estimation.

[0123] First, the interference of the pilot on the data is eliminated, and the data signal can be obtained Subsequently, through detection using the message passing algorithm, the symbols obtained by coarse detection can be expressed as:

[0124]

[0125] where represents the set of constellation points, and Q represents the number of elements in the constellation point set.

[0126] For the initial estimation, due to the interference of the data part on the pilot part, the estimation of the number of channel paths, time delay, Doppler frequency shift, and channel coefficients is inaccurate. Therefore, subsequent fine estimations of the number of channel paths and channel coefficients are performed.

[0127] In the second stage, fine estimation of the number of channel paths is performed.

[0128] Assume the i-th iteration of fine estimation of the number of paths. We can use the result of the (i - 1)-th iteration to eliminate the interference of the data signal on the pilot signal; when it is the first iteration, the result of coarse channel estimation is used for interference cancellation:

[0129]

[0130] When i = 1:

[0131]

[0132] After eliminating the interference of the data signal, the signal used for channel estimation can be obtained through matrix A:

[0133]

[0134] All pilots used for channel estimation are obtained in the same way as the coarse estimation and formed into a vector Then, from the first elements with the largest modulus values are selected as the positions of the estimated time delay and Doppler frequency shift:

[0135]

[0136] where F(·) represents the maximum sorting function. Subsequently, the channel coefficients and data

[0137] In addition, select the largest elements in the same way as the estimated positions of the delay and Doppler shift:

[0138]

[0139] and the corresponding channel coefficients and data

[0140] respectively use the number of signal path and to perform signal detection, and compare with The ||·|| represents the 2-norm of the matrix, and the estimated result corresponding to the smaller value is used as the result of the refined estimation of the number of signal paths this time. Among them:

[0141]

[0142] The condition for stopping the iteration is that the estimated number of paths no longer increases.

[0143] In the third stage, iterative channel refinement estimation and signal refinement detection.

[0144] After completing the refined estimation of the number of paths and the positions of the delay and Doppler in the grid, let:

[0145]

[0146] This means that subsequent estimations of the signal paths and the positions of the delay and Doppler shift in the grid are no longer performed, but only the refined estimation of the channel coefficients is carried out.

[0147] First, at the initial iteration, let Assume that the r-th iteration of the channel coefficient refined estimation is performed, then the channel coefficient of the i-th path is:

[0148]

[0149] Among them, After estimating the channel coefficients, perform data detection:

[0150]

[0151] When the maximum number of iterations is reached or the difference between the results of two channel estimations is very small, the iteration ends.

[0152] The following shows the comparison of the signal peak-to-average ratio and reception performance of the present invention with those of the traditional method. It should be noted that the parameters in the simulation do not affect the generality of the present invention.

[0153] Exemplarily, set the OTFS frame parameters as N = 32, M = 64, the carrier frequency f c and the bandwidth are 4 GHz and 1 MHz respectively, the number of channel paths L is 4, set N1 to 256, M1 to 8, use quadrature phase shift keying (QPSK) modulation for bit mapping, and set the number of pilots to N p = 3M1 = 24. The channel gain is generated through distribution and the maximum number of iterations for channel fine estimation is set to 3, and the maximum number of iterations for joint estimation and detection is set to 2.

[0154] Simulation Experiment 1

[0155] Figure 3 shows the complementary cumulative distribution function (CCDF) curves of various pilot schemes and pure data transmission when the pilot power is 20 dB higher than the signal. It can be observed from the figure that the single superimposed pilot scheme has the highest PAPR, which is approximately 1 dB higher than the sparse superimposed pilot method. In contrast, the proposed method achieves almost the same PAPR as the embedded pilot and pure data signals.

[0156] Figure 4 shows the BER curves of various pilot schemes and in the ideal case. It can be seen from the figure that the performance of the sparse superimposed pilot scheme is very poor, and although the single superimposed pilot scheme is slightly better than the sparse superimposed pilot scheme, it still performs poorly. However, for the proposed pilot scheme and channel estimation method, its BER performance is not only almost the same as the ideal case, but also slightly better than the channel estimation scheme based on embedded pilots. This is mainly because the proposed pilot scheme uses more pilots, which suppresses the interference to the data, and the proposed channel estimation method uses data-aided refinement to estimate the channel parameters.

[0157] Simulation Experiment 2

[0158] Figure 5 shows the CCDF curves of various pilot schemes and pure data transmission when the pilot power is 26 dB higher than the signal power. It can be observed from the figure that the PAPR of the single superimposed pilot scheme is approximately 1 dB higher than the embedded pilot scheme, while the PAPR of the sparse superimposed pilot scheme is approximately 1 dB lower than the embedded pilot method. For the proposed pilot scheme, its PAPR is almost the same as pure data transmission and is approximately 2 dB lower than the sparse superimposed pilot scheme.

[0159] Figure 6are the bit error rate (BER) curves for various pilot schemes and in the ideal case. It can be observed from the figure that, compared with the sparse superimposed pilot scheme, the single superimposed pilot method achieves a performance gain of approximately 2 dB at high signal-to-noise ratio (SNR). As the pilot power increases, the performance of both schemes improves. However, the proposed pilot scheme and channel estimation method achieve almost the same BER performance as the ideal case and are slightly better than the embedded pilot estimation scheme.

[0160] In summary, the present invention can solve the problems of pilot deteriorating the peak-to-average power ratio of the signal and large data interference in the channel estimation in the current OTFS signal superimposed pilot method, and has the characteristics of low peak-to-average power ratio, low signal generation complexity, and high channel estimation accuracy.

[0161] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

Claims

1. A method for transmitting OTFS signals with low additional peak-to-average ratio pilots, applied to a signal transmitter, characterized in that, It includes the following steps: Step T1: In the time-delay - Doppler domain, map the data signal onto a two-dimensional grid with a time-delay parameter of M and a Doppler parameter of N, and map the pilot signal with a low peak-to-average power ratio onto a two-dimensional grid with a time-delay parameter of M1 and a Doppler parameter of N1 to form a pilot matrix; where MN = M1N1, M = kM1, and the pilot matrix is a matrix composed of multiple embedded pilots and guard intervals. The pilot signals are continuous along the time-delay axis and have guard intervals on the Doppler axis; Step T2: In the time-delay - Doppler domain, through parameter transformation, convert the inverse symplectic Fourier transform and Heisenberg transform with different parameters for the data signal and the pilot signal respectively to OTFS modulation under unified parameters, and convert the time-domain superposition of the pilot signal and the data signal to time-delay - Doppler domain superposition to obtain the transmitted signal and transmit it.

2. The OTFS signal transmission method with a low additional peak-to-average ratio pilot according to claim 1, wherein, In the pilot matrix, the energy of each pilot is the same, and the length of the guard interval is twice the maximum Doppler frequency shift expansion.

3. The OTFS signal transmission method with a low additional PAPR pilot according to claim 1, characterized in that In Step T2, convert the inverse symplectic Fourier transform and Heisenberg transform with different parameters for the data signal and the pilot signal respectively to OTFS modulation under unified parameters through a parameter transformation matrix; the parameter transformation matrix is: Among them, represents rounding down, (i) k represents the remainder of i divided by k, P represents the row-column permutation interleaving matrix, (·) T represents the transpose of a matrix; represents the diagonal matrix with parameter as the main diagonal elements, i = 0, 1,..., MN - 1; I N represents the N-order identity matrix, F k represents the one-dimensional k-order DFT matrix, (·) H represents the conjugate transpose of a matrix, represents the Kronecker product.

4. A method for receiving an OTFS signal with a low peak-to-average power ratio of pilots, which is applied to a signal receiving end and is used to obtain information in the signal transmitted by the OTFS signal transmission method described in any one of claims 1 to 3, characterized in that, It includes the following steps: Step R1: For the received signal, convert the Wigner transform and symplectic Fourier transform with different parameters to OTFS demodulation under the same parameter, and separate the pilot signal through parameter transformation; Step R2: Use the pilot signal to perform a rough channel estimation based on a threshold to obtain the rough estimated channel information; according to the rough estimated channel information, perform a rough detection of the data signal; Step R3: According to the roughly detected data signal, eliminate the interference of the pilot signal; Step R4: Use the pilot signal after interference cancellation to perform a fine estimation of the number of channel paths; Step R5: According to the finely estimated number of channel paths, iteratively perform fine estimations of the channel time delay, Doppler frequency shift, and channel coefficient, and fine detection of the signal based on the finely estimated channel information.

5. A method for receiving an OTFS signal with a low additional peak-to-average ratio pilot, according to claim 4, characterized in that, The specific method of Step R1 is: Step R101: Perform the Wigner transform and symplectic Fourier transform with parameters M and N on the received time-domain signal to obtain a time-delay - Doppler domain signal; Step R102: Perform parameter transformation on the time-delay - Doppler domain signal to obtain a pilot matrix with parameters M1 and N1, which is the pilot signal.

6. The OTFS signal receiving method with a low additional PAPR pilot according to claim 5, characterized in that, In Step R2, perform parameter transformation according to the roughly estimated channel information, perform interference cancellation on the received signal with parameters M and N, obtain a roughly estimated received data matrix, and use the message passing algorithm to perform a rough detection of the data signal.

7. A method for receiving an OTFS signal with a low additional peak-to-average ratio pilot, as claimed in claim 5, wherein In Step R3, perform parameter transformation using the roughly detected data signal to perform interference cancellation on the pilot matrix with parameters M1 and N1.

8. A method for receiving an OTFS signal with a low additional peak-to-average ratio pilot, as claimed in claim 5, wherein The specific method of Step R4 is: Step R401: Use the pilot signal after interference cancellation to increase the number of channel paths, obtain new channel information, calculate the channel matrix according to the new channel information, and perform interference cancellation and signal detection; Step R402: According to the new channel matrix and the signal detection result, calculate the 2-norm of the difference between it and the received signal, and select the channel information corresponding to the result with the minimum 2-norm as the updated channel information; Step R403: Iteratively execute Step R401 and Step R402 until the number of channel paths no longer increases. The number of channel paths obtained at this time is the accurately estimated number of channel paths.

9. A method for receiving an OTFS signal with a low additional peak-to-average ratio pilot, as claimed in claim 8, wherein The specific method of Step R5 is as follows: Step R501: Perform interference cancellation and signal detection on the data signal according to the accurately estimated number of channel paths and their channel information. Step R502: According to the signal detection result, perform interference cancellation on the pilot signal, and then perform threshold-based channel estimation. At this time, the number of channel paths is fixed, and only the channel coefficient, time delay, and Doppler frequency shift are re-estimated. Step R503: Iteratively execute Step R501 and Step R502 until the preset number of iterations is reached. The value range of the preset number of iterations is 2 to 5.

Images