Method for multipath time delay estimation based on orthogonal linear frequency modulation wave division multiplexing signal

Abstract

The application discloses a multi-path time delay estimation method based on orthogonal linear frequency modulation wave division multiplexing signals, and the process is as follows: an OCDM signal is transmitted by a transmitting end, a time domain receiving signal is obtained after the OCDM signal passes through a multi-path channel, a fractional domain Fourier transform (FrFT) is performed on the time domain receiving signal to obtain a fractional domain receiving signal, a covariance matrix of the fractional domain receiving signal is calculated after the fractional domain receiving signal is de-coherent, a subspace of the covariance matrix is decomposed and a spatial spectrum is calculated by using a MUSIC algorithm, and a time delay corresponding to a spectrum peak of the spatial spectrum is a time delay estimation value. The multi-path time delay estimation method disclosed by the application can accurately estimate a time delay of a dense multi-path channel with a Doppler effect, effectively reduces a relative root mean square error of time delay estimation, is still relatively accurate in estimating a time delay when the multi-path is dense, and has a certain resistance to a Doppler frequency offset.

Description

Technical Field

[0001] This invention relates to the fields of wireless communication and underwater acoustic communication, specifically to a multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (OFM) signals. Background Technology

[0002] Linear Frequency Modulated (LFM) signals, also known as chirp signals, are signals whose frequency varies linearly with time. They possess excellent autocorrelation and spectral characteristics and are widely used in radar, sonar, medical imaging, and time delay estimation. Furthermore, chirp signals can be effectively applied to super-resolution algorithms such as the MUSIC algorithm to achieve accurate estimation of multipath transmission delays. In wireless communication, chirp spread spectrum technology is listed as a physical layer technology in the IEEE 802.15.4 standard. In underwater acoustic communication, chirp signals are widely used for time delay estimation and target detection in time-varying channels where there is relative movement between the transmitter and receiver due to their good performance against Doppler frequency offset. Orthogonal Chirp Division Multiplexing (OCDM) signals are obtained by adding several orthogonal chirp signals with the same modulation slope and equal starting frequency spacing in the time domain; therefore, OCDM signals also possess the characteristics of chirp signals.

[0003] In wireless communication, especially in shallow water acoustic communication, and in urban and suburban areas, delay estimation in dense multipath channels is a challenging problem. This is because signals may travel through multiple paths, including reflection and refraction, to reach the receiver, and these paths are often densely packed. Furthermore, the relative movement speeds of the transmitter and receiver can lead to Doppler effects and further dense transmission delays. Traditional methods often struggle to accurately estimate signal delays in dense multipath channels, impacting signal reception quality and system reliability. Therefore, finding effective techniques for accurately estimating transmission delays in dense multipath channels has been a long-standing research focus. Summary of the Invention

[0004] The purpose of this invention is to reduce the error caused by the Doppler effect in time delay estimation in dense multipath channels where there is relative movement speed at the transmitting and receiving ends. It provides a multipath time delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing signals, which can reduce the relative root mean square error of multipath time delay estimation in the presence of Doppler frequency offset.

[0005] The objective of this invention can be achieved by adopting the following technical solutions:

[0006] A multipath delay estimation method based on orthogonal linear frequency modulation wavelength division multiplexing (OFM) signals, the multipath delay estimation method comprising the following steps:

[0007] S1. Construct the OCDM transmit signal s(t):

[0008] Chirp signals are orthogonal if they have the same modulation slope and their starting frequencies differ by a factor of M (1 / T), where M is a positive integer and T is the duration of the chirp signal. We now prove the orthogonality of two chirp signals s1(t) and s2(t) with the same modulation slope and starting frequencies differing by a factor of M (1 / T):

[0009]

[0010]

[0011] Where f0 is the initial frequency of s1(t), Let B be the frequency modulation slope, and B be the bandwidth of s1(t) and s2(t). The cross-correlation function of s1(t) and s2(t) is:

[0012]

[0013] Since the cross-correlation function of s1(t) and s2(t) is 0, they are orthogonal to each other. That is, two Chirp signals with the same frequency modulation slope and a starting frequency difference of 1 / T are orthogonal to each other.

[0014] An OCDM signal, composed of the time-domain summation of N orthogonal Chirp signals, is used as the test signal. The waveform of this signal in the time-frequency plane is as follows: Figure 1 As shown, there are N line segments with equal slope and length, and their time-domain expression is:

[0015]

[0016] Where f0+k*deltaF,k=0,1,…,N-1 is the starting frequency of the kth subcarrier; T is the duration of each subcarrier; deltaF is an integer multiple of 1 / T; μ=B / T is the modulation frequency of each subcarrier; and B is the signal bandwidth of each subcarrier.

[0017] S2. Calculate the covariance matrix R of the received signal y(t) using FrFT. yy :

[0018] The multipath channel model is as follows:

[0019]

[0020] Where D is the number of multipath paths in the channel, h i and τ i These represent the attenuation coefficient and propagation delay of the i-th path, respectively.

[0021] The time-domain received signal after passing through channel h(t) is:

[0022]

[0023] Where, n k (t) represents the k-th subcarrier with a mean of 0 and a variance of σ. 2 Additive complex Gaussian white noise.

[0024] Performing an optimal-order (0) fractional Fourier transform (FrFT) on y(t), the received signal on the k-th subcarrier is:

[0025]

[0026] Among them, F p [x(t)] represents an O-order FrFT on x(t). The following calculation... According to the FrFT definition:

[0027]

[0028] in, To achieve the optimal FrFT transformation angle, It is a constant related to β.

[0029] Since the integral in equation (8) is over t, we should focus on the terms containing t, including the linear terms. and quadratic terms Adjust the transformation angle β so that μ + cotβ = 0, at which point the quadratic term... The exponent part is 0, so the quadratic term can be eliminated. After eliminating the quadratic term, let... Make a first term The exponent part is 0, so a term can be eliminated once. At this point, only terms not containing t remain. Substitution This is clearly a non-zero term, so the entire integral is not zero. If There exists a term of one kind. Based on the periodicity of sint and cost, their infinite integrals are equal to 0. Furthermore, according to Euler's formula, the following equation holds:

[0030]

[0031] Therefore, the existence of a linear term will make the entire integral zero; that is, the integral is only zero in the linear integral. It has a value at one location and is 0 elsewhere, similar to an impulse function. Therefore, equation (8) can be rewritten as:

[0032]

[0033] Where C is a constant.

[0034] The time-shifting characteristic of FrFT is:

[0035]

[0036]

[0037] The time-shift characteristics of FrFT can be calculated from equation (10) and equation (12). The value of is obtained, thus yielding the calculation result of equation (7). Combining equations (10) and (12), we can obtain:

[0038]

[0039] Substituting equation (13) into equation (7), we get:

[0040]

[0041] Among them, b i =ch i .

[0042] Equation (14) can be written in vector form as follows:

[0043] Y = AX + n (15)

[0044] in, A=[a(τ0),a(τ1),…,a(τ D-1 )], X = [x0, x1, ..., x D-1 ] T ,

[0045] Since multipath signals are generally coherent signals, and the performance of the Multiple Signal Classification (MUSIC) algorithm drops sharply when processing coherent signals, it is necessary to first decohere the received signal Y. The decoherence method used is Modified Spatial Smoothing Processing (MSSP).

[0046] The covariance matrix R of the decoherent received signal in the fractional domain yy for:

[0047] R yy =E(YY) H ) = AE(XX HA H +σ 2 I = AR XX A H +σ 2 I (16)

[0048] S3. Estimate the delay using the MUSIC algorithm:

[0049] R yy eigenvalues ​​λ i and the eigenvector u corresponding to this eigenvalue i Defined as:

[0050] R yy u i =λ i u i (17)

[0051] Matrix R yy N eigenvalues ​​λ i The numbers i = 0, 1, ..., N-1 are arranged in descending order as follows:

[0052] λ0≥λ1≥…≥λ N-1 (18)

[0053] The N eigenvectors corresponding to these eigenvalues ​​are u i ,i=0,1,…,N-1, Representing equation (17) in matrix form, we have:

[0054] R yy U=UΛ (19)

[0055] Among them, U=[u0,u1,…,u N-1 ],Λ=diag[λ0,λ1,…,λ N-1 ]. Matrix R yy It is a Hermitian matrix, meaning all eigenvalues ​​are real numbers, and the eigenvectors corresponding to different eigenvalues ​​are mutually orthogonal. That is:

[0056]

[0057] Matrix R yy It consists of two parts: a signal subspace and a noise subspace.

[0058]

[0059] Among them, Λ S From matrix R yy The first D larger eigenvalues ​​λ0, λ1, ..., λ D-1 Composition, Λ S =diag[λ0,λ1,…,λ D-1];Λ N From matrix R yy The last ND smaller eigenvalues ​​λ D ,λ D+1 ,…,λ N-1 Composition, Λ N =diag[λ D ,λ D+1 ,…,λ N-1 ];U S =[u0,u1,…,u D-1 ] is a matrix R yy The signal subspace formed by the eigenvectors corresponding to the D largest eigenvalues; U N =[u D ,u D+1 ,…,u N-1 ] is a matrix R yy The noise subspace is spanned by the eigenvectors corresponding to the smallest ND eigenvalues ​​in the equation. According to equation (20), the noise subspace U can be obtained. N With the steering vector a of the received signal H (τ i ) are orthogonal, that is:

[0060] a H (τ i )U N =0 (22)

[0061] Therefore, it is possible to search the noise subspace U. N The orthogonal steering vector is used to estimate the steering vector of the signal. Since the length of the received signal is finite in practical situations, R... yy It can be simplified to:

[0062]

[0063] Where L is the number of snapshots. For the covariance matrix... Eigenvalue decomposition can yield the noise subspace. Because of a H (τ i (cannot be with) Since the signals are completely orthogonal, an optimization search method is needed to estimate their time delay. The spatial spectrum estimation formula for the MUSIC algorithm is:

[0064]

[0065] The D maximum peaks in the spatial spectrum correspond to the estimated time delays of the S multipath components in the signal:

[0066]

[0067] The present invention has the following advantages and effects compared with the prior art:

[0068] (1) The multipath delay estimation method disclosed in this invention uses the OCDM signal as the transmitted signal. When the Doppler effect causes the position of the receiving array element to shift, the fractional domain form of the OCDM signal will multiply the shift by a sine factor less than or equal to 1, thereby reducing the array element shift and thus reducing the estimation error. Therefore, this method can resist the Doppler effect to a certain extent when there is a relative velocity at the transmitting and receiving ends, reduce the estimation error caused by the Doppler frequency offset, and improve the accuracy of multipath delay estimation.

[0069] (2) The multipath delay estimation method disclosed in this invention includes the MUSIC algorithm, which is based on subspace decomposition and is therefore a super-resolution estimation algorithm. Thus, this method can be applied to dense multipath channels, and the minimum resolvable delay interval can break through the Rayleigh limit to achieve super-resolution delay estimation accuracy. Attached Figure Description

[0070] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0071] Figure 1 This is a flowchart illustrating the multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (WDM) signals in Embodiment 1 of the present invention.

[0072] Figure 2 This is a waveform diagram of the OCDM signal composed of 20 orthogonal Chirp signal subcarriers in the time-frequency plane in Embodiment 1 of the present invention;

[0073] Figure 3 This is a schematic diagram of the MUSIC spatial spectrum using the OCDM signal as the transmitted signal in Embodiment 1 of the present invention. Figure 3 RRMSE = -49.86 dB;

[0074] Figure 4 This is a schematic diagram of the MUSIC spatial spectrum using the OCDM signal as the transmitted signal in Embodiment 2 of the present invention. Figure 4 RRMSE = -39.93dB;

[0075] Figure 5 This is a schematic diagram of the MUSIC spatial spectrum using OFDM signals as the transmitted signals in Embodiment 2 of the present invention. Figure 5 RRMSE = -37.68 dB;

[0076] Figure 6This is a graph showing the time delay estimation (RRMSE) as a function of SNR when the relative speed between the transmitter and receiver is 5 m / s in Embodiment 2 of the present invention, using OCDM and OFDM signals as the transmitted signals respectively. Detailed Implementation

[0077] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0078] Example 1

[0079] This embodiment discloses a multipath delay estimation method based on orthogonal linear frequency modulation wavelength division multiplexing (OFM) signals. The method uses an OCDM signal obtained by time-domain summation of several orthogonal Chirp signals with the same modulation slope and equal initial frequency spacing as the transmitted signal. After passing through a multipath channel, a time-domain received signal is obtained. A fractional Fourier transform (FrFT) is performed on the time-domain received signal to obtain a fractional-domain received signal, which is then decohered. The MUSIC algorithm is used to perform eigenvalue decomposition on the covariance matrix of the decoherent fractional-domain received signal, thereby achieving super-resolution delay estimation for the multipath channel. Specifically, the method includes the following steps:

[0080] S1. Constructing the OCDM transmit signal s(t): The transmitter transmits an OCDM signal composed of 20 orthogonal chirp signals as subcarriers. The bandwidth B, duration T, and frequency modulation slope μ of each subcarrier are 1kHz, 100ms, and 10kHz / s, respectively, and the starting frequency interval deltaF between adjacent subcarriers is 25 / T. The time-domain expression of this transmit signal is:

[0081]

[0082] The number of subcarriers is N = 20, and the starting frequency f0 of the first subcarrier is 0.7 kHz. Since the starting frequency interval between adjacent subcarriers is an integer multiple of 1 / T, these 20 chirp signals are mutually orthogonal.

[0083] S2. Calculate the covariance matrix R of the received signal y(t) using FrFT. yy The multipath channel model is as follows:

[0084]

[0085] The number of multipath channels is 3; the attenuation coefficients h0, h1, and h2 are 1, 0.94, and 0.8, respectively; and the propagation delays τ0, τ1, and τ2 are 0.1 ms, 0.6 ms, and 1.1 ms, respectively. The transmitted signal s(t) in equation (26) passes through the multipath channel h(t) in equation (27) to obtain the received signal y(t) in the time domain as follows:

[0086]

[0087] Where, n k (t) represents the k-th subcarrier with a mean of 0 and a variance of σ. 2 Additive complex white Gaussian noise. Performing an optimal p-order FrFT on y(t), the received signal on the k-th subcarrier is:

[0088]

[0089] The optimal transformation order p of y(t) is calculated to be 1.0059. Based on the time-shifting characteristics of FrFT and Euler's formula, the value in equation (29) is... It can be rewritten as follows:

[0090]

[0091] The optimal transformation angle β can be derived from... The calculated value is 1.58. Equation (30) can be written in vector form as follows:

[0092] Y = AX + n (31)

[0093] in, A = [a(τ0), a(τ1), a(τ2)], X = [x0, x1, x2] T ,

[0094] After performing MSSP decoherence on the received signal, calculate its covariance matrix in the fractional domain:

[0095] R yy =E(YY) H ) = AE(XX H A H +σ 2 I = AR XX A H +σ 2 I (32)

[0096] Because the length of the received signal is limited in real-world situations, R yy It can be simplified to:

[0097]

[0098] S3. Estimating delay using the MUSIC algorithm: For the matrix Perform eigenvalue decomposition to make It is divided into two parts: a signal subspace and a noise subspace.

[0099]

[0100] Among them, Λ S From matrix R yy Composed of the first three larger eigenvalues, Λ S =diag[λ0,λ1,λ2];Λ N From matrix R yy It consists of the last 17 smaller eigenvalues, Λ N =diag[λ3,λ4,…,λ 19 ]; It is a signal subspace composed of the eigenvectors corresponding to the three largest eigenvalues; It is a noise subspace spanned by the eigenvectors corresponding to the 17 smallest eigenvalues. Because the noise subspace... With the steering vector a of the received signal H (τ i Since the signal's steering vector is orthogonal to the noise subspace, the steering vector of the signal can be estimated by searching for a steering vector orthogonal to the noise subspace. The spatial spectrum estimation formula of the MUSIC algorithm is:

[0101]

[0102] The three maximum peaks in the spatial spectrum correspond to the time delay estimates of the three multipath components in the signal.

[0103]

[0104] Figure 3 The MUSIC spatial spectrum is generated when the SNR is 0dB and the OCDM signal is used as the transmitted signal. The actual values ​​of the three multipath delays of this channel are 0.1000ms, 0.6000ms, and 1.1000ms, respectively. Figure 3 The estimated MUSIC spatial spectrum delay values ​​using the OCDM signal are 0.1003 ms, 0.5993 ms, and 1.0982 ms, with a relative root mean square error (RRMSE) of -49.86 dB compared to the true value. This embodiment demonstrates that even with a low signal-to-noise ratio (SNR = 0 dB), it can accurately estimate dense multipath delay when using the OCDM signal as the transmit signal.

[0105] Example 2

[0106] This embodiment discloses a super-resolution multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (OFM) signals, specifically including the following steps:

[0107] S1. Constructing the OCDM transmit signal s(t): The transmitter remains stationary, while the receiver moves towards the transmitter at a speed of 5 m / s. The transmitter transmits an OCDM signal composed of 20 orthogonal chirp signals as subcarriers. The bandwidth B, duration T, and frequency modulation slope μ of each subcarrier are 1 kHz, 100 ms, and 10 kHz / s, respectively, and the initial frequency interval deltaF between adjacent subcarriers is 25 / T. The time-domain expression of this transmit signal is:

[0108]

[0109] The number of subcarriers is N = 20, and the starting frequency f0 of the first subcarrier is 0.7 kHz. Since the starting frequency interval of each subcarrier is an integer multiple of 1 / T, these 20 chirp signals are mutually orthogonal.

[0110] S2. Calculate the covariance matrix R of the received signal y(t) using FrFT. yy The multipath channel model is as follows:

[0111]

[0112] The number of multipath paths in the channel is 3; the attenuation coefficients h0, h1, and h2 are 1, 0.94, and 0.8, respectively; and the propagation delays τ0, τ1, and τ2 are 0.1 ms, 0.6 ms, and 1.1 ms, respectively. The transmitted signal s(t) in equation (37) passes through the channel h(t) in equation (38) to obtain the received signal y(t) as follows:

[0113]

[0114] Where, n k (t) represents the k-th subcarrier with a mean of 0 and a variance of σ. 2 Additive complex white Gaussian noise. Performing an optimal order p fractional Fourier transform on y(t), the received signal on the k-th subcarrier is:

[0115]

[0116] Calculations show that the optimal transformation order p of y(t) is 1.0059. Based on the time-shifting characteristics of FrFT and Euler's formula, the value in equation (40) is... It can be rewritten as follows:

[0117]

[0118] The optimal transformation angle β can be derived from... The calculated value is 1.58. Equation (41) can be written in vector form as follows:

[0119] Y = AX + n (42) where, A = [a(τ0), a(τ1), a(τ2)], X = [x0, X1, X2] T ,

[0120] When there is a relative velocity v at the transmitting and receiving ends, a Doppler frequency shift will occur, causing the steering vector a(τ) to... i ) changes, that is, in the guiding vector a(τ) i The exponential part of the product is multiplied by a Doppler scaling factor. Where c is the speed of sound in the medium, v>0 represents the transmitting and receiving ends traveling towards each other, and v<0 represents the transmitting and receiving ends traveling away from each other. The steering vector then becomes... When the relative velocity between the transmitter and receiver is 5 m / s, the Doppler scale factor At this time, the guide vector is... Become

[0121] The covariance matrix of the received signal in the fractional domain is:

[0122] R yy =E(YY) H ) = AE(XX H A H +σ 2 I = AR XX A H +σ 2 I (43)

[0123] Among them, A=[a′(τ0),a′(τ1),a′(τ2)],

[0124]

[0125] Because the length of the received signal is limited in real-world situations, R yy It can be simplified to:

[0126]

[0127] S3. Estimating delay using the MUSIC algorithm: For the matrix Perform eigenvalue decomposition to make It is divided into two parts: a signal subspace and a noise subspace.

[0128]

[0129] Among them, Λ S From matrix R yy Composed of the first three larger eigenvalues, Λ S =diag[λ0,λ1,λ2];Λ N From matrix R yy It consists of the last 17 smaller eigenvalues, Λ N =diag[λ3,λ4,…,λ 19 ]; It is a signal subspace composed of the eigenvectors corresponding to the three largest eigenvalues; It is a noise subspace spanned by the eigenvectors corresponding to the 17 smallest eigenvalues. Because the noise subspace... With the steering vector a of the received signal H (τ i Since the signal's steering vector is orthogonal to the noise subspace, the steering vector of the signal can be estimated by searching for a steering vector orthogonal to the noise subspace. The spatial spectrum estimation formula of the MUSIC algorithm is:

[0130]

[0131] The three maximum peaks in the spatial spectrum correspond to the time delay estimates of the three multipath components in the signal.

[0132]

[0133] Figure 4 , Figure 5 The MUSIC spatial spectra using OCDM and OFDM signals as the transmit signals are shown under the same conditions (SNR of 0dB and relative speed between the transmitter and receiver of 5m / s). The actual values ​​of the three multipath delays of this channel are 0.1000ms, 0.6000ms, and 1.1000ms, respectively. Figure 4 The estimated time delay values ​​of the MUSIC spatial spectrum using the OCDM signal are 0.1056ms, 0.6006ms, and 1.1022ms, with an RRMSE of -39.93dB compared to the true value, and the maximum peak-to-valley difference of the spatial spectrum is 70.13dB. Figure 5 The estimated time delay values ​​of the MUSIC spatial spectrum using OFDM signals are 0.1016 ms, 0.6046 ms, and 1.1062 ms, with a relative ground truth (RRMSE) of -37.68 dB and a maximum peak-to-valley difference of 55.47 dB. Combined with... Figure 4 , Figure 5It can be seen that, compared with OFDM signals, using OCDM signals as the transmitted signal results in a larger maximum peak-to-valley difference in the spatial spectrum, sharper, clearer, and more distinguishable spectral peaks, and a smaller RRMSE between the obtained time delay estimate and the true value.

[0134] Figure 6 The graphs show the time delay estimation (RRMSE) as a function of signal-to-noise ratio (SNR) when the relative velocity between the transmitter and receiver is 5 m / s, using OCDM and OFDM signals as the transmitted signals, respectively. Figure 6 As can be seen, the RRMSE of the time delay estimation algorithm decreases with the increase of SNR, and the estimation error using the OCDM signal as the transmit signal is smaller than that using the OFDM signal. This advantage is particularly obvious at low SNR (-5 to 5 dB). In summary, this embodiment can reduce the RRMSE of multipath time delay estimation, reduce the adverse effects of the Doppler effect, and improve the accuracy of time delay estimation when there is a relative speed at the transmitting and receiving ends, especially under low SNR conditions.

[0135] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (FMWDM) signals, characterized in that, The multipath delay estimation method includes the following steps: Construct OCDM transmission signal The transmitter sends several orthogonal chirp signals with the same frequency modulation slope and equal starting frequency spacing as subcarriers, with the starting frequency spacing being... , The OCDM transmit signal is an integer multiple of the duration of the Chirp signal. It is obtained by adding the above Chirp signals in the time domain; Calculate the received signal using fractional Fourier transform covariance matrix OCDM transmission signal After multipath channel The received signal in the time domain is obtained after propagation. ,right Make the best order Fractional domain received signal obtained by FrFT ,Bundle Written The vector form, for After decoherence, calculate its covariance matrix. FrFT is short for Fractional Fourier Transform. Represents the guidance matrix. This represents a matrix related to the transmitted signal. This indicates that the mean is 0 and the variance is 0. Additive complex white Gaussian noise, Represents the vector form of the received signal in the fractional domain; The received signal is calculated using fractional Fourier transform. covariance matrix middle, ， ， ， ， ， in, The optimal transform order for FrFT is... The number of multipath paths in the channel. The optimal transformation angle for FrFT. They represent the 0th, 1st, ..., ... The received signal on each subcarrier passes through Fractional field form after FrFT of order 1, Representation and delay The relevant guide vector, Indicates and , and latency Relevant constants, , For the first The attenuation coefficient of the path, It is a constant. The number of subcarriers; Estimating latency using the MUSIC algorithm: Subspace decomposition is performed to obtain the signal subspace. and noise subspace The MUSIC spatial spectrum is calculated using the orthogonality of the steering vector of the received signal and the noise subspace. The time delay corresponding to the spectral peak is the time delay estimate.

2. The multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (FMWDM) signals according to claim 1, characterized in that, The construction of OCDM transmission signal In the middle, the frequency modulation slopes are equal and the initial intervals are integer multiples. The chirp signal subcarriers are mutually orthogonal. The time-domain expression is as follows: , in, For the first The starting frequency of each subcarrier; The duration of the chirp signal is equal to the duration of each subcarrier. The frequency modulation slope for each subcarrier, The signal bandwidth for each subcarrier, The starting frequency of the first subcarrier. The starting frequency difference between adjacent subcarriers is denoted by t, where t represents time.

3. The multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (FMWDM) signals according to claim 2, characterized in that, In the time delay estimation using the MUSIC algorithm, the signal subspace From the matrix The largest The eigenvectors corresponding to each eigenvalue are constructed. They represent the 0th, 1st, ... 1st, ... The eigenvectors corresponding to the largest eigenvalues, and the noise subspace. From the matrix smallest The eigenvectors corresponding to each eigenvalue are constructed. They represent the first The, the The... The eigenvectors corresponding to the largest eigenvalues.

4. The multipath delay estimation method based on orthogonal linear frequency modulated wavelength division multiplexing (FMWDM) signals according to claim 3, characterized in that, In the time delay estimation using the MUSIC algorithm, the MUSIC spatial spectrum calculation formula is as follows: ， in, As the guide vector, For the noise subspace, the time delay value corresponding to the peak of the MUSIC spatial spectrum is the time delay estimate.

Images