OTFS signal peak-to-average ratio suppression method optimized by data symbol block selection

Abstract

This invention discloses an OTFS peak-to-average power ratio (PAPR) suppression method through optimized selection of data symbol blocks, belonging to the field of communication multi-carrier modulation technology. This invention introduces the Grey Wolf optimization algorithm into effective discrete encoding and decoding to solve discrete combinatorial optimization problems. It also introduces crossover and mutation operations to improve the global search capability of the optimization algorithm, and introduces neighborhood search operations to further enhance the local search capability. Addressing the problem of high PPR of transmitted signals easily exceeding the linear range of power amplifiers and causing nonlinear signal distortion, this invention optimizes the selection of OTFS frame data symbol blocks to obtain a better sub-block partitioning scheme, thereby achieving better OTFS signal PPR suppression performance. This ensures the signal operates within the linear range of the power amplifier, guaranteeing reliable communication transmission.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and specifically to a method for suppressing the peak-to-average power ratio (PAPR) of OTFS signals through optimized selection of data symbol blocks. Background Technology

[0002] Orthogonal Time-Frequency Space (OTFS) modulation technology can support reliable information transmission under high Doppler conditions and can be applied to high-dynamic communication scenarios such as satellite communication. As a multi-carrier technology, OTFS also suffers from a high peak-to-average power ratio (PAPR) due to the superposition of in-phase signals. When high-power signals enter the nonlinear region of the power amplifier, it causes nonlinear distortion of the signal, thereby affecting the overall performance of the system. Therefore, reducing PAPR is of great significance for improving communication reliability when applying OTFS modulation. Current work mainly includes distortion-based methods and probabilistic methods to address this problem.

[0003] Distortion-based methods preprocess signals with power exceeding a power threshold at the transmitting end by "clipping" them down to the threshold range before transmitting them. However, these methods cause nonlinear distortion in the signal, affecting the reliability of communication.

[0004] Probabilistic algorithms reduce the likelihood of peak signal power by employing interfering subcarrier phase shifting. Since the interfering subcarrier phase shifting process is linear, probabilistic methods typically do not cause nonlinear distortion to the signal, maintaining communication reliability. Partial Transfer Sequence (PTS) is a common probabilistic method. The PTS algorithm divides data symbols into blocks and selects appropriate phase rotation factors to obtain a signal with relatively low peak-to-average power (PAPS). However, dividing data symbols into blocks and selecting appropriate phase rotation factors for each block is a typical non-deterministic polynomial-hard (NP-hard) problem. Its solution can only be obtained through a global search. For a data symbol count of N, randomly and evenly divided into V sub-blocks, it is necessary to evaluate N! / (((NV)!) n There are 1000 possible partitioning methods (N and V) to find the optimal solution. Even with small N and V, solving the data symbol partitioning problem using a global search method is computationally quite complex.

[0005] Current PTS algorithms rarely study the optimal selection of data symbol blocks. Commonly used block methods include adjacent blocks, random blocks, and interleaved blocks. However, these coarse block methods often have limited performance in suppressing the peak-to-average power ratio (PAPR) of OTFS signals. Summary of the Invention

[0006] This invention provides a method for suppressing the peak-to-average ratio (PAPR) of OTFS signals through optimized selection of data symbol blocks. By improving the Grey Wolf optimization algorithm, a discrete encoding and decoding strategy is proposed to adapt to the discrete combinatorial optimization problem of this issue. Crossover and mutation operations from the genetic algorithm are introduced to enhance the global search capability. At the same time, the destruction and repair operation of the neighborhood search algorithm is introduced to further enhance the local search capability, thereby obtaining the optimal sub-block partitioning scheme and achieving better PPR suppression performance of OTFS signals.

[0007] The technical solution adopted in this invention is as follows:

[0008] The OTFS signal peak-to-average ratio suppression method, which optimizes the selection of data symbol blocks, includes the following steps:

[0009] Step 1, parameter initialization, including:

[0010] Set the OTFS signal Doppler number N and the number of sub-blocks V; set the gray wolf pack size I and the maximum number of iterations G. max Probability of mutation P m The destruction operator parameter D is initialized, the population consists of I gray wolf individuals, and the generation number g is initialized. Each gray wolf individual is a coding sequence of length N, where each element of the coding sequence takes a value from integer 1 to N. The coding sequence is then divided into V groups in a sequential, independent and uniform manner, i.e., each group corresponds to a sub-block.

[0011] Step 2: Calculate the peak-to-average power ratio (PAPR) for each gray wolf in the current generation, denoted as ρ. i (i=1,…,I), and set the three gray wolves with the smallest PAPR as the alpha wolves, define the three alpha wolves in ascending order of PAPR value as gray wolf α, gray wolf β, and gray wolf δ, and define ω to represent the level identifier of any non-alpha wolf;

[0012] Step 3: Determine if the current algebra g is less than G. max If yes, proceed to step 4; otherwise, proceed to step 8.

[0013] Step 4, Crossover operation: Iterate through each gray wolf individual of level ω and cross the current gray wolf individual with the three alpha wolves with equal probability. That is, in the crossover operation, only the gray wolf individuals of level ω are updated, and the three alpha wolves are retained.

[0014] Step 5, Mutation Operation: For any gray wolf individual of level ω, the mutation probability is P.m Mutation occurs;

[0015] Step 6, Neighborhood Search: Perform neighborhood search processing on the three alpha wolves based on the destruction operator parameter D;

[0016] Step 7, update the wolf pack: The transformed wolf pack is re-ranked according to their quality (i.e., objective function, PAPR). The three gray wolves with the lowest PAPR are set as the new alpha wolves, resulting in gray wolf α, gray wolf β, and gray wolf δ. The rank of the remaining gray wolves is set as ω, thus obtaining the gray wolf individuals of the g+1th generation.

[0017] Increment the algebra g by 1 and return to step 4;

[0018] Step 8: Decode the current gray wolf α. Based on the sub-block partitioning result corresponding to the current gray wolf α, obtain the optimal result of data symbol block optimization selection (i.e., PAPR optimal result). That is, rotate the sub-blocks corresponding to each group of the current gray wolf α's encoding sequence by the phase rotation factor, and then obtain the data symbol sequence corresponding to a delay grid based on all the groups after the phase rotation factor rotation.

[0019] Furthermore, the PAPR value for each gray wolf in the current generation is calculated as follows:

[0020]

[0021] Where E[] is the expectation operator, x i [k] represents the k-th coded symbol in the coded sequence of the i-th gray wolf, where k = 0, 1, ..., N-1.

[0022] Furthermore, the crossover operation in step 4 is specifically as follows:

[0023]

[0024] Where cross[·] represents the crossover operation between two gray wolves, and rand is a random number between 0 and 1, randomly generating the left and right crossover positions L. l L r Place two individual gray wolves at the intersection L l With L r Copy and paste the fragments from one instance to the beginning of the other instance's sequence, and then delete the duplicate elements at the second position from beginning to end.

[0025] Furthermore, the mutation operation in step 5 is similar to chromosome translocation in genetics. Specifically, it involves: first, randomly generating two sets of sequences of length L, where L ≤ N / 2, with the sequence elements of the two sets of sequences taking values ​​from 1 to N and being distinct from each other; the elements in the two sets of sequences are used to characterize the coding position index of the gray wolf individual to be mutated; then, the sequence elements of the two sets of sequences of length L are interchanged, and the coding elements of the gray wolf individual to be mutated are adjusted based on the interchanged coding position indexes to obtain the mutated gray wolf individual.

[0026] Furthermore, in step 8, the phase rotation factor used during phase rotation is: Where the sub-block index v = 1, 2, ..., V, Let v represent the phase of the v-th sub-block, which follows a uniform distribution on [0, 2π).

[0027] The technical solution provided by this invention brings at least the following beneficial effects:

[0028] Compared to commonly used segmentation methods, this invention has better OTFS signal PAPR suppression performance; at the same time, this invention can maintain the same transmission reliability as other segmentation methods while achieving lower PAPR. Attached Figure Description

[0029] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0030] Figure 1 This is a block diagram of the OTFS system;

[0031] Figure 2 This is a schematic diagram of the present invention used to optimize the PTS algorithm;

[0032] Figure 3 This is a diagram illustrating the cross operation;

[0033] Figure 4 This is a diagram illustrating the mutation operation;

[0034] Figure 5 This is a schematic diagram illustrating the correspondence between gray wolf encoded sequence elements and OTFS frame Doppler data in an embodiment;

[0035] Figure 6 This is a schematic diagram of the data blocks corresponding to OTFS sub-signals in the embodiment;

[0036] Figure 7 This is a comparison chart of the PAPR performance of this invention with other different methods;

[0037] Figure 8 This is a comparison chart of the BER performance of this invention with other different methods under the TDL-A model. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be described in detail and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Generally, the components of the embodiments of the present invention described and shown in the accompanying drawings can be arranged and designed using different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of the present invention.

[0039] OTFS system block diagram as follows Figure 1 As shown, at the system transmitter, the data symbols x[k,l], k=0,1,…N-1, l=0,1,…,M-1, after constellation mapping, are arranged in a two-dimensional time-delay-Doppler domain. Here, N is the number of Doppler grids of the OTFS signal, and M is the number of time-delay grids. The signal is converted to a two-dimensional video domain signal X[n,m] using an inverse symplectic finite Fourier transform (ISFFT). Then, multi-carrier modulation using the Heisenberg transform converts the two-dimensional time-frequency domain signal into a time-domain signal s(t). After transmission through the wireless channel, the transmitted signal is demodulated at the receiver using a Wigner transform to obtain the two-dimensional time-frequency domain signal Y[n,m]. Finally, the signal is recovered to a two-dimensional time-delay-Doppler domain signal using a symplectic finite Fourier transform (SFFT), where SFFT and Wigner transform are the inverse transforms of ISFFT and Heisenberg transform, respectively. OTFS modulation transforms the time-frequency domain bicolor dispersion channel into a time-delay-Doppler domain quasi-flat channel, which is more resistant to the time-varying characteristics of the channel.

[0040] The ISFFT is shown in equation (1):

[0041]

[0042] The Heisenberg transform is shown in equation (2):

[0043]

[0044] Among them, g tx(t) represents the transmitted pulse waveform, N represents the number of Doppler grids (number of symbols), M represents the number of delay grids (number of subcarriers), T represents the time-domain sampling duration, and Δf represents the subcarrier interval.

[0045] The PAPR of the OTFS signal is represented as:

[0046]

[0047] Where E[] is the expectation operator. Due to the randomness of communication data, the time-domain transmitted signal of OTFS exhibits envelope fluctuation characteristics. The complementary cumulative distribution function (CCDF) is typically used to describe the PAPR performance of the OTFS signal. The CCDF curve represents the probability that the PAPR exceeds a certain threshold of the OTFS signal. The CCDF expression for the OTFS signal is:

[0048]

[0049] Where Pr() represents the probability function and γ is a constant.

[0050] The time-delay-Doppler domain data symbol x[k,l] is converted into a time-frequency domain signal X[n,m] by ISFFT. Based on a specific partitioning method, the time-frequency signal X[n,m] is divided into V independent sub-data blocks, denoted as Xn. v [n,m], v=1,2…,V, where no two sets of data blocks overlap, and the number of delay grid information in each data block is N / V, as shown in the following formula:

[0051]

[0052] Next, we introduce the phase rotation factor. v = 1, 2, ..., V, where The data follows a uniform distribution on [0, 2π). The sub-data blocks are multiplied by the corresponding phase rotation factor to achieve phase rotation of the sub-data blocks, resulting in a new time-frequency domain signal X′[n, m]:

[0053]

[0054] Performing a Heisenberg transform on the new time-frequency domain signal yields a new time-domain signal s′(t):

[0055]

[0056] A schematic diagram of this invention, which utilizes the Grey Wolf algorithm to optimize PTS data symbol segmentation to reduce the peak-to-average power ratio (PAPR) of OTFS signals, is shown below. Figure 2 As shown.

[0057] The Grey Wolf Optimization Algorithm, which simulates a series of social behaviors such as wolf hierarchy, encirclement, and hunting, is an optimization problem in a continuous domain and cannot be directly solved for discrete combinatorial problems. Furthermore, the Grey Wolf Algorithm is prone to getting trapped in local optima. This invention improves the Grey Wolf Optimization Algorithm by introducing a discrete encoding / decoding strategy, along with operations such as crossover, mutation, and neighborhood search, to adapt it to discrete combinatorial optimization problems and select the optimal sub-block partitioning and combination.

[0058] For the problem of partitioning data symbols into sub-blocks with N Doppler lattice numbers and V number of blocks, the specific implementation steps of the Grey Wolf algorithm for sub-block selection include:

[0059] Population initialization: Set the gray wolf population size to I, and the maximum number of iterations for the gray wolf pack to G. max Each individual gray wolf is encoded by a random arrangement of 1 to N different integers. This encoded sequence is then divided into V groups in a sequential, independent and uniform manner for initialization.

[0060] Choosing the Alpha Wolf: For the g-th generation (the initial value of g is usually set to 1, but it can also start from 0), calculate the PAPR value of each gray wolf, denoted as ρ. i (i=1,…,I) and set the three gray wolves with the smallest PAPR as the alpha wolves. Arrange the gray wolves in order of PAPR value from smallest to largest as gray wolf α, gray wolf β, gray wolf δ, and the other gray wolves as the lowest level gray wolf ω. That is, set the level of the gray wolves that are not alpha wolves as ω.

[0061] Crossover operation: Let the current gray wolf individual be w (a gray wolf individual of level ω), and the gray wolf α individual be w α The number of individuals of gray wolf β is w β The individual gray wolf δ is w δ The update strategy for the individual gray wolf w is as follows:

[0062]

[0063] Here, `cross[]` represents a crossover operation between two individual gray wolves, and `rand` is a random number between 0 and 1. The crossover operation is as follows: Figure 3 As shown, the intersection position L is randomly generated. l L r (In the figure, 2 and 5) Copy and paste the segment between the intersection of the two individuals to the beginning of the other individual sequence, then delete the duplicate element at the second position from beginning to end, and finally update only the current gray wolf individual w, keeping the alpha wolf individual.

[0064] Mutation operation: Any individual gray wolf w will have a certain probability P m Mutations occur, enhancing global search capabilities. The mutation pattern is similar to chromosomal translocations in genetics, such as... Figure 4As shown, the specific process is as follows: First, two sets of sequences of length L are randomly generated, where L ≤ N / 2. The sequence elements of the two sets of sequences take values ​​from 1 to N and are distinct from each other. The elements in the two sets of sequences are used to represent the coding position indices of the gray wolf individuals to be mutated. Then, the elements of the two sets of sequences of length L are interchanged, and the coding elements of the gray wolf individuals to be mutated are adjusted based on the interchanged coding position indices to obtain the mutated gray wolf individuals. From Figure 4 As can be seen, in this embodiment, two sequences of length 3 (6,3,7 and {1,2,4}) were randomly selected. The mutation process involves replacing the element at index 6 with the element at index 1, which corresponds to replacing the encoded element 3 with the encoded element 1 in the original gray wolf individual; replacing the element at index 3 with the element at index 2, which corresponds to replacing the encoded element 4 with the encoded element 2 in the original gray wolf individual; and replacing the element at index 7 with the element at index 4, which corresponds to replacing the encoded element 8 with the encoded element 5 in the original gray wolf individual. This yields the encoded result of the mutated gray wolf w'.

[0065] Neighborhood Search: Perform a neighborhood search on the three alpha wolves, namely gray wolf α, gray wolf β, and gray wolf δ, to obtain a better gray wolf individual. The neighborhood search is a destruction-repair operation. Use the destruction operator to randomly select D elements from the gray wolf individual, and then use the repair operator to scramble the D elements and fill them into the positions of the missing elements in the gray wolf individual.

[0066] Updating the wolf pack: The (g+1)th generation wolf pack is obtained from the (g)th generation wolf pack through operations such as crossover, mutation, and neighborhood search. The (g+1)th generation wolf pack is then re-ranked according to the quality of each gray wolf (i.e., the objective function, PAPR value).

[0067] Example

[0068] As one possible implementation, the embodiments of the present invention provide specific implementation steps for the OTFS signal peak-to-average ratio suppression method selected through symbol block optimization, including:

[0069] Step S1: Parameter initialization, including:

[0070] Set the OTFS signal Doppler grid number N, delay grid number M, and sub-block number V; set the wolf pack size I and the maximum number of iterations G. max Probability of mutation P m This disrupts the operator parameter D and the initialization algebra g = 1;

[0071] Step S2, calculate the PAPR (parameters per second) of each gray wolf in the current generation, i.e., ρ. i (i=1,…,I), select 3 alpha wolves according to PAPR values ​​from smallest to largest and label them as gray wolf α, gray wolf β, and gray wolf δ, and the remaining gray wolf individuals are labeled as gray wolf w;

[0072] Step S3, if g < G max If yes, proceed to step S4; otherwise, proceed to step S8.

[0073] Step S4: Perform a crossover operation on the gray wolf w;

[0074] Step S5: Perform a mutation operation on the gray wolf w;

[0075] Step S6: Perform a neighborhood search operation on gray wolf α, gray wolf β, and gray wolf δ;

[0076] Step S7: Set the updated individuals as generation g+1, and reclassify them according to their quality (i.e., objective function, PAPR), updating the iteration number g = g+1.

[0077] Step S8: Output the sub-block partitioning scheme corresponding to the current gray wolf α and the optimal PAPR result, and transmit the low peak-to-average ratio OTFS time-domain signal.

[0078] That is, in this embodiment, such as Figure 5 As shown, for any individual gray wolf, its encoded sequence elements correspond to the Doppler data index of the original OTFS frame. Its encoded sequence is independently and uniformly divided into V groups, denoted as group v, where v = 1, 2, ..., V. The original OTFS frame is then divided into V groups according to the partitioning rules. Each group has MN / V non-zero elements, and their positions correspond to the Doppler data of the partitioned encoded sequence elements. The signal of group v is denoted as X. v [n, m]. The specific partitioning method is as follows: Figure 6 As shown.

[0079] To further verify the performance of the method of this invention, simulation verification was conducted through test experiments. The specific parameter configurations for this test experiment are as follows: OTFS modulation symbol number N = 16, number of subcarriers M = 128, subcarrier spacing 15kHz, carrier frequency 4GHz, modulation scheme QPSK, number of sub-blocks V = 4, phase rotation factor set [1,j,-1,-j], wolf pack size I = 10, and maximum number of iterations G. max =30, mutation probability P m =0.2, and the destruction operator parameter D=3. The channel selected is the TDL-A channel model in the satellite-to-ground scenario defined by 3GPP, and its parameters are shown in Table 1. The satellite parameter settings are shown in Table 2.

[0080] Table 1

[0081]

[0082] Table 2

[0083] parameter Value Satellite altitude 1500km Satellite speed 7.11km / h Central angle 40°

[0084] Figure 7 This is a comparison chart of the PAPR performance of the method of this invention with other different methods, namely the original OTFS signal, adjacent block segmentation, random block segmentation, and interleaved block segmentation methods. From Figure 7 As can be observed, the method of the present invention has better peak-to-average power ratio (PAPR) suppression performance compared to other methods. When CCDF = 10... -3 At that time, compared to the original OTFS signal, the PAPR performance of the method of the present invention decreased by approximately 3.6 dB. Compared with traditional sub-block partitioning methods such as adjacent block partitioning, random block partitioning, and interleaved block partitioning, the PAPR performance of the method of the present invention is improved by approximately 1–1.5 dB. Therefore, it is demonstrated that the method of the present invention is effective in suppressing the PAPR of the OTFS signal.

[0085] Figure 8 This chart compares the bit error rate (BER) performance of the method described in this invention with other methods under the satellite-to-ground TDL-A channel model, specifically the original OTFS signal, adjacent block segmentation, random block segmentation, and interleaved block segmentation methods. Figure 6 As can be seen, the method of this invention has almost the same BER performance as other methods and the original OTFS signal. Therefore, it is proven that the method of this invention can maintain good communication transmission reliability while effectively suppressing the peak-to-average power ratio of the OTFS signal.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0087] The above descriptions are merely some embodiments of the present invention. Those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A peak-to-average power ratio (PAPR) suppression method for OTFS signals selected through data symbol block optimization, characterized in that, Includes the following steps: Step 1, parameter initialization, including: Set the OTFS signal Doppler grid number N, delay grid number M, and sub-block division number V; set the gray wolf pack size I and the maximum number of iterations G. max Probability of mutation P m The destruction operator parameter D is used to initialize the I gray wolf individuals in the population, and the generation number g is initialized. Step 2: Calculate the peak-to-average power ratio (PAPR) for each gray wolf in the current generation, denoted as ρ. i (i=1,…,I), and set the three gray wolves with the smallest PAPR as the alpha wolves, define the three alpha wolves in ascending order of PAPR value as gray wolf α, gray wolf β, and gray wolf δ, and define ω to represent the level identifier of any non-alpha wolf; Step 3: Determine if the current algebra g is less than G. max If yes, proceed to step 4; otherwise, proceed to step 8. Step 4, Crossover operation: Iterate through each gray wolf individual of level ω and cross the current gray wolf individual with the three alpha wolves with equal probability. That is, in the crossover operation, only the gray wolf individuals of level ω are updated, and the three alpha wolves are retained. Step 5, Mutation Operation: For any gray wolf individual of level ω, the mutation probability is P. m The mutation occurs; the mutation is similar to chromosomal translocation in genetics, that is, any two sets of equal-length sequences in a gray wolf individual are replaced by each other; Step 6, Neighborhood Search; Perform a neighborhood search on the 3 alpha wolves to obtain the better gray wolf individuals in the current generation; The neighborhood search operation is also a destruction and repair operation. Use the destruction operator to randomly select D elements from the gray wolf individuals of level ω, and then use the repair operator to scramble the D elements and fill them into the positions of the missing elements in the gray wolf individuals. Step 7, update the wolf pack: After the transformation in steps 4-6, the wolf pack is re-ranked according to PAPR: The three gray wolves with the lowest PAPR are set as the new alpha wolves, resulting in gray wolf α, gray wolf β, and gray wolf δ, and the rank of the remaining gray wolves is set as ω, thus obtaining the gray wolf individuals of the g+1 generation. Increment the algebra g by 1 and return to step 4; Step 8: Perform phase factor rotation on the sub-block partitioning result corresponding to the current gray wolf α to obtain the optimal result of data symbol block optimization selection.

2. The method as described in claim 1, characterized in that, The PAPR value for each gray wolf in the current generation is calculated as follows: Where E[] is the expectation operator, x i [k] represents the k-th coded symbol in the coded sequence of the i-th gray wolf, where k = 0, 1, ..., N-1.

3. The method as described in claim 1, characterized in that, The crossover operation in step 4 is as follows: Where cross[·] represents the crossover operation between two gray wolves, and rand is a random number between 0 and 1, randomly generating the left and right crossover positions L. l L r Place two individual gray wolves at the intersection L l With L r Copy and paste the fragments from one instance to the beginning of the other instance's sequence, and then delete the duplicate elements at the second position from beginning to end.

4. The method as described in claim 1, characterized in that, The mutation operation in step 5 is as follows: two sets of sequences of length L are randomly generated, where L≤N / 2. The sequence elements of the two sets of sequences take values ​​from 1 to N and are different from each other. The elements in the two sets of sequences are used to represent the coding position index of the gray wolf individual to be mutated. Then, the sequence elements of the two sets of sequences of length L are replaced with each other. The coding elements of the gray wolf individual to be mutated are adjusted based on the mutually replaced coding position index to obtain the mutated gray wolf individual.

5. The method as described in claim 1, characterized in that, In step 8, the phase rotation factor used during phase rotation is: Where the sub-block index v = 1, 2, ..., V, Let v represent the phase of the v-th sub-block, which follows a uniform distribution on [0, 2π).

Images