A communication method based on integrated waveform design of orthogonal time-frequency-space

Abstract

A communication method based on an integrated waveform design of orthogonal time-frequency space relates to the field of communication technology. The present invention aims to solve the problem that the integrated waveform generated during the communication process in existing communication methods has poor adaptability, resulting in a single waveform application scenario. The present invention includes: performing constellation mapping on the serial data and pre-processing to obtain a baseband data signal vector x, performing serial / parallel conversion on x along the row to obtain a baseband data signal matrix X; thereby obtaining a one-dimensional time domain signal s α,β (t); intercept s α,β (t) The preset length signal at the end is used as a cyclic prefix to obtain s' α,β (t), using the transmitter to transmit s' α,β (t), get the receiver received signal r' α,β (t); r' α,β (t) is converted into the time domain received signal matrix Y′, thereby obtaining the received signal vector r α,β ; for r α,β Equalize to obtain #imgabs0#, perform WFRWHT and WFRFT on #imgabs1# along the rows, and restore #imgabs2# to the baseband received signal matrix Y; perform row-parallel / serial transformation on Y to obtain the baseband received signal vector y, perform constellation inverse mapping on y to obtain the transmitted information, and complete communication. The present invention is used for communication based on waveform design.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a communication method based on orthogonal time-frequency-space integrated waveform design. Background Art

[0002] Wireless communication technology has evolved dramatically over the past few decades, from the earliest analog signal transmissions to modern digital communication systems. With the growth in the number of mobile devices, increasing user demands, and the expansion of communication frequency bands, the next generation of physical layer waveforms must offer high reliability, low latency, and the ability to overcome complex and changing channel environments.

[0003] The paper "On the Performance of Integrated Orthogonal Time-Frequency-Space Framework based on WFRFT" proposes a WFRFT-based OTFS waveform system framework. This framework can be characterized as waveforms such as OTFS, adaptive OFDM, HC, OFDM, and SC by flexibly selecting the WFRFT order and information symbol length. Compared to single waveform systems, the WFRFT-OTFS waveform offers better compromise and flexibility, adapting to changing channel environments and complex requirements. However, while this waveform framework enables two-dimensional parameter adjustment, the selected parameters are the WFRFT order and number of points. As a result, the waveforms represented by this waveform framework are more homogeneous and cannot be successfully characterized as OTSM waveforms, resulting in poor adaptability of the integrated waveform. The paper "A Novel OTFS System Based on DFrFT-OFDM" proposes a novel OTFS system design based on the discrete fractional Fourier transform (DFrFT) OFDM system. By adjusting the DFrFT order, this system improves bit error rate (BER) and peak-to-average power ratio (PAPR) performance without increasing receiver complexity. However, the DFrFT dimension expansion direction of this system framework is relatively specific, and it cannot achieve the integration of existing waveforms, especially the unification of OTSM and SC waveforms. This results in poor adaptability of the generated integrated waveform, which in turn leads to a limited number of waveform application scenarios. Summary of the Invention

[0004] The purpose of the present invention is to solve the problem that the integrated waveform generated during the communication process in the existing communication methods has poor adaptability, resulting in a single waveform application scenario, and proposes a communication method based on orthogonal time-frequency space integrated waveform design.

[0005] A communication method based on orthogonal time-frequency-space integrated waveform design has the following specific processes:

[0006] Step 1: Perform constellation mapping on the serial data to be sent, and then preprocess the constellation mapping result to obtain a baseband data signal vector x with a length of M×N in the fractional domain. Perform serial / parallel conversion on x along the row to obtain a baseband data signal matrix X=[x0,x1,...,x M-1 ] T ;

[0007] Where M is the number of subcarriers, N is the number of symbols in each column of the signal vector, and the amount of serial data to be sent is (Ml max )×N;

[0008] The constellation mapping result is pre-processed to obtain the baseband data signal vector x of M×N columns in the fractional domain:

[0009] First, the serial data to be sent after the constellation mapping result is padded with zeros;

[0010] Among them, the total length of zero padding is l max ×N;l max is the zero-padded length of each column of the signal vector;

[0011] Then, the zero-filled x is divided into M signal vectors to obtain the baseband data signal vector of M×N columns in the fractional domain

[0012] Step 2: Use the X obtained in step 1 to obtain the one-dimensional time domain signal s α,β (t);

[0013] Step 3: Intercept s α,β (t) The signal with the preset length at the end is used as a cyclic prefix, and the cyclic prefix is ​​connected to the signal s α,β (t) before obtaining s' α,β (t), and then use the transmitter to transmit s' α,β (t), thereby obtaining the signal r' received by the receiver α,β (t);

[0014] Step 4: The signal r' received by the receiver α,β (t) is converted into the time domain received signal matrix Y′, and Y′ is used to obtain the received signal vector r α,β ;

[0015] Step 5: Receive signal vector r αβ Perform equalization to obtain the equalized time domain received signal matrix

[0016] Step 6: Perform N-point β-order WFRWHT and N-point 1-α-order WFRFT along the row. Restore to baseband received signal matrix Y;

[0017] Among them, α and β are values ​​between 0 and 1;

[0018] Step 7: Perform parallel / serial conversion on Y to obtain the baseband received signal vector Then perform constellation inverse mapping on y to obtain the transmitted information and complete the communication process.

[0019] Furthermore, the one-dimensional time domain signal s is obtained by using the X obtained in step 1 in step 2. α,β (t), comprising the following steps:

[0020] Step 2. Perform N-point α-1 order WFRFT and N-point β-order WFRWHT along the row of X to obtain the data X′ in the time domain, as shown in the following formula:

[0021]

[0022] in, is the N-point α-1 order WFRFT transform, is the N-point-β-order WFRWHT transformation;

[0023] Step 22: Perform DFT along the columns of X′ obtained in step 21 to obtain the frequency domain signal X″, as shown in the following formula:

[0024] X″=F M X′

[0025] Among them, F M is the DFT transform;

[0026] Step 2: Perform Heisenberg transform on X″ to obtain a one-dimensional time domain signal s α,β (t).

[0027] Furthermore, in steps 2 and 3, the Heisenberg transform is performed on X″ to obtain a one-dimensional time domain signal s α,β (t), as follows:

[0028]

[0029] Among them, X″[m,n] represents the element in the mth row and nth column of X″, g tx (t) is the time-domain continuous baseband pulse shaping function, Δf is the subcarrier bandwidth, T is the subsymbol duration, t is time, and j is the imaginary unit.

[0030] Furthermore, the signal r' received by the receiver in step 3 α,β (t), as follows:

[0031]

[0032] Where h(τ,t) represents the continuous time-varying channel impulse response in the delay-time domain, and n(t) represents the variance σ 2 Continuous Gaussian white noise, τ is the time delay, τ max is the maximum delay.

[0033] Furthermore, in step 4, the signal r' received by the receiver is α,β (t) is converted into the time domain received signal matrix Y′, and Y′ is used to obtain the received signal vector r α,β , including the following steps:

[0034] Step 4.1. Remove r' α,β (t) to obtain the cyclic prefix of the received signal r α,β (t), r α,β (t) Perform Wigner transform to obtain discrete frequency domain signal Y″[m,n], thereby obtaining the frequency received signal matrix Y″, as shown in the following formula:

[0035]

[0036] Among them, g rx (t) is the receiving shaping filter, Y″[m,n] is the element in the mth row and nth column of the frequency domain received signal matrix Y″, υ represents the Doppler shift, [·] * represents the conjugation of [·];

[0037] Step 42: Perform IDFT transformation along the columns of Y″ obtained in step 41 to obtain the time domain received signal matrix Y′, thereby obtaining the received signal vector r α,β .

[0038] Furthermore, in step 42, the Y″ obtained in step 41 is subjected to IDFT transformation along the columns to obtain the time domain received signal matrix Y′, thereby obtaining the received signal vector r α,β , specifically:

[0039] First, perform IDFT along the columns of Y″ obtained in step 4-1 to obtain the time domain received signal matrix Y′:

[0040]

[0041] in,[·] H represents the conjugate transpose;

[0042] Then, use Y′ to obtain the received signal vector rα,β :

[0043]

[0044] Furthermore, the received signal vector r in step 5 α,β Equalization is performed using MMSE equalization.

[0045] Furthermore, the received signal vector r in step 5 α,β GS equalization is used for equalization.

[0046] Furthermore, the received signal vector r in step 5 α,β Balancing is achieved by:

[0047] First, use MMSE equalization to receive the signal vector r α,β The MMSE equalization result is obtained by equalization, and then the MMSE equalization result is subjected to GS equalization.

[0048] Furthermore, in step six, Perform N-point β-order WFRWHT and N-point 1-α-order WFRFT along the row. Restored to the baseband received signal matrix Y, as follows:

[0049]

[0050] The beneficial effects of the present invention are:

[0051] The present invention proposes a waveform integration framework with adjustable two-dimensional parameters of WFRFT-WFRWHT-OTFS, which combines the two modules of WFRFT and WFRWHT in a novel way, achieves the unification of multiple waveforms while characterizing the OTFS waveform, and improves the success rate of characterizing the OTFS waveform, thereby improving the adaptability of the integrated waveform. The present invention can flexibly select appropriate two-dimensional parameters according to the needs of the user and the changes in the channel environment, thereby realizing the joint regulation of BER and PAPR performance in different scenarios. The present invention provides an efficient and flexible waveform design method for the communication system, improves the adaptability of the integrated waveform, and can adapt to different communication scenarios and application requirements. The present invention can realize communication based on different waveforms. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a flowchart of the process of the present invention;

[0053] Figure 2 This is a diagram of the emission waveform mechanism of the present invention;

[0054] Figure 3The bit error rate performance curve of the transmission waveform of the present invention when using different receivers;

[0055] Figure 4 is the PAPR performance curve of the present invention. DETAILED DESCRIPTION

[0056] Orthogonal Time Frequency Space (OTFS) modulation has excellent Doppler shift resistance and performs well in time-varying channels, making it a hot topic of research. However, OTFS has disadvantages such as high modulation complexity and high PAPR, which greatly limit its use. In addition, Orthogonal Time Sequency Multiplexing (OTSM) modulation can multiplex information symbols in the delay-sequence domain, allowing each information symbol to independently pass through a dual-dispersion channel, thus achieving Doppler shift resistance similar to OTFS. Compared with OTFS, the OTSM modulation scheme maintains the same performance but has lower modulation complexity. Therefore, the OTSM modulation scheme has broader prospects and application potential in future research and applications.

[0057] On the other hand, hybrid carrier (HC) technology based on weighted fractional Fourier transform (WFRFT) has been widely studied and applied in modern communication systems. HC technology based on WFRFT realizes the fusion of traditional single carrier (SC) and multi-carrier (MC) waveforms, and can simultaneously achieve the performance advantages of MC waveforms and SC waveforms. By changing the order of WFRFT, HC technology based on WFRFT can flexibly adjust the distribution of time-frequency energy to better adapt to complex and changing channel environments. In addition, WFRFT can be regarded as a precoding technology with high flexibility and adaptability, thereby expanding the dimension of the traditional MC waveform to obtain a more generalized HC waveform. This feature makes the HC system based on WFRFT more flexible to design and implement.

[0058] The present invention uses four weighted fractional Fourier transforms and weighted fractional Walsh-Hadamard transforms, as follows:

[0059] A.WFRFT Theory

[0060] The 4-item weighted N-point WFRFT matrix of order α can be expressed as:

[0061]

[0062] Among them, I N is the N-dimensional identity matrix, F N is the N-dimensional DFT matrix. Where the weighting coefficient ω l (α) can be expressed as follows:

[0063]

[0064] Among them, when α = 0, the WFRFT matrix becomes the unit matrix, when α = 1, the WFRFT matrix becomes the DFT matrix (discrete Fourier transform matrix), and j is the imaginary unit. In addition, the weighted fractional Fourier transform satisfies the rotation additivity, that is,

[0065] B.WFRWHT theory

[0066] The expression of the 4-term weighted N-point WFRWHT with order β is as follows:

[0067]

[0068] Among them, W N is the Walsh-Hadamard matrix;

[0069] Weighting coefficient ω l (β) is expressed as follows:

[0070]

[0071] Among them, when β = 0, the WFRWHT matrix becomes the identity matrix, and when β = 1, the WFRWHT matrix becomes the WHT matrix (Walsh-Hadamard matrix). In addition, the weighted fractional Walsh-Hadamard transform satisfies the rotation additivity, that is,

[0072] Next, the present invention will be described with reference to specific embodiments.

[0073] Specific implementation method 1: Figure 1 As shown, the specific process of the communication method based on orthogonal time-frequency-space integrated waveform design in this embodiment is as follows:

[0074] Step 1: (Ml max )×N serial data to be sent are constellation mapped, and then the constellation mapping result is preprocessed to obtain the M×N column baseband data signal vector x in the fractional domain, and x is serial / parallel converted along the row to obtain the baseband data signal matrix X=[x0,x1,...,x M-1 ] T ;

[0075] The constellation mapping result is preprocessed to obtain an M×N column baseband data signal vector x in the fractional domain, specifically:

[0076] Fill the serial data to be sent after constellation mapping with zeros max ×N length, obtaining a baseband data signal vector x of M×N columns in the fractional domain;

[0077] Where M is the number of subcarriers, N is the number of symbols per column vector; max is the length of zero padding for each column vector, l max It needs to be long enough to ensure that no inter-block interference occurs.

[0078] Step 2: Use the X obtained in step 1 to obtain the one-dimensional time domain signal s α,β (t), specifically:

[0079] Step 2. Perform N-point α-1 order WFRFT and N-point β-order WFRWHT along the row of X to obtain the data X′ in the time domain, which is expressed as:

[0080]

[0081] in, is the N-point α-1 order WFRFT transform, is the N-point-β-order WFRWHT transformation;

[0082] Step 22: Perform DFT along the columns on the time domain data X′ obtained in step 21 to convert it into a frequency domain signal X″, which is expressed as:

[0083] X″=F M X′ (6)

[0084] Among them, F M is the discrete Fourier transform;

[0085] Step 2: Perform Heisenberg transform on X″ to obtain a one-dimensional time domain signal s α,β (t)

[0086]

[0087] Among them, X″[m,n] represents the element in the mth row and nth column of X″, g tx (t) is the time-domain continuous baseband pulse shaping function, Δf is the subcarrier bandwidth, T is the subsymbol duration, and t is time.

[0088] Step 3: Intercept the time domain signal s α,β The signal with a preset length at the end of (t) is used as a cyclic prefix to ensure that this frame signal will not be affected by the previous frame signal. α,β(t) before, in order to avoid the mutual influence between each signal, and then use the transmitter to transmit the signal s' α,β (t), s' α,β (t) After passing through the channel, obtain the signal r' received by the receiver α,β (t):

[0089]

[0090] Where h(τ,t) represents the continuous time-varying channel impulse response in the delay-time domain, and n(t) represents the variance σ 2 Continuous Gaussian white noise, τ is the time delay, τ max is the maximum delay.

[0091] Step 4: The signal r' received by the receiver α,β (t) is converted into the time domain received signal matrix Y′, and Y′ is used to obtain the received signal vector r α,β , including the following steps:

[0092] Step 4.1. In the receiver, remove r' α,β (t) cyclic prefix, and obtain the received signal r α,β (t), r α,β (t) After the Wigner transform, the discrete frequency domain signal Y″[m,n] is obtained, thereby obtaining the frequency reception signal matrix Y″:

[0093]

[0094] Among them, g rx (t) is the receiving shaping filter, Y″[m,n] is the element in the mth row and nth column of the frequency domain received signal matrix Y″, υ represents the Doppler shift, [·] * represents the conjugation of [·];

[0095] Step 42: Perform IDFT (Inverse Discrete Fourier Transform) along the columns of Y″ obtained in step 41 to obtain the time domain received signal matrix Y′, thereby obtaining the received signal vector r α,β :

[0096]

[0097] in,[·] H represents the conjugate transpose.

[0098] Let the received signal vector r of the MN column be α,β Expressed as the received signal r α,β (t) The result of sampling at intervals of T / M, r α,β Divide into N vectors, that is

[0099] Step 5: Receive signal vector r α,β Perform equalization to obtain the equalized time domain received signal matrix

[0100] The present invention uses two equalization methods, namely MMSE equalization and GS equalization.

[0101] For the received signal vector r in the time domain received signal matrix Y′ α,β Equalization can be performed using any of the following three methods:

[0102] 1) For r α,β First perform MMSE equalization, then perform GS equalization on the result of MMSE equalization;

[0103] 2) For r α,β Perform MMSE equalization;

[0104] 3) α,β Perform GS balancing;

[0105] The MMSE equilibrium expression is as follows:

[0106]

[0107] in, is the time domain signal vector obtained after MMSE equalization. P is the transmission power, I is the unit matrix, and H is the time domain channel matrix model;

[0108] matrix With vector The relationship is as follows:

[0109]

[0110] in, Reconstruct (·) along the columns into a matrix with M rows and N columns.

[0111] GS balance is shown in Table 1:

[0112] Table 1

[0113]

[0114]

[0115] Where δ is a relaxation parameter used to improve the convergence of the detector for high-order modulation schemes, R, D, L, b n 、T n 、 is an intermediate variable, O M,N represents a zero matrix with M rows and N columns, represents hard decision, tril(·) represents the lower triangular matrix of (·), diag[·] represents the matrix that places the elements of vector (·) on the diagonal, is the GS equalization result, I is the number of iterations, H n Represents r n,α,β The time domain channel matrix, H n The relationship with H is:

[0116]

[0117] If MMSE equalization is performed first and then GS equalization is performed on the result of MMSE equalization, the parameters input for GS equalization include If GS equalization is performed separately, there is no

[0118] Step 6: Perform N-point β-order WFRWHT and N-point 1-α-order WFRFT along the row. Restored to the baseband received signal matrix Y, as follows:

[0119]

[0120] Step 7: Perform parallel / serial conversion on Y to obtain the baseband received signal vector Then perform constellation inverse mapping on y to obtain the transmitted information and complete the communication process.

[0121] Where Y=[y0,y1,…,y M-1 ] T .

[0122] The communication process of the present invention has the following characteristics:

[0123] The baseband received signal vector y and the baseband data signal vector x have the following relationship:

[0124]

[0125] in, is the equivalent channel matrix, is the equivalent noise matrix, The expression is as follows:

[0126]

[0127] Among them, H is the time domain channel matrix model, I M is the unit matrix, P is the row-column interleaver matrix, is the Kronecker product. The expression is as follows:

[0128]

[0129] Where n represents the variance σ 2 discrete Gaussian white noise.

[0130] The WFRFT-WFRWHT-OTFS integrated waveform framework of the present invention can be regarded as an OTFS waveform with α-order WFRFT and β-order WFRWHT as pre-coded waveforms. The equivalent relationship between different waveform modes is as follows: Figure 2 When α = 0, β = 0, the waveform is characterized as an OTFS waveform; when α = 1, β = 0, the waveform is characterized as an SC waveform; when α = 1, β = 1, the waveform is characterized as an OTSM waveform. The relationship between the specific two-dimensional parameters selected for WFRFT-WFRWHT-OTFS and the characterized waveform is shown in Table 2.

[0131] Table 1 Relationship between 2D parameter selection and waveform characterization in the WFRFT-WFRWHT-OTFS framework

[0132]

[0133]

[0134] Example:

[0135] The present invention is applied to communication scenarios with complex channel environments and diverse performance requirements to achieve signal communication. To verify the beneficial effects of the present invention, the present invention conducted simulation experiments:

[0136] The simulation conditions are:

[0137] Number of subcarriers: M = 128

[0138] Number of symbols: N = 32

[0139] Subcarrier mapping mode: 4QAM

[0140] Carrier frequency: 6×10 9 Hz

[0141] Subcarrier spacing: 15kHz

[0142] Channel parameters: Using the EVA model, the delay of each path is [0 30 150 310 370 710 1090 17302510] ns, and the corresponding power attenuation is [0 -1.5 -1.4 -3.6 -0.6 -9.1 -7.0 -12.0 -16.9] dB.

[0143] User moving speed: 150km / h

[0144] Equalization mode: MMSE equalization, GS iterative equalization

[0145] Figure 3 The BER performance curves of the WFRFT-WFRWHT-OTFS waveform are shown for MMSE and GS receivers at a user speed of 150 km / h. It can be seen that under high-speed mobile channel conditions, the BER performance of the WFRFT-WFRWHT-OTFS waveform using a GS receiver is better than that using an MMSE receiver. Furthermore, when using a GS receiver, the WFRFT-WFRWHT-OTFS waveform achieves even better BER performance than OTFS.

[0146] Figure 4 The PAPR performance curves of the WFRFT, WFRWHT-OTFS waveforms are compared at a 10x oversampling factor. The figure shows that the OTFS and OTSM waveforms have the worst PAPR performance, while the WHT-OTFS waveform has better PAPR performance, and the SC waveform has the best PAPR performance. This shows that by adjusting the two-dimensional parameters, the integrated waveform framework proposed in this invention can achieve suitable PAPR performance.

Claims

1. A communication method based on orthogonal time-frequency-space integrated waveform design, characterized in that The specific process of the method is: Step 1: Perform constellation mapping on the serial data to be sent, and then preprocess the constellation mapping result to obtain a baseband data signal vector x with a length of M×N in the fractional domain. Perform serial / parallel conversion on x along the row to obtain a baseband data signal matrix X=[x0,x1,...,x M-1 ] T ; Where M is the number of subcarriers, N is the number of symbols in each column of the signal vector, and the amount of serial data to be sent is (Ml max )×N; The constellation mapping result is pre-processed to obtain the baseband data signal vector x of M×N columns in the fractional domain: First, the serial data to be sent after the constellation mapping result is padded with zeros; Among them, the total length of zero padding is l max ×N;l max is the zero-padded length of each column of the signal vector; Then, the zero-filled x is divided into M signal vectors to obtain the baseband data signal vector of M×N columns in the fractional domain Step 2: Use the X obtained in step 1 to obtain the one-dimensional time domain signal s α,β (t), comprising the following steps: Step 2. Perform N-point α-1 order WFRFT and N-point β-order WFRWHT along the row of X to obtain the data X′ in the time domain, as shown in the following formula: in, is the N-point α-1 order WFRFT transform, is the N-point-β-order WFRWHT transformation; Step 22: Perform DFT along the columns of X′ obtained in step 21 to obtain the frequency domain signal X″, as shown in the following formula: X″=F M X′ Among them, F M is the DFT transform; Step 2: Perform Heisenberg transform on X″ to obtain a one-dimensional time domain signal s α,β (t), specifically: Among them, X″[m,n] represents the element in the mth row and nth column of X″, g tx (t) is the time-domain continuous baseband pulse shaping function, Δf is the subcarrier bandwidth, T is the subsymbol duration, t is time, and j is the imaginary unit; Step 3: Intercept s α,β (t) The signal with a preset length at the end is used as a cyclic prefix, and the cyclic prefix is ​​connected to the signal s α,β (t) before obtaining s' α,β (t), and then use the transmitter to transmit s' α,β (t), thereby obtaining the signal r' received by the receiver α,β (t), as follows: Where h(τ,t) represents the continuous time-varying channel impulse response in the delay-time domain, and n(t) represents the variance σ 2 Continuous Gaussian white noise, τ is the time delay, τ max is the maximum delay; Step 4: The signal r' received by the receiver α,β (t) is converted into the time domain received signal matrix Y′, and Y′ is used to obtain the received signal vector r α,β , including the following steps: Step 4.

1. Remove r' α,β (t) cyclic prefix, and obtain the received signal r α,β (t), r α,β (t) Perform Wigner transform to obtain discrete frequency domain signal Y″[m,n], thereby obtaining the frequency received signal matrix Y″, as shown in the following formula: Among them, g rx (t) is the receiving shaping filter, Y″[m,n] is the element in the mth row and nth column of the frequency domain received signal matrix Y″, υ represents the Doppler shift, [·] * represents the conjugation of [·]; Step 42: Perform IDFT transformation along the columns of Y″ obtained in step 41 to obtain the time domain received signal matrix Y′, thereby obtaining the received signal vector r α,β , specifically: First, perform IDFT along the columns of Y″ obtained in step 4-1 to obtain the time domain received signal matrix Y′: in,[·] H represents the conjugate transpose; Then, use Y′ to obtain the received signal vector r α,β : Step 5: Receive signal vector r α,β Perform equalization to obtain the equalized time domain received signal matrix Step 6: Perform N-point β-order WFRWHT and N-point 1-α-order WFRFT along the row. Restore to baseband received signal matrix Y; Among them, α and β are values ​​between 0 and 1; Step 7: Perform parallel / serial conversion on Y to obtain the baseband received signal vector Then perform constellation inverse mapping on y to obtain the transmitted information and complete the communication process.

2. The communication method based on orthogonal time-frequency-space integrated waveform design according to claim 1, characterized in that: The received signal vector r in step 5 α,β Equalization is performed using MMSE equalization.

3. The communication method based on orthogonal time-frequency-space integrated waveform design according to claim 2, characterized in that: The received signal vector r in step 5 α,β GS equalization is used to achieve equalization.

4. The communication method based on orthogonal time-frequency-space integrated waveform design according to claim 3, characterized in that: The received signal vector r in step 5 α,β Balancing is achieved by: First, use MMSE equalization to receive the signal vector r α,β The MMSE equalization result is obtained by equalization, and then the MMSE equalization result is subjected to GS equalization.

5. The communication method based on orthogonal time-frequency-space integrated waveform design according to claim 4, characterized in that: In step six, Perform N-point β-order WFRWHT and N-point 1-α-order WFRFT along the row. Restored to the baseband received signal matrix Y, as follows:

Images