A high-resolution multipath time-delay estimation method and system based on multi-frequency cooperation

Abstract

The present invention discloses a high-resolution multipath delay estimation method and system based on multi-frequency cooperation, constructs a multi-band system model, and obtains a multi-band channel state information matrix by performing equally-spaced frequency-domain sampling on different frequency bands; performs subspace decomposition on the obtained multi-band channel state information matrix, and constructs an arrival delay matrix by using the coherent subspace; and estimates the multipath delay by using the properties of the obtained arrival delay matrix. The present invention utilizes subsystems on different frequency bands in a multi-frequency cooperation system to obtain a larger available bandwidth, and while greatly improving the delay estimation accuracy, it will not bring huge system losses due to the large-bandwidth transceiver RF link.

Description

Technical Field

[0001] The present invention belongs to the technical field, and particularly relates to a high-resolution multipath delay estimation method and system based on multi-frequency cooperation. Background Art

[0002] Arrival delay estimation is an important signal processing problem and is widely applied in projects such as radar, sonar, and positioning. Existing methods for estimating multipath delay mainly focus on single-frequency scenarios. Generally speaking, a larger signal bandwidth can bring a larger delay resolution. With the continuous innovation of communication technologies, the overall signal bandwidth used for delay estimation has gradually expanded with the improvement of accuracy. Therefore, in order to greatly improve the accuracy of delay estimation, constructing a system using signals with a larger bandwidth is a very direct research idea. However, due to physical limitations, during the development towards ultra-high bandwidth, the difficulty and cost of the transceiver system will also increase accordingly. In recent years, the ultra-wideband UWB system, which is popular in the field of high-precision delay estimation research, its core is to utilize characteristics such as ultra-high bandwidth and extremely short signal duration to obtain higher delay estimation accuracy, but there are still problems of too high system complexity and cost, thus restricting the large-scale commercial use of this system. In order to estimate high-precision delay without increasing system cost and by making the best use of existing resources as much as possible, a new research idea is to expand the available bandwidth of the signal by using the cooperation of multiple frequency bands on the basis of the existing communication frequency band, so as to construct an equivalent ultra-wideband system to obtain high-precision delay estimation.

[0003] Defects and deficiencies of the prior art:

[0004] 1. Most of the existing delay estimation methods are applicable to the delay estimation of single-frequency signals and are not applicable to multi-frequency cooperation systems.

[0005] 2. The performance of traditional delay estimation algorithms based on single-frequency systems is limited by the bandwidth of the transmitted signal, and due to hardware limitations, a single broadband system has not been widely used in existing communication systems. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a high-resolution multipath delay estimation method and system based on multi-frequency cooperation for the deficiencies in the above-mentioned prior art, which extends the traditional single-frequency signal model to multiple frequencies to solve the technical problem of low delay estimation accuracy caused by limited system bandwidth.

[0007] The present invention adopts the following technical solutions:

[0008] A high-resolution multipath delay estimation method based on multi-frequency cooperation, comprising the following steps:

[0009] S1. Construct a multi - band system model, and perform equally - spaced frequency - domain sampling for different bands to obtain a multi - band channel state information matrix;

[0010] S2. Perform subspace analysis on the multi - band channel state information matrix obtained in step S1, and construct an arrival delay matrix using the coherent subspace;

[0011] S3. Use the multi - resolution estimation method to estimate the arrival delay matrix obtained in step S2 and determine the arrival delay.

[0012] Specifically, in step S1, the multi - band channel state information matrix x N is:

[0013] x N = H N + w N = VΘ N-1 a1 + w N

[0014] where H N is the signal matrix corresponding to the Nth band, w N is the noise matrix in the Nth band, V is the Vandermonde matrix composed of multi - path signal delays, Θ is the additional phase matrix composed of the frequency band interval and multi - path delays, and a1 is the multi - path attenuation matrix.

[0015] Furthermore, the multi - band system model is specifically:

[0016]

[0017] where x1(l) is the l - th sampling data of the first frequency band, H1(f l ) is the signal part in the l - th frequency - domain sampling, w1(l) represents additive white Gaussian noise with a mean of zero and a variance of , L p is the number of multi - path delays in the channel, α k is the complex amplitude of the corresponding k - th path, f0 is the carrier frequency of the initial band, Δf is the sampling frequency interval, and τ k is the propagation delay of the corresponding k - th path.

[0018] Specifically, in step S2, solve the autocovariance matrix of the first band and perform eigenvalue decomposition to reconstruct the signal subspace Solve the cross - covariance matrix R 21 between the adjacent higher - frequency band and this band, and determine the first arrival delay matrix according to the signal subspace 21 and the cross - covariance matrix R Then, solve the arrival delay matrices of all adjacent bands in turn and take the average as the final arrival delay matrix.

[0019] Furthermore, the arrival delay matrix R TOA is as follows:

[0020]

[0021] where N is the number of frequency bands in the system, so there are N - 1 adjacent frequency bands in the system, is the arrival delay matrix obtained from the channel state information matrix of the i-th pair of adjacent frequency bands.

[0022] Specifically, in step S3, the sampling multi-resolution estimation method for estimating the arrival delay is as follows:

[0023] S301. Use the ESPRIT algorithm to obtain the rotation invariant matrix corresponding to matrix V, and use the sampling frequency interval Δf within the frequency band to obtain a rough estimate of the arrival delay;

[0024] S302. Estimate the phase winding number n k ;

[0025] S303. Use the frequency band interval B to obtain a fine estimate of the high-resolution arrival delay, and repeat steps S301 to S303 for the signals of different paths to estimate the arrival delays of all paths.

[0026] Furthermore, in step S301, the rough estimate τ of the k-th arrival delay k is as follows:

[0027]

[0028] where Φ k is the k-th diagonal element of the rotation invariant matrix Φ, and the rotation invariant matrix Φ is:

[0029]

[0030] Furthermore, in step S302, the estimated value k of the phase winding number n is as follows:

[0031]

[0032] where are respectively the estimated values of the k-th diagonal element φ k of the rotation invariant matrix Φ, and the actual phase θ k of the k-th diagonal element of matrix Θ.

[0033] Furthermore, in step S303, the arrival delay corresponding to the k-th propagation path is as follows:

[0034]

[0035] Among them, is the estimated value of the actual phase θ k of the k-th diagonal element of the matrix Θ, is the estimated value of the phase winding number n k of

[0036] In a second aspect, an embodiment of the present invention provides a high-resolution multipath delay estimation system based on multi-frequency cooperation, including:

[0037] A sampling module that constructs a multi-band system model and obtains a multi-band channel state information matrix by performing equally spaced frequency-domain sampling for different bands;

[0038] An analysis module that performs subspace analysis on the multi-band channel state information matrix obtained by the sampling module and constructs an arrival delay matrix using a coherent subspace;

[0039] An estimation module that estimates the arrival delay matrix obtained by the analysis module using a multi-resolution estimation method to determine the arrival delay.

[0040] Compared with the prior art, the present invention has at least the following beneficial effects:

[0041] A high-resolution multipath delay estimation method based on multi-frequency cooperation extends the traditional single-band signal model to multiple bands, constructs a multi-frequency cooperation model, and can achieve the performance of a broadband system without increasing additional system overhead; a high-resolution delay estimation method based on a coherent subspace applicable to the multi-frequency cooperation model is proposed, which links multiple bands together, constructs a signal with a larger available bandwidth using the band interval, and can obtain high-precision delay estimation results through a multi-resolution estimation method, which is superior to the traditional single-band delay estimation performance. While greatly improving the delay estimation accuracy, it will not cause huge system losses due to the large-bandwidth transceiver RF link.

[0042] Furthermore, the multi-band channel state information matrix is given, revealing the relationship between the multi-band channel state information matrices, that is, the channel state information matrix corresponding to the subsequent band can be deduced from the structure of the first band.

[0043] Furthermore, the multi-frequency system model shows the sampling results, and then obtains the channel state information matrices of different bands; similarly, by sequentially setting the sampling start frequency to the carrier frequencies corresponding to the other bands and selecting the same sampling interval and the number of sampling points, the multi-band channel state information matrices can be obtained.

[0044] Further, an arrival time delay matrix is constructed using the obtained multi-band channel information matrix. The purpose of constructing the arrival time delay matrix is to utilize the coherence relationship between adjacent frequency bands. Taking the initial adjacent dual-frequency bands as an example, first, the autocovariance matrix of the first frequency band is solved and eigenvalue decomposition is performed. Then, the arrival time delay matrices of all adjacent frequency bands are successively solved, and the average value is taken as the final arrival time delay matrix. The constructed arrival time delay matrix contains the time delay information between all frequency bands and can be used to obtain time delay estimation.

[0045] Further, the arrival time delay matrix R TOA incorporates the coherence information between all adjacent frequency bands, which includes the phase information of the combination of multipath time delays with the frequency band interval and the sampling interval respectively. That is, after performing eigenvalue decomposition on the obtained arrival time delay matrix R TOA the matrices Θ and V can be obtained. Among them, the diagonal elements of the matrix Θ are the non-zero eigenvalues of the arrival time delay matrix R TOA and contain the phase information of the combination of multipath time delays with the frequency band interval; the corresponding column vectors of the matrix V are the eigenvectors corresponding to the non-zero eigenvalues of the arrival time delay matrix R TOA and contain the phase information of the combination of multipath time delays and the sampling interval.

[0046] Further, based on the obtained matrices Θ and V, a multi-resolution estimation method is used to estimate the arrival time delay. This is because in practice, the adjacent frequency band interval B is generally much larger than the sampling interval Δf, resulting in the phase 2πBτ k being much larger than π, which will cause aliasing of the 2π phase. By using the multi-resolution estimation method, the original phase 2πBτ k can be solved, thereby obtaining a high-precision time delay estimation result.

[0047] Further, the rotation invariant matrix Φ corresponding to the matrix V is extracted using the ESPRIT algorithm.

[0048] Further, based on the result of the coarse time delay estimation, the phase winding number is estimated. Theoretically, the result of the coarse time delay estimation is the same as that of the high-resolution fine estimation. Estimating the phase winding number n k is to obtain the original phase information of the combination of the frequency band interval and the multipath time delay.

[0049] Further, finally, the result of estimating the multipath time delay using the multi-frequency band interval is obtained. According to the obtained result of estimating the phase winding number, the fine estimation result of the arrival time delay corresponding to the k-th propagation path is finally obtained.

[0050] It can be understood that the beneficial effects of the second aspect above can be referred to the relevant descriptions in the first aspect above, and will not be elaborated here.

[0051] In summary, the present invention aims to utilize the subsystem intervals on different frequency bands in a multi-frequency collaborative system to obtain a larger available bandwidth and thus achieve a high-precision time-delay estimation result, without causing huge system losses; a model of the multi-frequency system is theoretically given. First, a time-delay matrix is constructed using the multi-frequency channel state information matrix to extract the coherence information between adjacent frequency bands, then the phase information related to the time delay is extracted using the properties of the time-delay matrix, and finally a high-precision time-delay estimation result is obtained by sampling a multi-resolution estimation method.

[0052] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Description of the Drawings

[0053] Figure 1 It is a time-frequency domain channel characteristic diagram of any adjacent frequency bands in the multi-frequency collaborative system;

[0054] Figure 2 It is a frequency band design diagram of the multi-frequency collaborative system;

[0055] Figure 3 It is a high-resolution time-delay estimation technical roadmap based on the multi-frequency collaborative system;

[0056] Figure 4 It is a flowchart of the time-delay estimation method in the present invention;

[0057] Figure 5 It is a curve diagram showing the change of time-delay estimation performance with signal-to-noise ratio under different frequency band intervals in a dual-frequency system;

[0058] Figure 6 It is a curve diagram showing the change of time-delay estimation performance with signal-to-noise ratio under different numbers of frequency bands in the multi-frequency collaborative system. Detailed Embodiments

[0059] The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0060] In the description of the present invention, it should be understood that the terms "include" and "comprise" indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0061] It should also be understood that the terms used in the description of the present invention are for the purpose of describing particular embodiments only and are not intended to limit the present invention. As used in the description of the present invention and the appended claims, unless the context clearly dictates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0062] It should be further understood that the term " / and / or" used in the description of the present invention and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the contextually related objects.

[0063] It should be understood that although the terms first, second, third, etc. may be used in the embodiments of the present invention to describe preset ranges, etc., these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range. Similarly, the second preset range may also be referred to as the first preset range.

[0064] Depending on the context, the word "if" as used herein can be interpreted as "when" or "while" or "in response to determining" or "in response to detecting". Similarly, depending on the context, the phrase "if determined" or "if detected (stated condition or event)" can be interpreted as "when determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)".

[0065] Structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the drawings. These figures are not drawn to scale, where for the purpose of clear expression, some details are enlarged and some details may be omitted. The shapes of the various regions and layers shown in the figures and their relative sizes and positional relationships are only exemplary. In practice, there may be deviations due to manufacturing tolerances or technical limitations, and those skilled in the art can design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0066] The present invention provides a high-resolution multipath delay estimation method based on multi-frequency cooperation. By constructing a multi-band system model, equally-spaced frequency-domain sampling is performed on different frequency bands to obtain a multi-band channel state information matrix. The obtained multi-band channel state information matrix is subjected to subspace decomposition, and a time-delay matrix is constructed using the coherent subspace. Using the properties of the obtained time-delay matrix, a multi-resolution method is employed to estimate the multipath delay. The present invention utilizes subsystems on different frequency bands in a multi-frequency cooperation system to obtain a larger available bandwidth, significantly improving the delay estimation accuracy without incurring huge system losses due to the large-bandwidth transceiver RF links.

[0067] Compared with various existing types of delay estimation technologies, the multi-band cooperation technology has significant characteristics: the propagation paths experienced by multiple frequency bands in a multi-band cooperation system are the same, and the channel state information of each subsystem in each frequency band contains the delay information carried by a signal transmission. When multiple frequency bands work together simultaneously, the goal of the system is to utilize subsystems operating on different frequency bands to bring a larger available bandwidth, thus significantly improving the delay estimation accuracy without incurring huge system losses due to the large-bandwidth transceiver RF links.

[0068] Please refer to Figure 1 , for the time-frequency domain channel characteristics of any two adjacent frequency bands in the multi-frequency cooperation system of the present invention. Theoretically, signals in adjacent frequency bands pass through the same path, so they have the same delay distribution in the time domain, except that they are distributed in different frequency bands in the frequency domain. Further, all frequency bands in the system have similar channel characteristics under the same environment.

[0069] Please refer to Figure 2 , for the frequency band distribution of the designed multi-frequency cooperation system. K frequency bands are equally-spaced and share the same subcarrier structure.

[0070] Please refer to Figure 3 , which shows the high-resolution delay estimation technical route based on the multi-frequency cooperation system. Signals in multiple frequency bands are processed by receivers in different frequency bands to obtain the channel state information of each frequency band, and then fed into the designed time-delay estimation method, finally outputting a high-resolution estimation result of the delay.

[0071] Please refer to Figure 4 , a high-resolution multipath delay estimation method based on multi-frequency cooperation according to the present invention, includes the following steps:

[0072] S1. Construct a multi-band system model, and perform equally-spaced frequency-domain sampling on different frequency bands to obtain a multi-band channel state information matrix;

[0073] Taking the initial adjacent dual-frequency band as an example, the l-th sampling data x1(l) of the first frequency band is modeled as:

[0074]

[0075] wherein, H1(f l ) is the signal part in the l-th frequency-domain sampling, w1(l) represents additive white Gaussian noise with a mean of zero and a variance of , L p is the number of multipath time delays in the channel, α k is the complex amplitude of the corresponding k-th path, f0 is the carrier frequency of the initial frequency band, i.e., the sampling start frequency, Δf is the sampling frequency interval, and τ k is the propagation delay of the corresponding k-th path.

[0076] The number of sampling points in each frequency band is L, and the corresponding vector representation is:

[0077] x1 = H1 + w1 = Va1 + w1

[0078] wherein, the vector representations of each component are:

[0079] x1 = [x1(0) x1(1)... x1(L - 1)] T

[0080] H1 = [H1(f0)... H1(f0 + (L - 1)Δf)] T

[0081] w1 = [w1(0) w1(1)... w1(L - 1)] T

[0082]

[0083]

[0084]

[0085] Suppose the frequency band interval between the second frequency band signal and the first frequency band signal is B, specifically, the sampling start frequency becomes f0 + B, and the rest of the structure is the same as that of the first frequency band, then there is:

[0086]

[0087]

[0088] Thus, the channel model of the second frequency band is obtained as:

[0089] x2 = H2 + w2 = Va2 + w2 = VΘa1 + w2,

[0090] Similarly, the channel model of the N-th frequency band is:

[0091] x N = H N + w N = VΘ N-1 a1 + w N

[0092] S2. Perform subspace analysis on the obtained multi - band channel state information matrix, and construct the time - delay - of - arrival matrix using the coherent subspace;

[0093] Still taking the initial adjacent dual - band as an example, first solve the autocovariance matrix of the first band and perform eigenvalue decomposition as follows:

[0094]

[0095] Among them, ∑ is a diagonal matrix composed of the eigenvalues of R1, Q is a unitary matrix composed of the eigenvectors of R1, ∑ is a diagonal matrix composed of the eigenvalues of R1, and Q is a unitary matrix composed of the eigenvectors of R1.

[0096] ∑ = diag[λ0, λ1,..., λ L-1

[0097] Among them, The first L p eigenvalues are much larger than the last L - L p eigenvalues, which are called signal eigenvalues. Let the eigenvector corresponding to λ l be q l , then the original signal subspace R 1s = VAV H is expressed as:

[0098]

[0099] Considering that the signal eigenvalues are much larger than the noise eigenvalues, approximately obtain the original signal subspace as:

[0100]

[0101] Then, solve the cross - covariance matrix between the adjacent high - frequency band and this band, and the result is:

[0102]

[0103] Then the first time - delay - of - arrival matrix is:

[0104]

[0105] Similarly, solve the time - delay - of - arrival matrices of all adjacent bands in turn, and their average value is the final time - delay - of - arrival matrix: ​

[0106]

[0107] The constructed arrival delay matrix has the following properties:

[0108] R TOA V = VΘ

[0109] The proof is as follows: First, according to the structure of, we can obtain:

[0110]

[0111] where Q s is the signal subspace obtained from the measurement data and is composed of L p signal eigenvectors. The remaining L - L p eigenvectors form the noise subspace Q n . V is a Vandermonde matrix representing the signal part and also belongs to the signal space. According to the orthogonality of the signal subspace and the noise subspace, we can obtain V H Q n =, then we have:

[0112]

[0113] Since V has full column rank and R 1s = VAV H , then we have:

[0114]

[0115] Finally, we can obtain:

[0116]

[0117] Next, it is explained that the above properties apply to any two adjacent frequency bands in the multi - frequency band.

[0118] According to the multi - frequency band channel model in step S1, the channel state information matrix models of the Nth frequency band and the (N + 1)th frequency band are respectively:

[0119] x N = H N + w N = VΘ N-1 a1 + w N

[0120] x N+1 = H N+1 + w N+1 = VΘ N a1 + w N+1 .

[0121] Let V' = VΘN-1 , then V′ is still a full column rank Vandermonde matrix, and its dimension is the same as that of matrix V. Thus, we have:

[0122] x N = V′a1 + w N

[0123] x N+1 = V′Θa1 + w N+1

[0124] Seeing that the structure at this time is the same as the initial adjacent dual-band, we have the following result:

[0125]

[0126] Substitute V′ = VΘ N-1 and expand it, we get:

[0127]

[0128] Since Θ is an invertible square matrix, then Θ N-1 is also invertible. Multiply both sides of the above equation by the inverse matrix of Θ N-1 on the right, and we obtain:

[0129]

[0130] Up to this point, it has been proved that the arrival delay matrix constructed from any adjacent frequency band in the constructed multi-frequency band model has the above properties. The final arrival delay matrix R TOA of the multi-frequency band is the average value of the arrival delay matrices constructed from all adjacent frequency bands.

[0131] S3. Use the obtained arrival delay matrix and adopt the multi-resolution estimation method to estimate the arrival delay.

[0132] After performing eigenvalue decomposition on the obtained arrival delay matrix R TOA , the diagonal elements of matrix Θ are the non-zero eigenvalues of the arrival delay matrix R TOA , and the corresponding column vectors of matrix V are the eigenvectors of the arrival delay matrix R TOA . In practice, since the adjacent frequency band interval B is generally much larger than the sampling interval Δf, the phase 2πBτ k is much larger than π, resulting in the aliasing of the 2π phase, that is:

[0133]

[0134] where, Θ k is the k-th diagonal element of matrix Θ, θ k is the actual phase of the k-th diagonal element, and n kThe number of periods corresponding to phase aliasing is a positive integer to be estimated. The final high-resolution arrival time delay estimation result is:

[0135]

[0136] The specific arrival time delay estimated by the multi-resolution estimation method is as follows:

[0137] S301. Arrival time delay τ k Coarse estimation

[0138] Use the sampling frequency interval Δf within the frequency band to obtain a coarse estimate of the arrival time delay. The matrix V is a Vandermonde matrix, and use the ESPRIT algorithm to obtain its corresponding rotation invariant matrix:

[0139]

[0140] Where, is the k-th diagonal element of Φ, corresponding to the k-th arrival time delay, and its coarse estimation result is

[0141] S302. Phase winding number n k Estimation

[0142] The coarse estimation and the high-resolution fine estimation should be the same theoretically, that is:

[0143]

[0144] At this time, the estimated result of n k is:

[0145]

[0146] Where, are the estimated values of n k , φ k , θ k respectively.

[0147] S303. Arrival time delay τ k Fine estimation

[0148] Use the frequency band interval B to obtain a high-resolution fine estimate of the arrival time delay, and its result is:

[0149]

[0150] Where, is the arrival time delay corresponding to the k-th propagation path.

[0151] Repeat steps S301 to S303 for the signals of different paths to estimate the arrival time delays of all paths.

[0152] In another embodiment of the present invention, a high-resolution multipath delay estimation system based on multi-frequency cooperation is provided. This system can be used to implement the above-mentioned high-resolution multipath delay estimation method based on multi-frequency cooperation. Specifically, the high-resolution multipath delay estimation system based on multi-frequency cooperation includes a sampling module, an analysis module, and an estimation module.

[0153] Among them, the sampling module constructs a multi-band system model and obtains a multi-band channel state information matrix through equidistant frequency-domain sampling for different bands.

[0154] The analysis module performs subspace analysis on the multi-band channel state information matrix obtained by the sampling module and constructs an arrival delay matrix using the coherent subspace.

[0155] The estimation module estimates the arrival delay matrix obtained by the analysis module using a multi-resolution estimation method to determine the arrival delay.

[0156] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components described and shown in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0157] Please refer to Figure 5 and Figure 6 , which respectively give the curves of the delay estimation performance varying with the signal-to-noise ratio under different band intervals in the dual-band system and different numbers of bands in the multi-band system. The estimation performance of the delay is evaluated using the root mean square error (RMSE) and the Cramér-Rao Lower Bound (CRLB). Generally, the smaller the root mean square error, the better the estimation effect. The Cramér-Rao Lower Bound sets a lower bound for all unbiased estimates, that is, the closer the RMSE of the estimation result is to the CRLB, the better the corresponding estimation effect. The relationship between RMSE and CRLB is:

[0158]

[0159] Please refer to Figure 5, taking the dual - band as an example, the influence of the frequency - band interval on the algorithm estimation performance is explored. Taking 3.5 GHz as the starting frequency band, setting the sampling interval as 240 kHz, the number of sampling points for each frequency band as 600, considering three frequency - band interval scenarios: the frequency - band interval is 1.2 GHz (S1), the frequency - band interval is 1.4 GHz (S2), and the frequency - band interval is 1.6 GHz (S3), and then 1000 Monte - Carlo experiments are carried out to analyze the time - delay estimation results corresponding to the earliest path.

[0160] From Figure 5 it can be seen that the algorithm of the present invention can approach the CRLB corresponding to the corresponding frequency - band interval under high signal - to - noise ratio, and the accuracy can reach 10 -2 ns. In addition, the larger the frequency - band interval, the better the theoretical performance, but the signal - to - noise ratio required to approach the performance bound will also increase accordingly.

[0161] Please refer to Figure 6 , and further explores the influence of different frequency - band selection schemes on the time - delay estimation performance in a multi - frequency collaborative system. Setting the total frequency - band interval as 1.2 GHz, considering three frequency - band selection schemes:

[0162] 1) Dual - band scheme, the interval between the two frequency bands is 1.2 GHz;

[0163] 2) Three - band scheme, the adjacent frequency - band interval is set as 0.6 GHz, and the total blank frequency - band interval is 1.2 GHz;

[0164] 3) Four - band scheme, the adjacent frequency - band interval is set as 0.4 GHz, and the total blank frequency - band interval is 1.2 GHz;

[0165] 4) Single - band, the total bandwidth is 1.2 GHz, and the traditional subspace method is used for estimation. Then 1000 Monte - Carlo experiments are carried out to analyze the time - delay estimation results corresponding to the earliest path.

[0166] From Figure 6 it can be seen that the algorithm of the present invention is applicable to the multi - frequency collaborative scheme and can achieve the performance under the equivalent total frequency - band interval under different frequency - band selection schemes. Further, in the three - band and four - band schemes of the present invention, the proposed algorithm can also approach the CRLB and achieve the best performance under low signal - to - noise ratio, while the traditional single - band scheme and its applicable algorithm have an order - of - magnitude difference in accuracy from the proposed algorithm, indicating the superiority of the proposed algorithm.

[0167] In summary, for the method and system for high-resolution multipath delay estimation based on multi-frequency collaboration of the present invention, after determining the multi-frequency system interval, the number of frequency bands, and obtaining the channel state information of each frequency band through parallel sampling, a high-resolution delay estimation result can be obtained according to the multi-frequency collaboration system model constructed by the present invention and the corresponding delay estimation method based on the arrival delay matrix. Moreover, the simulation shows that the present invention can achieve the theoretical performance of the system, far superior to the traditional single-frequency band signal delay estimation scheme. In addition, the present invention is applicable to existing multi-frequency band communication systems and can be processed by a parallel receiver without adding additional overhead to the system.

[0168] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0169] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocks Figure 1 one or more of the blocks.

[0170] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocks Figure 1 one or more of the blocks.

[0171] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide means for implementing the functions specified in Figure 1 one or more of the processes or multiple processes and / or blocksFigure 1 Steps of the functions specified in one or more boxes.

[0172] The above content is only to illustrate the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims

1. A high-resolution multipath delay estimation method based on multi-frequency cooperation, characterized in that, Including the following steps: S1. Construct a multi-band system model, and perform equally-spaced frequency-domain sampling for different frequency bands to obtain a multi-band channel state information matrix. The multi-band channel state information matrix is as follows: Among them, is the signal matrix corresponding to the N th frequency band, is the noise matrix in the N th frequency band, is the Vandermonde matrix composed of the time delays of multipath signals, is the additional phase matrix composed of the frequency band interval and the multipath time delay, is the multipath attenuation matrix; S2. Perform subspace analysis on the multi-band channel state information matrix obtained in step S1, and construct an arrival time delay matrix using the coherent subspace; S3. Estimate the arrival time delay matrix obtained in step S2 using the multi-resolution estimation method to determine the arrival time delay. The specific method for estimating the arrival time delay using the multi-resolution estimation method is as follows: S301. Obtain the rotation invariant matrix corresponding to the Vandermonde matrix V using the ESPRIT algorithm, and use the sampling frequency interval within the frequency band to obtain a rough estimate of the arrival time delay; S302. Estimate the phase winding number ; S303. Use frequency band interval B Obtain a fine estimate of the arrival time delay with high resolution, and repeat steps S301 to S303 for the signals of different paths to estimate the arrival time delays of all paths.

2. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 1, characterized in that, In step S1, the multi-band system model is specifically as follows: Among them, is the l th sampling data of the first frequency band, is the signal part in the l th frequency-domain sampling, represents additive white Gaussian noise with a mean of zero and a variance of , is the number of multipath delays in the channel, is the complex amplitude of the corresponding k th path, is the carrier frequency of the initial frequency band, is the sampling frequency interval, is the propagation delay of the corresponding k th path.

3. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 1, characterized in that, In step S2, the autocovariance matrix of the first frequency band is solved and eigenvalue decomposition is performed to reconstruct the signal subspace. , and the cross-covariance matrix between the adjacent high-frequency band and this frequency band is solved. , according to the signal subspace and the cross-covariance matrix to determine the first arrival time delay matrix , and then the arrival time delay matrices of all adjacent frequency bands are solved in sequence, and the average value is taken as the final arrival time delay matrix.

4. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 3, characterized in that, Arrival delay matrix is as follows: Among them, is the number of medium frequency bands in the system, then there are N -1 pairs of adjacent frequency bands, is the arrival delay matrix obtained from the channel state information matrix of the -th pair of adjacent frequency bands.

5. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 1, characterized in that, The k coarse estimate of the arrival delay is Among them, is a rotation-invariant matrix of the k th diagonal element, and the rotation-invariant matrix is: 。 6. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 1, characterized in that, Phase winding number estimated value is as follows: Among them, , are the -th k diagonal elements of the rotation-invariant matrix , and the estimated value of the actual phase of the -th k diagonal element of the matrix .

7. The high-resolution multipath delay estimation method based on multi-frequency cooperation according to claim 1, characterized in that, The k fine estimate of the arrival delay corresponding to the th propagation path is: Among them, is a matrix the k estimated value of the actual phase of the n-th diagonal element, is the estimated value of the phase winding number .

8. A high-resolution multipath delay estimation system based on multi-frequency cooperation, characterized in that, Including: Sampling module, constructing a multi-band system model, performing equally-spaced frequency-domain sampling for different frequency bands to obtain a multi-band channel state information matrix, the multi-band channel state information matrix is as follows: Among them, is the signal matrix corresponding to the N th frequency band, is the noise matrix in the N th frequency band, is the Vandermonde matrix composed of the time delays of multipath signals, is the additional phase matrix composed of the frequency band interval and the multipath time delay, is the multipath attenuation matrix; An analysis module that performs subspace analysis on the multi-band channel state information matrix obtained by the sampling module and constructs an arrival time delay matrix using the coherent subspace; An estimation module that estimates the arrival time delay matrix obtained by the analysis module using the multi-resolution estimation method to determine the arrival time delay. The specific method for estimating the arrival time delay using the multi-resolution estimation method is as follows: The rotation invariant matrix corresponding to the Vandermonde matrix V is obtained by using the ESPRIT algorithm, and the coarse estimation of the arrival time delay is obtained by using the sampling frequency interval within the frequency band. Estimate the phase wrapping number ; Use the frequency band interval B to obtain the fine estimation of the high-resolution arrival time delay. Repeat the above steps for the signals of different paths to estimate the arrival time delays of all paths.

Images