FM-SC-OFDM waveform-based terahertz communication sensing integrated method

Abstract

The invention belongs to the technical field of communication and sensing integration, and particularly relates to a terahertz communication and sensing integration method based on FM-SC-OFDM waveforms. According to the method, the DFT spread spectrum technology is utilized, the PAPR of an FM-OFDM instantaneous frequency OFDM sequence is reduced, the instantaneous frequency / phase sequence property is optimized, the phase demodulation error is relieved, and finally, the inherent communication and sensing performance is further enhanced under the condition that the waveform constant envelope and the frequency offset resistance property are inherited in a lossless mode, so that a complete ISAC architecture is provided. The FM-SC-OFDM is proposed by taking SC-OFDM as an instantaneous frequency sequence, the phase demodulation error effect faced by the FM-OFDM is relieved, the BER performance and the perception precision performance are broken through, and a novel integrated waveform which can avoid the power amplifier nonlinear effect and has strong frequency offset robustness is provided for a THz ISAC system.

Description

Technical Field

[0001] The present invention belongs to the technical field of synaesthesia integration, and in particular relates to a terahertz communication perception integration method based on FM-SC-OFDM waveform. Background Art

[0002] ISAC (Integrated Signal Acoustic Response) is one of the important technologies in the sixth generation (6G) mobile communication network. It gives the existing communication network the ability to perceive by sharing spectrum and hardware resources, providing a key foundation for the intelligent integration of communication and radar. THz technology further empowers ISAC, providing Tbps communication rate and extremely high perception resolution through extremely large bandwidth. In recent years, ISAC waveform design has received widespread attention, among which orthogonal frequency division multiplexing (OFDM) is the most commonly used solution. However, OFDM waveform has two serious problems in THz system: (1) OFDM has an extremely high peak-to-average power ratio, and the high power amplifier (HPA) must use power backoff to avoid clipping distortion of the transmitted signal, which directly limits the coverage range and equipment energy efficiency of THz ISAC; (2) THz system has a serious frequency offset effect, and OFDM's orthogonality is easily destroyed by frequency offset, resulting in deterioration of BER performance. These problems limit the direct application of OFDM in THz ISAC system and prompt researchers to explore other solutions.

[0003] Constant envelope OFDM (CE-OFDM) achieves 0dB PAPR by modulating the OFDM signal onto the phase of a constant envelope carrier, making it a candidate for THz ISAC systems. However, the frequency offset issue remains unresolved. To address this, researchers have recently proposed a new waveform, FM-OFDM, which frequency modulates the OFDM signal before performing phase modulation (also known as accumulation). This means that CE-OFDM carries the OFDM signal at the instantaneous phase, while FM-OFDM carries it at the instantaneous frequency.

[0004] Compared to instantaneous phase, the impact of frequency offset on instantaneous frequency is greatly reduced. For example, constant carrier frequency offset (CFO) causes a constant offset effect on the OFDM sequence at the instantaneous frequency. By reserving a DC null carrier, this effect will be completely eliminated. Therefore, FM-OFDM not only has extremely strong frequency offset robustness, but also retains the advantage of constant envelope. However, the high PAPR effect of the OFDM signal at the instantaneous frequency still exists, which can easily lead to severe phase jumps and phase demodulation errors, and greatly limit its upper limit of synaesthesia performance. Since FM-OFDM is relatively new, this issue has not been studied in FM-OFDM. Based on this problem, the present invention uses DFT spread spectrum technology to optimize the phase properties of FM-OFDM, alleviate phase demodulation errors, and improve both the communication and perception performance of FM-OFDM.

[0005] Similar phase demodulation error issues also exist in CE-OFDM. Typical solutions include peak clipping, companding, selective mapping, or DFT spreading of the phase OFDM sequence. Peak clipping and companding inevitably introduce nonlinear distortion, while selective mapping is too complex. In contrast, DFT spreading successfully mitigates the phase jump problem with acceptable complexity and a linear, lossless transformation. Summary of the Invention

[0006] To address the above problems, the present invention utilizes DFT spread spectrum technology to reduce the PAPR of the FM-OFDM instantaneous frequency OFDM sequence, optimize the instantaneous frequency / phase sequence properties, and alleviate phase demodulation errors. Ultimately, while losslessly inheriting the waveform's constant envelope and frequency offset resistance properties, it further enhances its inherent communication and perception performance, thereby proposing a complete ISAC architecture.

[0007] The technical solution of the present invention is:

[0008] The integrated terahertz communication and sensing method based on FM-SC-OFDM waveform includes:

[0009] The integrated transmitter generates a communication-aware integrated signal by performing M-QAM modulation on the source bits to obtain a QAM sequence. This is then transformed through DFT, zero-padding, conjugate symmetric mapping, and IFFT to obtain a real-number SC-OFDM sequence. The SC-OFDM is frequency modulated to obtain a phase sequence, and finally, phase modulation is performed to obtain an FM-SC-OFDM waveform, i.e., the communication-aware integrated signal.

[0010] The communication receiver receives the FM-SC-OFDM waveform signal and demodulates it to obtain the received bits. Specifically, the received signal is first subjected to cyclic prefix removal, channel estimation, and equalization. Then, phase demodulation is performed using an inverse tangent machine and deconvolution. The instantaneous frequency SC-OFDM sequence is extracted using cyclic differencing. Finally, SC-OFDM demodulation is performed to obtain the received bits.

[0011] The perception receiver extracts target distance and speed information based on the perceived FM-SC-OFDM waveform signal. Specifically, it performs 2D-FFT processing on the acquired FM-SC-OFDM integrated frame containing multiple symbols to extract target distance and speed information.

[0012] Furthermore, in the integrated transmitter, the source bits are defined as being modulated by M-QAM to obtain Dot symbol sequence , then in the integrated transmitter, The specific processing process is as follows: Perform DFT operation to obtain its frequency domain sequence ,Will Perform conjugate symmetric frequency domain mapping to obtain a frequency domain sequence , ,in represent The zero vector of the point is used to achieve oversampling, and the oversampling multiple is recorded as ,in Represents the effective number of subcarriers, Represents the number of IFFT points. Perform IFFT to obtain the instantaneous frequency SC-OFDM sequence and normalize its power:

[0013] ,

[0014] For instantaneous frequency series Perform frequency modulation to obtain instantaneous phase sequence :

[0015] ,

[0016] in is the modulation index;

[0017] Through Perform phase modulation to obtain an FM-SC-OFDM signal with a constant amplitude of A:

[0018] .

[0019] Furthermore, in the communication receiver, the baseband received signal after cyclic prefix removal, channel estimation, and equalization is defined as:

[0020] ,

[0021] in, represents additive white Gaussian noise (AWGN), represents the normalized frequency deviation, and represent the analog frequency offset and subcarrier spacing respectively, Represents the initial phase deviation constant. In a quasi-static channel environment, is considered as a constant within the symbol period; the baseband received signal is reconstructed as:

[0022] ,

[0023] in represents the phase noise introduced by AWGN, represents the frequency deviation factor;

[0024] The process of phase demodulation of the baseband received signal is:

[0025] ,

[0026] in, represents the imaginary and real parts, represents the inverse tangent machine, Represents the phase unwinder, which is used to identify phase jump values ​​exceeding the threshold point, and the phase sequence from that point onwards is compensation to relieve The output phase is blurred, represents the noise introduced by the unwinding error, Mathematically expressed as:

[0027] ,

[0028] in,

[0029] ,

[0030] definition is a cyclic difference function, Do cyclic differencing:

[0031] ,

[0032] in Represents the modulo-N operation, so the received SC-OFDM sequence is expressed as:

[0033] ,

[0034] Will After the sequence completes FFT, demapping, and IDFT, the received QAM sequence is:

[0035] ,

[0036] in represents the original transmitted QAM symbol sequence, and Corresponding to and The noise term introduced.

[0037] Furthermore, in the perceptual receiver, the definition of FM-SC-OFDM integrated frame includes symbols, and the echo signal obtained by the sensing receiver after being reflected by the target is:

[0038] ,

[0039] in , , , , , as well as They represent subcarrier spacing, symbol period, echo delay, Doppler shift, rectangular window function, speed of light, and carrier frequency respectively. represents the frequency domain matrix of the integrated frame, and Represent the subcarrier index and symbol index respectively; the echo signal is converted to the frequency domain as follows:

[0040] ,

[0041] in, Represents the frequency domain matrix of Gaussian white noise in the time domain. and Doing matrix dot division yields a matrix containing delay and Doppler information. ,in Represents the frequency domain matrix of the signal after threshold processing, represents the amplitude threshold, It can be expressed as

[0042] ,

[0043] Perform IFFT and FFT on the distance dimension and velocity dimension respectively, and we get:

[0044]

[0045] in and Represents the index of distance and speed information respectively, and Represents the number of points of IFFT and FFT respectively. For a single target scene, and represent The relationship between the index of the maximum value and the distance and speed information of the target is:

[0046] .

[0047] The beneficial effects of the present invention are as follows: the present invention proposes FM-SC-OFDM using SC-OFDM as the instantaneous frequency sequence, alleviates the phase demodulation error effect faced by FM-OFDM, breaks through the BER performance and perception accuracy performance, and provides a new integrated waveform for the THz ISAC system that can avoid the nonlinear effect of the power amplifier and has strong frequency deviation robustness. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1Schematic diagram of the ISAC architecture based on FM-SC-OFDM of the present invention;

[0049] Figure 2 This is a schematic diagram of the BER-SNR curve under nonlinear power amplifier;

[0050] Figure 3 It is a schematic diagram of the MSE and CRLB curves of distance estimation;

[0051] Figure 4 It is a schematic diagram of BER-MI and CRLB-MI curves under fixed SNR;

[0052] Figure 5 This is a schematic diagram of the THz ISAC experimental scene setup. DETAILED DESCRIPTION

[0053] The present invention will be further described in detail below with reference to the accompanying drawings.

[0054] The ISAC architecture based on FM-SC-OFDM of the present invention is as follows Figure 1 As shown, compared with traditional FM-OFDM, the difference lies in the real SC-OFDM modulation and demodulation module marked in the red box, which specifically includes:

[0055] Integrated transmitter:

[0056] The bit stream is first modulated by M-QAM to obtain power normalization Dot symbol sequence .right Perform DFT operation to obtain its frequency domain sequence ,Will Perform conjugate symmetric frequency domain mapping to obtain a frequency domain sequence This makes the SC-OFDM obtained by the subsequent IFFT real-valued, and the relationship between the two is . Let the number of effective subcarriers be , the number of IFFT points is , the oversampling multiple is recorded as . Note that the first subcarrier, that is, the DC subcarrier is set as a null carrier. Perform IFFT to obtain the instantaneous frequency SC-OFDM sequence and normalize its power:

[0057] (1)

[0058] For instantaneous frequency series Perform frequency modulation to obtain instantaneous phase sequence , where is the modulation index (MI).

[0059] (2)

[0060] Finally, through Perform phase modulation to obtain an FM-SC-OFDM signal with a constant amplitude of A

[0061] (3)

[0062] Communication receiving end:

[0063] like Figure 1 As shown, after CP removal, the baseband received signal can be expressed as

[0064] (4)

[0065] in represents additive white Gaussian noise (AWGN), represents the normalized frequency deviation ( and represent the analog frequency offset and subcarrier spacing, respectively), Indicates the initial phase deviation constant. In a quasi-static channel environment, Caused by CFO, it is considered constant within the symbol period. The received signal is further reconstructed as

[0066] (5)

[0067] in represents the phase noise introduced by AWGN, represents the frequency deviation factor. The operation of the phase demodulator can be expressed as

[0068] (6)

[0069] in represents the imaginary and real parts, represents the inverse tangent machine, Represents the phase unwinder, which identifies the phase jump value exceeding the threshold point, and the phase sequence from that point onwards is compensation to relieve The output phase is blurred. Mathematically, It can be expressed as

[0070] (7)

[0071] in,

[0072] (8)

[0073] represents the noise introduced by the unwinding error (which will be analyzed in detail later). Do circular differencing,

[0074] (9)

[0075] in represents the modulo-N operation, and the received SC-OFDM sequence can be expressed as

[0076] (10)

[0077] It can be seen that the cyclic difference will shift the initial phase Completely eliminate, and also the frequency deviation factor Convert to a constant term , this constant term is located at the DC null carrier by the subsequent FFT operation and will not affect the useful information. Finally, after completing SC-OFDM demodulation, that is, FFT, demapping and IDFT, the received QAM sequence can be expressed as

[0078] (11)

[0079] in and They are and The noise term introduced.

[0080] Perception receiving end:

[0081] Consider a The integrated frame of symbols, which is reflected by a target in the environment and captured by the sensing receiver, can be expressed as

[0082] (12)

[0083] in , , , and They represent subcarrier spacing, symbol period, echo delay, Doppler shift and rectangular window function respectively. represents the frequency domain matrix of the integrated frame, where and Represent the subcarrier index and symbol index respectively. Converting the above formula to the frequency domain, we can get:

[0084] (13)

[0085] Will and Doing matrix dot division can get a matrix containing delay (distance) and Doppler (speed) information.

[0086] (14)

[0087] in It is called frequency domain equivalent noise. Due to the complex frequency domain properties of FM-SC-OFDM, its spectrum often contains frequency points with very small values. The frequency domain equivalent noise at the corresponding position will be severely amplified, causing a sharp drop in estimation accuracy. For this reason, there is an adjusted version of the 2D-FFT algorithm for CE-OFDM in the industry, which is also applicable here. When dividing the frequency domain matrix, Replace with , The definition is as follows

[0088] (15)

[0089] in This can directly reduce the originally severely amplified noise to zero and obtain the adjusted distance-speed information matrix Then perform IFFT and FFT on the distance dimension and velocity dimension respectively, and we can get

[0090] (16)

[0091] in and Represent the indexes of distance and speed information respectively, and and Represents the number of points of IFFT and FFT respectively. For a single target scenario, and represent The relationship between the index of the maximum value and the distance and speed information of the target is as follows

[0092] (17)

[0093] Analysis of FM-SC-OFDM ISAC performance: Based on the FM-SC-OFDM ISAC architecture described in this paper, the theoretical lower bound for BER and the expression for perceptual accuracy (CRLB) of FM-SC-OFDM are derived. The important conclusion is that a larger MI leads to better theoretical BER and perceptual accuracy. Subsequently, the specific mechanism for extending the upper bound of the MI value for SC-OFDM sequences at the instantaneous frequency, as well as its robustness to phase demodulation errors, are analyzed. Specifically, the following aspects are discussed:

[0094] BER theoretical lower bound:

[0095] Based on formula (11), ignoring the phase demodulation error, that is, , and Approximately Gaussian white noise, the theoretical BER lower bound of FM-SC-OFDM under AWGN is

[0096] (18)

[0097] in is the Gaussian-Q function, represents the modulation index, represents the time domain signal-to-noise ratio of the received signal, , and They represent the amplitude of the received signal, the AWGN of the channel, and the expected value respectively. ,in Indicates the index set of valid subcarriers.

[0098] Focus on formula (11), since To act on the received QAM sequence is the only noise on the surface and is considered as Gaussian white. Its power is given by (10) Decision. The nature of the discussion began.

[0099] In (4) Convert to ,in and The power is equal to Half the power, that is , thus, the phase noise term can be approximated as

[0100] (19)

[0101] when When it is large enough, that is, When , we can further approximate , which obeys the Gaussian distribution as shown below

[0102] (20)

[0103] After cyclic difference, the noise term in (11) is obtained . Note that, due to There is a correlation between adjacent items, so its autocorrelation function is not an impulse function located at the origin, that is, It is not white noise. The autocorrelation function for

[0104] (twenty one)

[0105] Furthermore, we can get its power spectral density function as

[0106] (twenty two)

[0107] During the demapping process, only the index The frequency domain noise on the effective subcarrier is retained, so we can further obtain The power can be expressed as

[0108] (twenty three)

[0109] At the same time, the signal-to-noise ratio of the M-QAM signal in the time domain is The BER formula under the AWGN channel is

[0110] (twenty four)

[0111] Since the power of the QAM sequence is normalized, we can obtain (19) by combining (20), (23) and (24).

[0112] According to formula (18), it can be seen that under ideal phase demodulation conditions, the larger the MI, the better the BER performance of FM-SC-OFDM.

[0113] Perceived CRLB:

[0114] Based on (14), the estimation of time delay and Doppler is equivalent to the frequency estimation of two single-frequency signals. The CRLB of this frequency estimation has been completed in previous studies and is summarized in Lemma 1:

[0115] Lemma 1. Consider a The sampling rate of the points is Discrete-time single-frequency signal ,in and represent constant amplitude and unknown initial phase respectively, The power is AWGN, for the simulation frequency The estimated CRLB is

[0116] (25)

[0117] In (14), the distance and velocity estimates are decoupled and can be viewed as the sum of the two simulation frequencies and Frequency estimation of a single-frequency signal. Taking distance estimation as an example, equation (14) can be simplified to ,in as well as Corresponding to the initial phase and frequency in Lemma 1, the target distance estimation can be expressed as . Further , as well as Substituting into (25), we can get the first The distance estimate CRLB of symbols is

[0118] (26)

[0119] The distance estimate for a frame is obtained by The average distance estimate of the symbols can be expressed as , the corresponding CRLB is In the same way, the CRLB of speed estimation can be obtained. Therefore, the CRLB of ranging and speed measurement of FM-SC-OFDM integrated frame can be obtained as follows:

[0120] (27)

[0121] (28)

[0122] because represents Gaussian white noise The frequency domain matrix of The size of The flatness is determined by The more uneven, The larger the MI, the lower the corresponding perception accuracy. However, as the MI increases, the spectrum components of FM-SC-OFDM are expanded. As the flatness improves, the accuracy of distance and speed measurement will be higher.

[0123] Phase demodulation performance improvement:

[0124] As previously analyzed, a larger MI can improve ISAC performance in FM-SC-OFDM. However, in traditional FM-OFDM, the MI value is severely limited by the phase demodulator's operational requirements. The following analysis will explore how FM-SC-OFDM can expand the MI range and demonstrate robustness against phase demodulation errors.

[0125] (a) Extended MI upper limit

[0126] As shown in Equations (7) and (8), when the noise factor is not considered, the unwinder will convert the original phase sequence into a large jump (i.e. ) is mistakenly identified as phase wrapping, and then a completely wrong phase sequence is restored. The probability of such misjudgment is equal to

[0127] (29)

[0128] in represents the probability of event B occurring. Under a relatively small MI value, the wrong judgment corresponds to At the peak, because It is power-normalized, so PAPR can be used to describe Minimum value of

[0129] (30)

[0130] Equation (30) illustrates the fundamental principle of optimizing phase demodulation, i.e., reducing dewarping misjudgments: reducing the PAPR of the instantaneous frequency sequence. SC-OFDM preprocesses data symbols through DFT spreading, reassigning spectrum energy and converting frequency-domain signals with high peak-to-average power ratio (PAPR) into quasi-single-carrier waveforms with limited amplitude fluctuations. Numerous existing studies indicate that SC-OFDM can reduce PAPR by 2 to 3 dB compared to OFDM. Consequently, compared to FM-OFDM, FM-SC-OFDM improves the upper limit of MI by approximately 30%.

[0131] (b) Robustness to phase demodulation errors

[0132] The misjudgment of the phase unwrapper can be modeled as acting on the instantaneous phase sequence with an amplitude of , rectangular pulse interference with random start and end positions The output of its cyclic difference can be expressed as

[0133] (31)

[0134] in, represents the trigger item that follows the Bernoulli distribution, and its probability can be expressed as In the demodulation process of FM-OFDM, after the FFT of the corresponding OFDM demodulation, the impulse train This will have a destructive effect on all frequency domain data symbols, resulting in the appearance of a BER degradation floor. However, for FM-SC-OFDM, QAM data symbols are distributed in the time domain, so there is an additional IDFT after the FFT, which can just convert the interference spread across the entire frequency band back to pulse interference in the time domain, that is, At this time, only a few QAM symbols at the pulse are destroyed, thus largely avoiding the occurrence of BER degradation flattening.

[0135] Simulation example:

[0136] The following simulations are used to analyze and evaluate the ISAC performance of FM-SC-OFDM, FM-OFDM, CE-OFDM, and SC-OFDM. BER and MSE are obtained by 5000 and 1000 Monte Carlo simulations, respectively. The distance estimation MSE is .

[0137] In the following simulations, the three constant modulus waveforms: FM-SC-OFDM, FM-OFDM, and CE-OFDM are modulated with 16-QAM, while SC-OFDM is modulated with QPSK. This is done to standardize the same bit rate to ensure a fair comparison. Figure 2 and Figure 3 In the example, the three constant modulus waveforms all use MI values ​​close to the upper limit: 0.80 (FM-SC-OFDM), 0.64 (FM-OFDM), and 0.70 (CE-OFDM). The nonlinear power amplifier sampled Saleh model. The number of IFFT points, i.e., the total number of subcarriers, is , the oversampling factor is , the number of symbols in a frame is , the sampling rate is , the carrier frequency is .

[0138] Figure 2 The BER-SNR curves of four waveforms are shown. It is observed that the BER simulation curve of FM-SC-OFDM (red solid line) converges to the BER theoretical lower bound derived in (18). When compared to FM-OFDM, CE-OFDM, and SC-OFDM, FM-SC-OFDM shows approximately 5dB, 3dB, and 3dB SNR gains, respectively. This is because FM-SC-OFDM has two key advantages: an extended MI upper limit and robustness to deconvolution errors.

[0139] Figure 3 The ranging MSE (solid line) and CRLB (dashed line) for the four waveforms are shown. FM-SC-OFDM's CRLB slightly outperforms the other three waveforms and exhibits a 4dB SNR gain compared to CE-OFDM. Furthermore, the MSE curve for FM-SC-OFDM converges to the CRLB curve when the signal-to-noise ratio is sufficient. This demonstrates FM-SC-OFDM's superior perceptual performance and further reinforces the conclusion previously stated that a larger MI leads to better perceptual accuracy.

[0140] Figure 4The BER-MI curves (at a fixed SNR of 14dB) and the CRLB-MI curves (at a fixed SNR of -20dB) of the three constant modulus waveforms are further shown. It is observed that the CRLB-MI curves show a monotonically decreasing relationship, with FM-SC-OFDM showing a slight advantage. The BER-MI curves are all "V-shaped": before the turning point, the BER performance improves with the increase of MI. After the turning point, the BER performance deteriorates sharply due to the large number of errors in phase demodulation. However, FM-SC-OFDM reaches the turning point at the upper limit of MI of about 0.82, corresponding to a minimum BER of about , compared with FM-OFDM and CE-OFDM, it shows about This phenomenon further illustrates the ISAC performance improvement achieved by FM-SC-ODM due to the extended MI upper limit.

[0141] Hardware Experiment

[0142] In order to further fully verify the actual performance of the present invention, namely FM-SC-OFDM and its ISAC architecture in THz scenarios, a THz ISAC prototype verification machine based on FM-SC-OFDM was built using USRP and THz RF front-end. Figure 5 As shown in the figure, the THz ISAC transmitter uses a host computer to control a USRP-X410 to generate an integrated frame baseband signal (parameters similar to those in the simulation). This signal is then up-converted to a 220 GHz RF frequency by the THz transmitter front-end for transmission. The sensing receiver and the integrated transmitter are co-located and receive the angular echo at a distance of 3 meters from the sensing receiver. This signal is down-converted to baseband and a modified version of 2D-FFT is used to estimate the distance and calculate the ranging mean square error (MSE). The communication receiver is 6 meters away from the integrated transmitter and down-converts the received signal to baseband via the THz receiver front-end. This signal is then transmitted to another USRP-X410, where it is demodulated and received bits are collected by another host computer for BER calculation.

[0143] Table 1 shows the BER performance of various waveforms with and without CFO correction:

[0144] Table 1. Comparison of BER performance with and without carrier frequency offset correction

[0145]

[0146] With CFO correction, FM-SC-OFDM achieves an SNR of about 14dB. The BER level shows an SNR gain of at least 4dB compared to FM-OFDM, and an even higher SNR gain compared to CE-OFDM and SC-OFDM. Without CFO correction, the BER performance of FM-SC-OFDM and FM-OFDM remains stable, while the BER of CE-OFDM and SC-OFDM deteriorates sharply. In this case, FM-SC-OFDM shows an SNR gain of at least 8dB compared to CE-OFDM.

[0147] Table 2 shows the MSE of the four waveforms:

[0148] Table 2. Ranging MSE experimental data

[0149]

[0150] In comparison, FM-SC-OFDM achieves the lowest MSE and slightly outperforms the other waveforms in perceptual accuracy.

[0151] These experimental observations are highly consistent with theoretical analysis, highlighting the superior performance of FM-SC-OFDM in both communication and perception, which stems from its higher upper limit on MI. Furthermore, the technology's constant envelope characteristics and frequency offset robustness further enhance its potential application in terahertz-band integrated interawareness (ISAC) systems.

Claims

1. A terahertz communication and perception integrated method based on FM-SC-OFDM waveform, characterized in that: include: The integrated transmitter generates a communication-aware integrated signal by performing M-QAM modulation on the source bits to obtain a QAM sequence. This is then transformed through DFT, zero-padding, conjugate symmetric mapping, and IFFT to obtain a real-number SC-OFDM sequence. The SC-OFDM is frequency modulated to obtain a phase sequence, and finally, phase modulation is performed to obtain an FM-SC-OFDM waveform, i.e., the communication-aware integrated signal. The communication receiver receives the FM-SC-OFDM waveform signal and demodulates it to obtain the received bits. Specifically, the received signal is first subjected to cyclic prefix removal, channel estimation, and equalization. Then, phase demodulation is performed using an inverse tangent machine and deconvolution. The instantaneous frequency SC-OFDM sequence is extracted using cyclic differencing. Finally, SC-OFDM demodulation is performed to obtain the received bits. The perception receiver extracts target distance and speed information based on the perceived FM-SC-OFDM waveform signal. Specifically, it performs 2D-FFT processing on the acquired FM-SC-OFDM integrated frame containing multiple symbols to extract target distance and speed information.

2. The terahertz communication and perception integration method based on FM-SC-OFDM waveform according to claim 1 is characterized in that: In the integrated transmitter, the source bits are defined as the M-QAM modulated Dot symbol sequence , then in the integrated transmitter, The specific processing process is as follows: Perform DFT operation to obtain its frequency domain sequence ,Will Perform conjugate symmetric frequency domain mapping to obtain a frequency domain sequence , ,in represent The zero vector of the point is used to achieve oversampling, and the oversampling multiple is recorded as ,in Represents the effective number of subcarriers, Represents the number of IFFT points. Perform IFFT to obtain the instantaneous frequency SC-OFDM sequence and normalize its power: ， For instantaneous frequency series Perform frequency modulation to obtain instantaneous phase sequence : ， in is the modulation index; Through Perform phase modulation to obtain an FM-SC-OFDM signal with a constant amplitude of A: 。 3. The terahertz communication perception integration method based on FM-SC-OFDM waveform according to claim 2 is characterized in that: In a communication receiver, the baseband received signal after cyclic prefix removal, channel estimation, and equalization is defined as: ， in, represents additive white Gaussian noise (AWGN), represents the normalized frequency deviation, and represent the analog frequency offset and subcarrier spacing respectively, Represents the initial phase deviation constant. In a quasi-static channel environment, is considered as a constant within the symbol period; the baseband received signal is reconstructed as: ， in represents the phase noise introduced by AWGN, represents the frequency deviation factor; The process of phase demodulation of the baseband received signal is: ， in, represents the imaginary and real parts, represents the inverse tangent machine, Represents the phase unwinder, which is used to identify phase jump values ​​exceeding the threshold point, and the phase sequence from that point onwards is compensation to relieve The output phase is blurred, represents the noise introduced by the unwinding error, Mathematically expressed as: ， in, ， definition is a cyclic difference function, Do cyclic differencing: ， in It represents the modulo-N operation. The output of the cyclic difference is the received SC-OFDM sequence. Then, FFT, demapping and IFFT are performed in sequence to extract the received symbol sequence.

4. The terahertz communication perception integration method of FM-SC-OFDM waveform according to claim 3 is characterized in that: In the perceptual receiver, the definition of FM-SC-OFDM integrated frame includes symbols, and the echo signal obtained by the sensing receiver after being reflected by the target is: ， in , , , , , as well as Represent subcarrier spacing, symbol period, echo delay, Doppler shift, rectangular window function, speed of light, and carrier frequency respectively; represents the frequency domain matrix of the integrated frame, and Represent the subcarrier index and symbol index respectively; the echo signal is converted to the frequency domain as follows: ， in, Represents the frequency domain matrix of time domain Gaussian white noise; and Doing matrix dot division yields a matrix containing delay and Doppler information. ,in Represents the frequency domain matrix of the signal after threshold processing, represents the amplitude threshold, Expressed as ， Respectively Perform IFFT and FFT on the distance and velocity dimensions to obtain: ， in and Represents the index of distance and speed information respectively, and Represents the number of points of IFFT and FFT respectively. For a single target scene, and represent The relationship between the index of the maximum value and the distance and speed information of the target is: 。

Images