A method for integrated sensing and communication

Abstract

The present invention discloses an integrated method for joint sensing and communication. First, at the transmitter, baseband data is subjected to weighted fractional Fourier transform and OTFS modulation, and the modulated signal is sent after preprocessing. After the signal passes through the channel and reaches the receiver, a part of the power enters the receiver antenna and is received. After demodulation and decision-making, the data information is restored. Another part of the power is reflected from the surface of the receiver back to the transmitter. The radar antenna part of the transmitter receives the echo signal. After OTFS demodulation and weighted inverse transform, the perception and estimation of the delay and Doppler frequency offset parameters in the channel are completed in the weighted fractional domain. Based on the perception and estimation results, the parameters of the pre-equalization module are set to realize the preprocessing of the new frame signal. The present invention reduces the system hardware cost, improves the frequency band utilization rate, and aids in enhancing the communication performance based on the perception results through the soft / hardware reuse between the sensing and communication systems.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and relates to an integrated method of joint sensing and communication, in particular to an integrated method of joint sensing and communication based on WFRFT-OTFS waveform. Background Technique

[0002] The increasing demand for air interface capabilities in new communication services drives the development of the next-generation mobile communication network (The Sixth Generation of Mobile Communications System, 6G) towards building a multi-dimensional system and direction. In the sixth-generation communication system, in addition to the improvement of traditional communication capabilities, capabilities such as computing, sensing, artificial intelligence, and security will also be provided. Among them, the sensing ability will become an important feature of future mobile communication networks, supporting new services such as driverless and unmanned manufacturing, and accelerating the development of wireless communication technology.

[0003] Sensing and communication are two technologies with different characteristics, and there is a huge gap in traditional concepts. In the 6G mobile communication system, it is expected to use higher frequency bands (millimeter waves and even terahertz), wider bandwidths, and larger-scale antenna arrays, which makes the new generation of communication systems have a working spectrum and hardware structure similar to those of sensing systems. With the development of technologies such as Massive MIMO and software-defined radio, there is a possibility of unifying communication and sensing in terms of waveforms and baseband digital signal processing. The concept of Integrated Sensing and Communication (ISAC) has gradually become popular in the academic community. The ISAC system has both communication and sensing capabilities, and the two sets of functions share the same set of hardware and waveforms, which can simultaneously achieve high-quality communication and high-precision sensing in one system, with characteristics such as improving spectrum efficiency and hardware resource utilization, and reducing application costs. Currently, the research on ISAC systems is mainly divided into two categories: integrated systems based on the sensing system and integrated systems based on the communication system. The integrated system based on the sensing system embeds communication information into the sensing detection signal, and common waveforms include chirp signals, frequency-modulated continuous wave signals, and phase-modulated continuous wave signals, etc. The integrated system based on the communication system adds radar functions at the transmitter of the communication signal, receives and processes the echo of the communication signal to obtain information about the target and the channel, and common waveforms include Single Carrier (SC), Orthogonal Frequency Division Multiplexing (OFDM), Orthogonal Time Frequency Space (OTFS), etc. Among them, the symbols of the OTFS waveform are located in the time-delay Doppler domain, allowing the transmitted signal to directly interact with the delay and Doppler frequency shift, which is exactly consistent with the purpose of sensing. Therefore, the OTFS waveform has become a candidate solution that has attracted much attention in the JSAC system.

[0004] OTFS modulation can be regarded as a generalized precoding scheme for traditional multi-carrier waveforms. On the premise of improving the bit error rate performance, it also maintains good compatibility with traditional multi-carrier modulation systems. However, like traditional multi-carrier modulation technologies, OTFS modulation also has the problem of high peak-to-average power ratio. The peak-to-average power ratio of the OTFS system increases with the increase in the quantization interval number of the Doppler frequency shift in the time-delay Doppler plane, which makes peak signals with relatively large power ratios usually appear in the OTFS system. When the peak power of the signal entering the radio frequency amplifier is greater than the linear dynamic range of the power amplifier, it will cause out-of-band dispersion and in-band interference, resulting in impaired performance of communication and sensing. Therefore, how to suppress the peak-to-average power ratio (PAPR) is one of the important issues that the JSAC system needs to address.

[0005] The literature "On the Performance of Integrated Orthogonal Time Frequency Space Framework based on WFRFT" constructs a two-dimensional parameter-adjustable and integrated fusion waveform framework based on the weighted fractional Fourier transform (WFRFT) and OTFS waveforms, and uses it to achieve high-quality communication in complex channels. Through the flexible selection of two-dimensional parameters, the WFRFT-OTFS waveform framework can be simplified into waveforms such as OTFS, A-OFDM, HC, OFDM, and SC. Compared with single-waveform systems, the WFRFT-OTFS waveform has better trade-off and flexibility. However, this literature does not fully utilize the high flexibility of the WFRFT-OTFS waveform. It only processes the signal at the receiver and ignores the preprocessing of the signal at the transmitter, and the signal processing process is greatly affected by channel noise.

[0006] The literature "Signal Interference Elimination Based on Autoencoder for Integrated Sensing and Communication" establishes an integrated communication and sensing system based on OFDM signals and eliminates interference during transmission to ensure the accuracy of sensing and communication. However, the OFDM waveform considered by this technology is not sensitive enough to time delay, so the sensing performance of signal time delay is not ideal.

[0007] The literature "Joint Radar-Communication With Cyclic Prefixed Single Carrier Waveforms" uses the Cyclic Prefix Single Carrier (CP-SC) waveform to achieve the integration of sensing and communication, and derives an effective algorithm for target range and speed detection / estimation based on the principle of the maximum likelihood algorithm. After verification, the estimation performance of this method is better than that of the method based on traditional channel estimation, and the cyclic prefix single carrier has an obvious low peak-to-average power ratio advantage compared with the integrated sensing and communication system using multi-carriers. However, the SC waveform lacks sensitivity to the Doppler frequency offset, which makes the frequency offset estimation performance of this system not ideal enough.

[0008] The literature "On the Effectiveness of OTFS for Joint Radar Parameter Estimation and Communication" uses the OTFS waveform to implement an integrated joint sensing and communication system, and derives an approximate maximum likelihood (ML) estimation algorithm based on the OTFS waveform. After verification, the performance of the method using the OTFS waveform for radar parameter estimation is very close to that of the method using the traditional chirp waveform, and it is significantly better than the channel estimation algorithm at the receiving end in terms of estimation, achieving accurate and effective estimation. This scheme only uses one waveform as the estimation scheme and lacks flexibility in the face of a complex channel environment. Summary of the Invention

[0009] Aiming at the above-mentioned existing technologies, the technical problem to be solved by the present invention is to provide an integrated joint sensing and communication method based on the weighted fractional Fourier transform and the orthogonal time-frequency-space waveform. By introducing adjustable weighted transformation coefficients, on the premise of maintaining the high resolution of the orthogonal time-frequency-space waveform in the time-delay Doppler domain, the peak-to-average power ratio of the signal is effectively reduced, and the waveform and hardware of the communication system and the sensing system are multiplexed through signal preprocessing at the transmitting end, improving the performance of the communication system.

[0010] To solve the above technical problems, an integrated joint sensing and communication method of the present invention includes:

[0011] Step 1: Perform constellation mapping and serial / parallel conversion on the N c -bit serial data to be transmitted. After mapping, the baseband data x[k, l] with N rows and M columns is obtained, where x[k, l] is located in the weighted fractional domain;

[0012] Step 2: Map the baseband data in the fractional domain to the time-delay Doppler domain by performing an N-point weighted fractional Fourier transform on x[k, l] column by column, obtaining the weighted symbol x[k′, l′];

[0013] Step 3: Perform a two-dimensional inverse symplectic Fourier transform on the data in the time-delay Doppler domain to map the data x[k′, l′] in the time-delay Doppler domain to the data X[n, m] in the time-frequency domain;

[0014] Step 4: Perform a Heisenberg transform on X[n, m] to obtain a one-dimensional time-domain signal

[0015] Step 5: Perform pre-equalization on the signal to suppress the self-interference of the signal in the channel, obtaining the pre-equalized transmitted signal where Q is the pre-equalization matrix, Q = H * / (Η = H * ) = H -1 , and H is the equivalent matrix of the NM = NM-dimensional time-domain channel;

[0016] Step 6: The signal s(t) is transmitted from the transmitter to the target receiver. After passing through the channel, the receiver receives the signal r(t). At the same time, a part of the signal is reflected by the target and returns to the transmitter position. The radar receiver on the transmitter detects the echo signal

[0017] Step 7: In the transmitter, perform OTFS demodulation and the corresponding weighted inverse fractional Fourier transform on the echo signal to obtain the restored sample signal y echo , calculate the estimated value of the main path of the received signal Then estimate the parameter θ according to the minimization of the log-likelihood function. The parameter θ includes the main path gain echo time delay and echo Doppler frequency offset The estimation process includes: First, according to the range of the time delay and Doppler frequency offset of the echo, construct a two-dimensional integer time-delay Doppler network, calculate the log-likelihood function at each grid node respectively, find the node with the minimum log-likelihood function, and perform a secondary estimation on the fractional time delay and Doppler frequency offset within a set range near the node with the minimum log-likelihood function to obtain the accurate value of the estimated parameter; In the receiver, perform OTFS demodulation and the corresponding weighted inverse fractional Fourier transform on the received signal to restore the transmitted information;

[0018] Step 8: The channel information estimated based on echo detection in the transmitter is used as known information for the preprocessing of the next frame of signal. In the receiver, the decision-reduced signal is output to complete the communication process.

[0019] Furthermore, the weighted symbol x[k′, l′] is specifically:

[0020]

[0021] where α is the order of the weighted transform, denotes the i-th Fourier transform of the symbol x[k, l], and the weighted coefficient of WFRFT is expressed as:

[0022]

[0023] Furthermore, mapping the data x[k′, l′] in the time-delay Doppler domain to the data X[n, m] in the time-frequency domain described in Step 3 is specifically:

[0024]

[0025] Furthermore, the one-dimensional time-domain signal is specifically:

[0026]

[0027] where g tx (t) is a continuous baseband pulse shaping function in the time domain, Δf is the subcarrier bandwidth, and T is the sub-symbol duration.

[0028] Furthermore, the equivalent matrix H of the NM×NM-dimensional time-domain channel is specifically:

[0029]

[0030] where is an NM×NM-dimensional permutation matrix depending on the time delay τ i , and is an NM×NM-dimensional frequency offset matrix depending on the Doppler frequency offset υ i .

[0031] Furthermore, the echo signal r(t) of the signal described in Step 6 is specifically:

[0032]

[0033] where P represents the number of paths in the channel, h(υ, τ) represents the channel response function, and τ and υ are the path delay and Doppler frequency offset respectively;

[0034]

[0035] where They are the channel fading, Doppler frequency offset, and time delay of the path through which the echo signal passes, respectively. In the echo channel, the time delay of the main path is approximately twice the time delay of the main channel through which the signal is transmitted, and they have the same Doppler frequency offset.

[0036] Furthermore, the sample signal y is obtained. echo Specifically:

[0037] Write the N×M matrix of the transmitted symbol and the received sample as an N c ×1-dimensional column vector, and the input-output relationship matrix of the echo is:

[0038]

[0039] where W α is the matrix form of the WFRFT, expressed as:

[0040]

[0041] F N represents the N-point Fourier transform, Θ OTFS is the transformation matrix of OTFS modulation, which includes two-dimensional symplectic Fourier transform and one-dimensional Heisenberg transform. is the OTFS demodulation matrix, which includes symplectic Fourier transform and Wigner transform, where the Wigner transform is the inverse operation of the Heisenberg transform.

[0042] Furthermore, the estimated value of the main path of the received signal Specifically:

[0043]

[0044] where represents the estimated value of the echo channel matrix, and x is the transmitted baseband data.

[0045] Furthermore, minimizing the log-likelihood function is specifically:

[0046]

[0047] where ρ(a,b) represents the correlation between vectors and satisfies:

[0048]

[0049] Advantages of the present invention: The present invention uses the WFRFT-OTFS waveform to build a new integrated sensing and communication method. On the one hand, by introducing adjustable weighting coefficients, the integration of multiple waveforms is achieved based on a set of waveform frameworks. Compared with traditional JSAC systems using multi-carrier systems such as OTFS / OFDM, the integrated sensing and communication system using the WFRFT-OTFS waveform has advantages such as lower PAPR and higher flexibility. Compared with the JSAC system using SC, the WFRFT-OTFS retains the sensitivity characteristics of the OTFS waveform to the delay and Doppler network, and has obvious advantages in high-speed complex channels. On the other hand, based on the sensing results, this patent proposes a preprocessing method at the transmitter end based on the WFRFT-OTFS waveform, adopting the idea of zero-forcing (ZF) pre-equalization, which significantly improves the performance of the communication system. The present invention reduces the system hardware cost, improves the frequency band utilization rate through the soft / hardware reuse between the sensing and communication systems, and aids in enhancing the communication performance based on the sensing results. The specific effects of the system are reflected by simulating and comparing the performance of target sensing, the peak-to-average power ratio of the signal, and the bit error rate performance. The present invention is applied to communication systems with a relatively high moving speed of the target and relatively high requirements for communication quality, such as vehicle-to-everything (V2X) and unmanned aerial vehicle (UAV) communication. Description of the Drawings

[0050] Figure 1 is the integrated sensing and communication system framework based on the WFRFT-OTFS waveform;

[0051] Figure 2 is the integrated communication and sensing transmission channel model;

[0052] Figure 3 is the two-dimensional delay-Doppler grid;

[0053] Figure 4 is the speed estimation performance;

[0054] Figure 5 is the distance estimation performance;

[0055] Figure 6 is the comparison of bit error rate performance;

[0056] Figure 7 is the comparison of peak-to-average power ratio performance. Detailed Embodiment

[0057] The present invention will be further described below in conjunction with the drawings of the specification and embodiments.

[0058] Based on the integrated sensing and communication method using the WFRFT-OTFS waveform proposed by the present invention, the system block diagram is as Figure 1 shown, and the specific implementation process of the present invention is as shown in the following steps:

[0059] Step 1: For the N-bit serial data to be transmitted, perform constellation mapping and serial-to-parallel conversion. After mapping, baseband data x[k, l] with N rows and M columns is obtained, where x[k, l] is in the weighted fractional domain. c Step 2: By performing an N-point weighted fractional Fourier transform on x[k, l] for each column, map the baseband data in the fractional domain to the time-delay Doppler domain, obtaining the weighted symbol x[k′, l′], which is expressed as:

[0060] where α is the order of the weighted transform,

[0061]

[0062] and represents performing the i-th Fourier transform on the symbol x[k, l]. The weighted coefficient of the WFRFT can be expressed as:

[0063]

[0064] The entire time-delay Doppler plane is divided into a discrete network Γ with time delay τ and Doppler shift υ as the unit dimensions, as shown in (3). Here, N and M represent the number of sub-symbols and sub-carriers in OTFS modulation respectively, and 1 / NT and 1 / MΔf represent the quantization units of the time delay τ and Doppler shift υ.

[0065] Γ = {(k′ / NT, l′ / MΔf), k = 0,..., N - 1, l = 0,..., M - 1} (3)

[0066] Step 3: Perform a two-dimensional inverse symplectic Fourier transform (ISFFT) on the data in the time-delay Doppler domain to map the data x[k′, l′] in the time-delay Doppler domain to the time-frequency domain data X[n, m]:

[0067]

[0068] Step 4: Perform a Heisenberg transform on X[n, m] to obtain a one-dimensional time-domain signal s(t). The Heisenberg transform is regarded as a generalized form of the Fourier transform.

[0069]

[0070] where g tx (t) is the continuous baseband pulse shaping function in the time domain.

[0071] Step 5: Based on the channel information obtained by perception estimation, including the path delay τ and Doppler frequency offset υ, a channel matrix model can be constructed at the transmitter end, which is expressed as:

[0072]

[0073] H represents a NM×NM dimensional complex channel matrix, is determined by the time delay τ i The NM×NM dimensional permutation matrix, is determined by the Doppler frequency deviation υ i The NM×NM-dimensional frequency offset matrix of the signal is Pre-equalization suppresses the self-interference of the signal in the channel and improves the accuracy of the communication system. In this method, the ZF criterion is used as the pre-processing method, and the matrix model of the communication process is considered as follows:

[0074]

[0075] Where Q is the pre-equalization matrix, H is the equivalent matrix of the time domain channel, based on the principle of the ZF pre-equalization algorithm, let:

[0076]

[0077] Then the pre-equalization matrix Q = H * / (H×H * )=H -1 The transmitted signal after pre-equalization is expressed as:

[0078]

[0079] Step 6: The signal s(t) is transmitted to the target receiver via the transmitter. The process is as follows: Figure 2 As shown. After passing through the channel, the signal received by the receiver is:

[0080]

[0081] Where P represents the number of paths in the channel, h(υ,τ) represents the channel response function, and τ and υ are the path delay and Doppler frequency deviation, respectively. In addition, part of the signal is reflected by the target and returns to the transmitter. The radar receiver on the transmitter detects the echo signal:

[0082]

[0083] They are the channel fading, Doppler frequency deviation and delay of the path that the echo signal passes through, and in the echo channel, the delay of the main path can be approximated to be twice the delay of the main channel of signal transmission, and the two have the same Doppler frequency deviation. Obviously, based on the echo signal and the transmitted signal, the speed, position and other information of the target can be effectively estimated, and then the parameters of the entire transmission channel can be perceived.

[0084] Step 7: Process the echo signal and the received signal at the transmitter radar end and the receiver respectively. Among them, in the transmitter, perform OTFS demodulation and the corresponding weighted inverse fractional Fourier transform on the echo signal to obtain the restored sample signal. Write the N×M matrix of the transmitted symbol and the received sample as an N c ×1-dimensional column vector, and the input-output relationship matrix of the echo can be obtained as:

[0085]

[0086] where W α is the matrix form of the WFRFT, expressed as:

[0087]

[0088] F N represents the N-point Fourier transform, Θ OTFS is the transformation matrix of OTFS modulation, including the two-dimensional symplectic Fourier transform and the one-dimensional Heisenberg transform. is the OTFS demodulation matrix, including the symplectic Fourier transform and the Wigner transform, where the Wigner transform is the inverse operation of the Heisenberg transform.

[0089] In the radar system, the parameters θ to be estimated mainly include the path gain echo time delay and echo Doppler frequency offset In the initial research, we give priority to considering the model with a large direct main path. Under this premise, the remaining paths are regarded as interference. According to the input-output relationship, the estimated value of the main path of the received signal is:

[0090]

[0091] represents the estimated value of the echo channel matrix. According to the maximum likelihood estimation principle, the logarithmic likelihood function to be minimized is given as:

[0092]

[0093] where x is the transmitted baseband data, y echo is the demodulated echo received data, and θ is the target parameter to be estimated. Obviously, estimating three unknowns simultaneously, the computational complexity is unacceptable to us. In the detection of the target speed and distance, we give priority to estimating the echo time delay and Doppler frequency offset . Therefore, improve formula (15). Let ρ(a,b) represent the correlation between vectors, then there is:

[0094]

[0095] Let Then minimizing the log-likelihood function can be rewritten as:

[0096]

[0097] To improve the efficiency of estimation, the process of channel parameter estimation is simplified to finding the point that minimizes the log-likelihood function \(l(y echo ||\theta,x)\) on a discrete two-dimensional time-delay Doppler grid, and an integer-fraction two-layer estimation is adopted to further reduce the number of points to be calculated. As Figure 3 shown, based on the ranges of the time delay and Doppler frequency shift of the echo, an integer time-delay Doppler network is first constructed, the log-likelihood function at each integer node is calculated respectively, the node with the minimum log-likelihood function is found, and then a secondary estimation is performed on the fractional time delay and Doppler frequency shift near the node with the minimum log-likelihood function value to obtain the exact value of the estimated parameter.

[0098] In the receiver, the received signal is demodulated by OTFS and the corresponding weighted inverse fractional Fourier transform is performed to restore the transmitted information.

[0099] Step Eight: The channel information estimated based on echo detection in the transmitter is used as known information for the preprocessing of the next frame of signal. The output decision restores the signal in the receiver to complete the communication process.

[0100] The invention will be further described below in conjunction with specific parameters and simulation results.

[0101] In the simulation, the channel is a time-delay Doppler channel with a main path. It is assumed that the main path time delay of the echo channel is twice that of the transmit channel, and the moving direction of the target is linear motion. The data length of the transmitted signal is 128 bit, the baseband modulation method is 4-QAM, the sampling frequency is 1 MHz, and the carrier frequency is 26 GHz. The maximum moving speed of the sensed target is 500 m / s, and the maximum distance between the transmitter and the receiver is 2 km.

[0102] For the JSAC system based on the WFRFT-OTFS waveform, the performance such as the sensing mean square error, bit error rate, and peak-to-average power ratio when different weighting parameters are selected is shown in the following simulation results:

[0103] Figure 4 and Figure 5They are the mean square errors of using the WFRFT-OTFS waveform to sense the velocity and distance of the target under different weighting coefficients. It can be seen from the figure that when the JSAC framework built by the WFRFT-OTFS waveform is sensing, its performance varies to some extent with the change of the weighting transformation coefficient. However, this gap in sensing performance is not obvious, and the sensing systems with different weighted parameter waveforms have the same lower limit of mean square error under high signal-to-noise ratio conditions.

[0104] Figure 6 In [reference], the bit error rate differences between the zero-forcing pre-equalization at the transmitter based on the WFRFT-OTFS waveform proposed in this patent and the traditional zero-forcing equalization method at the receiver are compared; Figure 7 Then, the peak-to-average power ratio of the WFRFT-OTFS waveform under different coefficients is simulated. In the simulation, the total number of symbols N of the signal c is 128 bit, the number of sampling points M on the time-delay axis is 16, the number of sampling points N on the Doppler axis is 8, and 4-fold oversampling is adopted when calculating the PAPR. Obviously, by adjusting the weighting coefficient α, the peak-to-average power ratio of the signal can be effectively reduced, while affecting the bit error rate performance of the signal. When α = 0, the WFRFT-OTFS waveform is simplified to the OTFS waveform. At this time, the peak-to-average power ratio of the signal is relatively high, but the OTFS has a high time-delay Doppler domain resolution and has certain advantages in subsequent signal processing. When α = 1, the WFRFT-OTFS waveform is simplified to a single carrier. At this time, the signal has a relatively low peak-to-average power ratio performance. On the other hand, although the traditional receiver equalization method can better suppress the self-interference of the signal, it will amplify the noise part in the signal, resulting in a loss of bit error rate. By adopting the pre-equalization method at the transmitter in this patent, the amplification of noise in the processing process is avoided, and relatively better bit error rate performance is obtained.

[0105] Therefore, the integrated communication and sensing method based on the WFRFT-OTFS waveform proposed in this invention realizes the further optimization of performance by accurately suppressing the peak-to-average power ratio and bit error rate through the weighting coefficient α compared with the JSAC system using waveforms such as OTFS / OFDM / SC. On the other hand, by using the sensing and estimation results of the target motion and position, zero-forcing pre-equalization is performed on the signal at the transmitter, significantly improving the transmission quality of the communication system.

Claims

1. An integrated method for joint sensing and communication, characterized in that, Including: Step 1: For the N-bit serial data to be transmitted c perform constellation mapping and serial / parallel conversion, and after mapping, obtain baseband data x[k, l] with N rows and M columns, where x[k, l] is located in the weighted fractional domain; Step 2: By performing an N-point weighted fractional Fourier transform on x[k, l] for each column, map the baseband data in the fractional domain to the time-delay Doppler domain to obtain the weighted symbol x[k′, l′]; Step 3: Perform an inverse symplectic Fourier transform in two dimensions on the data in the time-delay Doppler domain to map the data x[k′, l′] in the time-delay Doppler domain to the data X[n, m] in the time-frequency domain; Step 4: Perform a Heisenberg transform on X[n,m] to obtain a one-dimensional time-domain signal Step 5: Perform pre - equalization on the signal to suppress the self - interference of the signal in the channel and obtain the pre - equalized transmitted signal where Q is the pre - equalization matrix, and Q = H * / (Η×H * ) = H -1 , and H is the equivalent matrix of the time - domain channel with dimensions NM×NM; Step 6: The signal s(t) is transmitted from the transmitter to the target receiver. After passing through the channel, the receiver receives the signal r(t). Meanwhile, a part of the signal is reflected by the target and returns to the position of the transmitter. The radar receiver on the transmitter detects the echo signal Step 7: In the transmitter, perform OTFS demodulation and the corresponding weighted inverse fractional Fourier transform on the echo signal to obtain the restored sample signal y echo , and calculate the estimated value of the main path of the received signal Then, estimate the parameter θ according to the minimization of the log-likelihood function. The parameter θ includes the main path gain echo time delay and echo Doppler frequency offset The estimation process includes: First, construct a two-dimensional integer time-delay Doppler network according to the ranges of the time delay and Doppler frequency offset of the echo, calculate the log-likelihood function at each grid node respectively, find the node with the minimum log-likelihood function, and perform a secondary estimation on the fractional time delay and Doppler frequency offset within a set range near the node with the minimum log-likelihood function to obtain the accurate value of the estimated parameter; In the receiver, perform OTFS demodulation and the corresponding weighted inverse fractional Fourier transform on the received signal to restore the transmitted information; Step 8: The channel information estimated based on echo detection in the transmitter is used as known information for preprocessing the next frame of signals. In the receiver, the signal obtained by decision restoration is output to complete the communication process.

2. The integrated method for joint sensing and communication according to claim 1, wherein: The weighted symbol x[k′, l′] is specifically: where α is the order of the weighted transform, denotes the i-th Fourier transform of the symbol x[k, l], and the weighted coefficient of the WFRFT is expressed as:

3. The integrated method of joint sensing and communication according to claim 1, characterized in that: The mapping of the data x[k′, l′] in the time-delay Doppler domain to the data X[n, m] in the time-frequency domain described in Step 3 is specifically:

4. The integrated method for joint sensing and communication according to claim 1, wherein: The one-dimensional time-domain signal described in Step 4 Specifically: where g tx (t) is a continuous baseband pulse shaping function in the time domain, Δf is the subcarrier bandwidth, and T is the sub-symbol duration.

5. The integrated method for joint sensing and communication according to claim 1, wherein: The equivalent matrix H of the NM×NM-dimensional time-domain channel is specifically: wherein, is an NM×NM dimensional permutation matrix depending on the time delay τ i , and is an NM×NM dimensional Doppler frequency offset matrix depending on the Doppler frequency offset υ i .

6. The integrated method for joint sensing and communication according to claim 1, wherein: The echo signal of the signal r(t) described in Step 6 Specifically: where P represents the number of paths in the channel, h(υ, τ) represents the channel response function, and τ and υ are the path delay and Doppler frequency shift respectively; wherein, are respectively the channel fading, Doppler frequency offset and time delay of the path through which the echo signal passes. In the echo channel, the time delay of the main path is approximately twice the time delay of the main signal transmission channel, and they have the same Doppler frequency offset.

7. The integrated method for joint sensing and communication according to claim 1, characterized in that: Obtain the sample signal y echo Specifically: Write the N×M matrix of transmission symbols and received samples as an N c ×1 dimensional column vector, and the input-output relationship matrix of the echo is obtained as follows: where W α is the matrix form of the WFRFT, expressed as: F N represents the Fourier transform of N points, Θ OTFS is the transform matrix of OTFS modulation, which includes the two-dimensional symplectic Fourier transform and the one-dimensional Heisenberg transform, is the OTFS demodulation matrix, which includes the symplectic Fourier transform and the Wigner transform, where the Wigner transform is the inverse operation of the Heisenberg transform.

8. The integrated method for joint sensing and communication according to claim 1, characterized in that: Estimated value of the main path of the received signal Specifically: Among them, represents the estimated value of the echo channel matrix, and x is the transmitted baseband data.

9. The integrated method of joint sensing and communication according to claim 1, wherein: Minimizing the log-likelihood function is specifically: where ρ(a, b) represents the correlation between vectors and satisfies:

Images