Broadband spectrum sensing method and system in multipath propagation environment

Abstract

The invention discloses a broadband spectrum sensing method and system in a multipath propagation environment, and relates to the technical field of wireless communication. Comprising the following steps: constructing a dual-channel sampling structure, and obtaining multi-band signal sampling data in a multipath propagation environment; establishing a broadband spectrum sensing model based on the sampling data; performing tensor decomposition on the model to obtain a factor matrix; and estimating narrowband signal parameters based on the factor matrix. According to the invention, a dual-channel sampling structure is constructed, and the sub-Nyquist sampling rate is adopted, so that the sampling rate is effectively reduced, and the data storage and processing burden is reduced; by establishing a broadband spectrum sensing model based on sampling data, the accuracy of spectrum sensing is improved; tensor decomposition is carried out on the model, so that effective separation of a complex multipath signal structure is realized; by estimating the narrowband signal parameters based on the factor matrix, accurate estimation of the carrier frequency and the power spectrum is realized, and a reliable basis is provided for spectrum access of a secondary user.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a broadband spectrum sensing method and system in a multipath propagation environment. Background Art

[0002] Cognitive radio (CR) is considered to be one of the key technologies of the next generation of wireless communication systems. The system allows secondary users to opportunistically access the authorized spectrum without interfering with the primary user, which can greatly improve the spectrum utilization rate and is an effective means to alleviate the shortage of spectrum resources. The CR system needs to monitor the use of the spectrum periodically. However, directly performing Nyquist sampling on a large bandwidth spectrum (such as the spectrum of up to tens of gigahertz involved in terahertz communication) will cause a heavy burden on the storage and processing of a large number of samples due to the high sampling rate. In this case, efficient broadband spectrum sensing technology is an important problem that the CR system needs to overcome. Traditional broadband spectrum sensing schemes all assume that the receiver receives multi-band signals through a line-of-sight (LOS) path. However, in actual applications, multi-band signals and receivers are often propagated in multiple paths, especially in application scenarios with rich scattering environments such as dense urban areas. Therefore, in this case, the broadband spectrum sensing solution based on the LOS propagation path assumption will experience serious performance degradation. To solve this problem, the present invention proposes an efficient broadband spectrum sensing scheme suitable for multi-path scenarios. Summary of the invention

[0003] In view of the problems existing in the prior art, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is how to realize the monitoring and perception of broadband spectrum at a sampling rate far lower than the Nyquist sampling frequency in a multipath propagation environment. Compared with the existing broadband spectrum sensing solution based on the LOS propagation assumption, the proposed method has better spectrum reconstruction and carrier frequency estimation accuracy in different signal-to-noise ratio environments, especially in low signal-to-noise ratio environments.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a method for broadband spectrum sensing in a multipath propagation environment, which includes constructing a dual-channel sampling structure to obtain multi-band signal sampling data in the multipath propagation environment;

[0007] Establish a broadband spectrum sensing model based on sampled data;

[0008] Perform tensor decomposition on the model to obtain the factor matrix;

[0009] Estimate narrowband signal parameters based on factor matrix.

[0010] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, the construction of a dual-channel sampling structure specifically includes:

[0011] A receiver with M antennas is used, wherein each antenna is connected to a direct sampling channel and a delayed sampling channel with a fixed time delay Δ;

[0012] The analog signals of the two sampling channels are sent to a synchronized analog-to-digital converter ADC for sampling at a sub-Nyquist sampling rate;

[0013] The sampling time interval satisfies T s represents the sampling time, f nyq is the Nyquist sampling rate, and the delay factor Δ of the delayed sampling channel needs to satisfy and The sampling rate is not less than the maximum bandwidth of these sources, that is, f s ≥B;

[0014] There are K far-field narrowband signals in the frequency band to be detected. These narrowband signals are wide-sense stationary and uncorrelated.

[0015] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, establishing a broadband spectrum sensing model based on sampled data specifically includes:

[0016] Get the multi-band signal model formed by K far-field narrowband signals:

[0017]

[0018] Among them, f k and represent the carrier frequency and baseband signal of the kth far-field narrowband source, respectively. are uncorrelated and are all band-limited signals, {s k (t)} are all narrow-band signals with different power spectra;

[0019] Considering the multipath propagation environment, the received signal at the receiving end is obtained:

[0020] x(t)=Hs(t)+w(t)

[0021] in, represents the channel between the radiation source and the base station, x m (t) represents the received signal of the mth antenna at time t, represents the set of K modulated signals at time t, w(t) represents the covariance σ2 Complex Gaussian white noise of I;

[0022] Sample the received signal to obtain the sampling data of the direct sampling channel:

[0023] x[l]=HE(lT s )s b [l]+w[l]

[0024] in,

[0025] The sampling data of the time-delay sampling channel can be expressed as:

[0026] y[l]=HE(lT s -Δ)s b (lT s -Δ)+w(lT s -Δ)

[0027] ≈HE(t-Δ)s b [l]+w′[l]

[0028] In the above formula, w′[l]=w(lT s -Δ),

[0029] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, wherein: obtaining the cross-correlation covariance of the sampling data of the direct sampling channel includes:

[0030] Direct sampling channel at any two times l1T s Moment and l2T s The cross-correlation covariance of the sampled data at time is:

[0031] R x (l1,l2)οE[x[l1]x H [l2]]

[0032] =HE(τ)Rb(τ)H H +E[w[l1]w H [l2]]

[0033] The cross-correlation covariance between the direct sampling channel and the delayed sampling channel sample data is:

[0034]

[0035] in, R b (τ) is a diagonal matrix as shown below:

[0036] R b (τ)=E[sb [l1]s b [l2]]=diag(r b (τ)

[0037] r b (τ) is defined as

[0038] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, the tensor decomposition of the model specifically includes:

[0039] Will and Represented as a third-order tensor The three modes of the tensor R represent the antenna index and sampling time difference index of the receiver;

[0040] Using κ to represent the sampling time difference index mode, then

[0041]

[0042] Among them, S = S1 ∪ S2;

[0043] Note that the slice of a tensor can be represented as:

[0044]

[0045] Among them, λ κ，k It is expressed as:

[0046]

[0047] Wherein, "ο" represents the vector outer product,

[0048]

[0049] in, Defined as The factor matrix {H,H * ,Λ} is obtained by performing CP decomposition on it. Λ contains the information of the carrier frequency and power spectrum of the far-field narrowband signal. The estimated value of Λ is used to estimate the carrier frequency and power spectrum of these source signals.

[0050] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, wherein: estimating narrowband signal parameters based on the factor matrix includes:

[0051] Estimate the carrier frequency of the kth far-field narrowband signal:

[0052]

[0053] Where β is defined as:

[0054]

[0055] mod(a,b) represents the modulus operator;

[0056]

[0057] When L is large enough, the approximate error in the formula can be ignored.

[0058] As a preferred solution of the broadband spectrum sensing method in a multipath propagation environment of the present invention, the uniqueness condition of the tensor decomposition is:

[0059] When K is known a priori, the CP decomposition of R involves the following optimization problem:

[0060]

[0061] like When k A represents the Kruskal rank of the matrix A;

[0062] In a multipath propagation environment, H is usually modeled as a random matrix composed of independent complex Gaussian terms. In this case, H is full rank, that is, P(k H =K)=1;

[0063] When k Λ When ≥2, the uniqueness condition of CP decomposition is satisfied.

[0064] In a second aspect, an embodiment of the present invention provides a broadband spectrum sensing system in a multipath propagation environment, which includes a sampling structure module, a broadband spectrum sensing model module, a tensor decomposition module, and a parameter estimation module;

[0065] The sampling structure module is used to construct a dual-channel sampling structure to obtain multi-band signal sampling data in a multipath propagation environment;

[0066] The wideband spectrum sensing model module is used to establish a wideband spectrum sensing model based on sampling data;

[0067] The tensor decomposition module is used to decompose the model into a factor matrix;

[0068] The parameter estimation module is used to estimate narrowband signal parameters based on the factor matrix.

[0069] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the broadband spectrum sensing method in a multipath propagation environment of the first aspect of the present invention are implemented.

[0070] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the broadband spectrum sensing method in a multipath propagation environment of the first aspect of the present invention are implemented.

[0071] The invention has the following beneficial effects: the broadband spectrum sensing method in a multipath propagation environment provided by the invention obtains multi-band signal sampling data in a multipath propagation environment by constructing a dual-channel sampling structure, each antenna is connected to a direct sampling channel and a delayed sampling channel with a fixed delay, and sampling is performed at a sub-Nyquist sampling rate, which effectively reduces the sampling rate and reduces the burden of data storage and processing; by establishing a broadband spectrum sensing model based on sampling data, the influence of the multipath propagation environment is considered, a signal model that is more in line with the actual scene is established, and the accuracy of spectrum sensing is improved; by performing tensor decomposition of the model to obtain a factor matrix, the effective separation of the complex multipath signal structure is achieved, and the uniqueness of the decomposition result under certain conditions is proved, and the reliability of the result is guaranteed; by estimating narrowband signal parameters based on the factor matrix, the accurate estimation of the carrier frequency and power spectrum is achieved, and even if there is scale ambiguity, the frequency band occupancy can be effectively judged, providing a reliable basis for the spectrum access of secondary users. The invention has achieved good results in sampling efficiency, model accuracy, signal separation and parameter estimation, and is particularly suitable for application scenarios with rich scattering environments such as dense urban areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0073] Figure 1 A flowchart of a broadband spectrum sensing method in a multipath propagation environment;

[0074] Figure 2 It is a curve diagram of carrier frequency NMAE versus SNR of the broadband spectrum sensing method in a multipath propagation environment;

[0075] Figure 3 This is a comparison chart of the original power spectrum and the estimated power of the broadband spectrum sensing method in a multipath propagation environment when the SNR is 10dB. DETAILED DESCRIPTION

[0076] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0077] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0078] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0079] Example 1

[0080] Reference Figure 1 to Figure 3 , is the first embodiment of the present invention, which provides a broadband spectrum sensing method in a multipath propagation environment, including:

[0081] For the problem of wideband spectrum sensing in a multipath propagation environment, consider a scenario where a receiver with M antennas is used to handle the problem of wideband spectrum sensing. Assume that there are K (K≤M) far-field narrowband signals in the frequency band to be monitored, and these narrowband signals are wide-sense stationary and uncorrelated.

[0082] S1: Construct a dual-channel sampling structure to obtain multi-band signal sampling data in a multipath propagation environment. The multi-band signal formed by K far-field narrowband signals can be modeled as:

[0083]

[0084] Among them, f k and represent the carrier frequency and baseband signal of the kth far-field narrowband source, respectively. are uncorrelated and are all band-limited signals, {s k (t)} are all narrow-band signals with different power spectra;

[0085] Considering the multipath propagation environment, the received signal at the receiving end is obtained:

[0086] x(t)=Hs(t)+w(t)

[0087] in, represents the channel between the radiation source and the base station, represents the received signal of the mth antenna at time t, represents the set of K modulated signals at time t, w(t) represents the covariance σ 2 Complex Gaussian white noise of I;

[0088] In order to achieve the goal of broadband spectrum sensing, the present invention adopts a sub-Nyquist sampling structure based on a phased array. Each antenna is connected to two channels, namely a direct sampling channel and a delayed sampling channel with a fixed delay. The analog signal passes through these two sampling channels and is sampled by a synchronized analog-to-digital converter (ADC) at a sub-Nyquist sampling rate, i.e. T s represents the sampling time, f nyq is the Nyquist sampling rate, and the delay factor Δ of the delayed sampling channel needs to satisfy and The sampling rate is not less than the maximum bandwidth of these sources, that is, f s ≥B;

[0089] definition as well as Then the sampling data of the direct sampling channel at the moment is

[0090] x[l]=HE(lT s )s b [l]+w[l]

[0091] in, Similarly, the sampling data of the time-delay sampling channel can be expressed as

[0092] y[l]=HE(lT s -Δ)s b (lT s -Δ)+w(lT s -Δ)

[0093] ≈HE(t-Δ)s b [l]+w′[l]

[0094] In the above formula, w′[l]=w(lT s -Δ), because Therefore, the above formula is approximately valid.

[0095] S2: Establish a wideband spectrum sensing model based on sampled data. First, consider the samples of the sampling channel directly. Given any two moments, such as l1T s Moment and l2T s At time , the cross-correlation covariance of these two samples can be written as

[0096] R x (l1,l2)οE[x[l1]x H [l2]]

[0097] =HE(τ)R b (τ)H H +E[w[l1]w H [l2]]

[0098] Among them, among them, R b (τ) is a diagonal matrix as shown below:

[0099] R b (τ)=E[s b [l1]s b [l2]]=diag(r b (τ)

[0100] r b (τ) is defined as

[0101] Obviously, R x (l1,l2) depends only on the sampling time difference τ, so R x (τ) to represent R x (l1,l2). Considering w(t) is white Gaussian noise, it can be simplified to

[0102]

[0103] Where E(0) is an identity matrix. Definition for

[0104]

[0105] Similarly, the cross-correlation between the sampled data of the direct sampling channel and the delayed sampling channel can be given by

[0106]

[0107] According to the definition of tensor, and It can be expressed as a third-order tensor The three modes of the tensor R represent the antenna index and sampling time difference index of the receiver. Using κ to represent the sampling time difference index mode, then, Further, define

[0108]

[0109] Among them, S = S1 ∪ S2;

[0110]

[0111]

[0112] Wherein, "ο" represents the vector outer product, Therefore, the tensor R can be decomposed by CP as follows:

[0113]

[0114] in, Defined as The factor matrix {H,H * ,Λ} is obtained by performing CP decomposition on it. Λ contains the information of the carrier frequency and power spectrum of the far-field narrowband signal. The estimated value of Λ is used to estimate the carrier frequency and power spectrum of these source signals.

[0115] S3: Perform tensor decomposition on the model to obtain the factor matrix;

[0116] First consider the case where K is known a priori. In this case, the CP decomposition of R involves the following optimization problem:

[0117]

[0118] As will be discussed later, the CP decomposition of R is unique. Therefore, an auxiliary matrix A is used instead of H*. Then, the optimization problem in the formula is effectively solved by the following iterative process:

[0119]

[0120] In the formula, i represents the number of iterations, R (x) represents the x-th modal expansion of R.

[0121] When K is unknown, some modified CP decomposition methods can be used to obtain the factor matrix of R. At this time, the optimization problem involved in the CP decomposition of the third-order tensor is:

[0122]

[0123] In the formula, K is the estimated tensor rank, and μ is a parameter used to weigh the low-rank property of the tensor R and the degree of data fitting. The formula can be similarly solved using the iterative least squares method, and the least squares optimization problem involved is:

[0124]

[0125] get After estimating the value, the rank of the third-order tensor R can be reduced by removing The items with smaller eigenvalues ​​are determined.

[0126] The uniqueness of CP decomposition is a necessary condition for the method proposed in the present invention.

[0127]

[0128] Established, where k A denotes the Kruskal rank of the matrix A, then the CP decomposition of R is essentially unique except for ambiguities in scaling and permutation order. * is the complex conjugate of H, so they have the same Kruskal rank, and for this purpose we only need to verify the Kruskal rank of H. In a multipath propagation environment, H is usually modeled as a random matrix consisting of independent complex Gaussian terms. In this case, H is column-full rank, so k H =K almost satisfies, that is, P(k H =K) = 1. In addition, assuming that any two baseband signals have different power spectra, this means that any two columns of Λ are linearly independent, that is, k Λ ≥ 2. Based on the above discussion, it can be concluded that the uniqueness condition of CP decomposition can be satisfied.

[0129] S4: Estimate narrowband signal parameters based on the factor matrix.

[0130] As before, the CP decomposition of the third-order tensor has both scaling ambiguity and permutation ambiguity. k The sum power spectrum of} depends only on and Each column of corresponds to a specific source signal, so the permutation ambiguity in CP decomposition can be ignored. Considering the scaling ambiguity, Each element in the kth column of

[0131]

[0132] Among them, φ k and a k are the phase ambiguity and amplitude ambiguity caused by the CP decomposition scaling ambiguity. Considering Δ·f k <1, carrier frequency f k It can be estimated as:

[0133]

[0134] Where β is defined as:

[0135]

[0136] mod(a,b) represents the modulus operator. For z=re jφ We have arg(z)=mod(φ,2π).

[0137] The power spectrum of the far-field radiation source involved can be obtained by DFT transform of the autocorrelation sequence.

[0138]

[0139] When L is large enough, the approximate error in the formula can be ignored. The estimated value of is:

[0140]

[0141] Due to the blurring of the scale caused by CP decomposition, a k and φ k is unknown. Simply use replace to estimate the power spectrum. However, this alternative approach results in the scaling of the power spectrum of the kth source signal. Despite the scaling ambiguity, the wide spectrum sensing task can still be considered accomplished because the above method can determine which frequency bands are not occupied and which frequency bands are available for secondary users.

[0142] Figure 2 The curve of NMAE changing with SNR is plotted in s =10 5 It can be seen that the proposed method performs the same as the comparison algorithm when the SNR is large, while it significantly improves the estimation accuracy in the low SNR area.

[0143] Figure 3 It shows that when the SNR is 10dB and N s =10 5 The original power spectrum and the reconstructed power spectrum are realized at one time. It can be seen that despite a certain degree of scaling ambiguity, the proposed method can accurately recover the considered wideband power spectrum at a sampling rate much lower than the Nyquist sampling rate. Although CP decomposition and unknown channels will bring amplitude deviations to the estimated power spectrum, it can provide idle spectrum information for secondary access.

[0144] Furthermore, this embodiment also provides a broadband spectrum sensing system in a multipath propagation environment, including a sampling structure module, a broadband spectrum sensing model module, a tensor decomposition module, and a parameter estimation module;

[0145] The sampling structure module is used to construct a dual-channel sampling structure to obtain multi-band signal sampling data in a multipath propagation environment;

[0146] The wideband spectrum sensing model module is used to establish a wideband spectrum sensing model based on sampling data;

[0147] The tensor decomposition module is used to decompose the model into a factor matrix;

[0148] The parameter estimation module is used to estimate narrowband signal parameters based on the factor matrix.

[0149] This embodiment also provides a computer device, which is suitable for the case of a broadband spectrum sensing method in a multipath propagation environment, and includes a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the broadband spectrum sensing method in a multipath propagation environment proposed in the above embodiment.

[0150] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0151] This embodiment further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing wideband spectrum sensing in a multipath propagation environment as proposed in the above embodiment is implemented.

[0152] In summary, the broadband spectrum sensing method in a multipath propagation environment provided by the present invention obtains multi-band signal sampling data in a multipath propagation environment by constructing a dual-channel sampling structure. Each antenna is connected to a direct sampling channel and a delayed sampling channel with a fixed delay, and sampling is performed at a sub-Nyquist sampling rate, which effectively reduces the sampling rate and reduces the burden of data storage and processing; by establishing a broadband spectrum sensing model based on sampling data, the influence of the multipath propagation environment is considered, and a signal model that is more in line with the actual scene is established, which improves the accuracy of spectrum sensing; by performing tensor decomposition of the model to obtain a factor matrix, the effective separation of the complex multipath signal structure is achieved, and the uniqueness of the decomposition result under certain conditions is proved, ensuring the reliability of the result; by estimating narrowband signal parameters based on the factor matrix, the accurate estimation of the carrier frequency and power spectrum is achieved, and even if there is scale ambiguity, the frequency band occupancy can be effectively judged, providing a reliable basis for the spectrum access of secondary users. The present invention has achieved good results in sampling efficiency, model accuracy, signal separation and parameter estimation, and is particularly suitable for application scenarios with rich scattering environments such as dense urban areas.

[0153] Example 2

[0154] The second embodiment of the present invention provides a broadband spectrum sensing method in a multipath propagation environment. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0155] In the simulation, it is assumed that the receiver is equipped with M = 8 antennas and the ΔT of different antennas is set to 10 -9 seconds. Let U(a,b) be a uniform distribution on the interval [a,b]. Consider K = 3 far-field sources, whose carrier frequencies and corresponding bandwidths are set to f1 = U(50,150) MHz, f2 = U(200,300) MHz, f3 = U(350,450) MHz, B1 = 10 MHz, B2 = 15 MHz, and B3 = 20 MHz. Each far-field source signal reaches the receiver through multiple paths. The channel matrix H can be represented by the following geometric model:

[0156]

[0157] where i k ∈{1,…,I k} represents the index of the signal path, as well as Respectively represent the i k The complex gain and direction of arrival angle associated with each path are given by d(·) and d(·) represents the array corresponding vector. In the experiment, the number of paths that different signals pass through is randomly selected from [10,50], and the direction of arrival angle corresponding to each path is uniformly selected from (0,π). The complex gain of each path is randomly generated from CN(0,1) and normalized to The complex baseband signal is generated by band-limited Gaussian noise, which is obtained by low-pass filtering with different cutoff frequencies. The sampling frequency is set to f s = 28 MHz, slightly higher than 20 MHz. L is set to 95, which is large enough to satisfy the approximate calculation of the power spectrum. SNR is defined as E(|s(t)| 2 ) / σ 2 In the experiment, the normalized mean absolute error (NMAE) is used to measure the estimation accuracy of the carrier frequency. NMAE is defined as where f k and are the true frequency and estimated frequency of the kth user respectively, and the Nyquist sampling frequency is f nyq =1GHz.

[0158] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A broadband spectrum sensing method in a multipath propagation environment, characterized by: include, Construct a dual-channel sampling structure to obtain multi-band signal sampling data in a multipath propagation environment; Establishing a broadband spectrum sensing model based on the sampled data; Performing tensor decomposition on the model to obtain a factor matrix; Narrowband signal parameters are estimated based on the factor matrix.

2. The broadband spectrum sensing method in a multipath propagation environment according to claim 1, characterized in that: The construction of the dual-channel sampling structure specifically includes: A receiver with M antennas is used, wherein each antenna is connected to a direct sampling channel and a delayed sampling channel with a fixed time delay Δ; Sending the analog signals of the two sampling channels to a synchronized analog-to-digital converter ADC for sampling at a sub-Nyquist sampling rate; The sampling time interval satisfies T s represents the sampling time, f nyq is the Nyquist sampling rate, and the delay factor Δ of the delayed sampling channel needs to satisfy and The sampling rate is not less than the maximum bandwidth of these sources, that is, f s ≥B; There are K far-field narrowband signals in the frequency band to be detected. These narrowband signals are wide-sense stationary and uncorrelated.

3. The broadband spectrum sensing method in a multipath propagation environment according to claim 2, characterized in that: The establishing of a broadband spectrum sensing model based on sampled data specifically includes: Get the multi-band signal model formed by K far-field narrowband signals: Among them, f k and represent the carrier frequency and baseband signal of the kth far-field narrowband source, respectively. are uncorrelated and are all band-limited signals, {s k (t)} are all narrow-band signals with different power spectra; Considering the multipath propagation environment, the received signal at the receiving end is obtained: x(t)=Hs(t)+w(t) in, represents the channel between the radiation source and the base station, x m (t) represents the received signal of the mth antenna at time t, represents the set of K modulated signals at time t, w(t) represents the covariance σ 2 Complex Gaussian white noise of I; The received signal is sampled to obtain sampling data of a direct sampling channel: x[l]=HE(lT s )s b [l]+w[l] in, The sampling data of the time-delay sampling channel can be expressed as: y[l]=HE(lT s -Δ)s b (lT s -Δ)+w(lT s -Δ) ≈HE(t-Δ)s b [l]+w′[l] In the above formula, w′[l]=w(lT s -Δ), 4. The broadband spectrum sensing method in a multipath propagation environment as claimed in claim 3, characterized in that: The obtaining of the cross-correlation covariance of the sampling data of the direct sampling channel comprises: Direct sampling channel at any two times l1T s Moment and l2T s The cross-correlation covariance of the sampled data at time is: The cross-correlation covariance between the direct sampling channel and the delayed sampling channel sample data is: in, R b (τ) is a diagonal matrix as shown below: R b (τ)=E[s b [l1]s b [l2]]=diag(r b (τ)) r b (τ) is defined as 5. The broadband spectrum sensing method in a multipath propagation environment according to claim 4, characterized in that: The tensor decomposition of the model specifically includes: Will and Represented as a third-order tensor The three modes of the tensor R represent the antenna index and sampling time difference index of the receiver; Using κ to represent the sampling time difference index mode, then Among them, S = S1 ∪ S2; Note that the slice of a tensor can be represented as: Among them, λ κ , k It is expressed as: in, represents the vector outer product, in, Defined as The factor matrix {H,H * ,Λ} is obtained by performing CP decomposition on it. Λ contains the information of the carrier frequency and power spectrum of the far-field narrowband signal. The estimated value of Λ is used to estimate the carrier frequency and power spectrum of these source signals.

6. The broadband spectrum sensing method in a multipath propagation environment according to claim 5, characterized in that: The estimating narrowband signal parameters based on the factor matrix comprises: Estimate the carrier frequency of the kth far-field narrowband signal: Where β is defined as: mod(a,b) represents the modulus operator; When L is large enough, the approximate error in the formula can be ignored.

7. The broadband spectrum sensing method in a multipath propagation environment according to claim 6, characterized in that: The uniqueness condition of the tensor decomposition is: When K is known a priori, the CP decomposition of R involves the following optimization problem: like When k A represents the Kruskal rank of the matrix A; In a multipath propagation environment, H is usually modeled as a random matrix composed of independent complex Gaussian terms. In this case, H is full rank, that is, P(k H =K)=1; When k Λ When ≥2, the uniqueness condition of CP decomposition is satisfied.

8. A broadband spectrum sensing system in a multipath propagation environment, based on the broadband spectrum sensing method in a multipath propagation environment according to any one of claims 1 to 7, characterized in that: It also includes a sampling structure module, a broadband spectrum sensing model module, a tensor decomposition module, and a parameter estimation module; The sampling structure module is used to construct a dual-channel sampling structure to obtain multi-band signal sampling data in a multipath propagation environment; The broadband spectrum sensing model module is used to establish a broadband spectrum sensing model based on the sampled data; The tensor decomposition module is used to perform tensor decomposition on the model to obtain a factor matrix; The parameter estimation module is used to estimate narrowband signal parameters based on the factor matrix.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the broadband spectrum sensing method in a multipath propagation environment according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the broadband spectrum sensing method in a multipath propagation environment according to any one of claims 1 to 7 are implemented.

Images