Non-circular signal sensing method based on weighted coherence matrix

Abstract

This invention provides a non-circular signal sensing method based on a weighted coherence matrix, belonging to the field of wireless communication technology. The method includes: acquiring a complex-valued received signal and calculating a real-valued augmented sample covariance matrix; constructing a real-valued sample coherence matrix based on the real-valued augmented sample covariance matrix and calculating corresponding weighting coefficients; constructing a test statistic based on the weighted sum of squares of the real-valued sample coherence matrix and the weighting coefficients; calculating a sensing threshold based on the false alarm probability and comparing the sensing threshold with the test statistic to determine the result of the spectrum sensing. This invention utilizes the real-valued augmented sample covariance matrix to describe the second-order statistical properties of non-circular signals and combines sample data-driven weighting coefficients to perform weighted square fusion of the real-valued sample coherence matrix, achieving superior sensing performance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a non-circular signal sensing method based on weighted coherence matrix. BACKGROUND

[0002] The increasing of wireless communication business leads to the problem of spectrum resource scarcity. As a basic module of cognitive radio network, spectrum sensing allows cognitive users to dynamically access the spectrum for communication without interfering with the primary users by detecting the idle state of the licensed spectrum, thereby improving the utilization efficiency of spectrum resources.

[0003] The traditional spectrum sensing method is based on the fact that the source signal obeys Gaussian distribution, and its statistics only depend on the standard covariance matrix. However, the signals in actual communication systems (such as BPSK and MSK modulated signals) often have non-circular characteristics, that is, the compensated covariance of the signal is not 0, and the augmented covariance matrix (including the standard covariance and the compensated covariance) is needed for algorithm design. Therefore, the spectrum sensing algorithm based on non-circular signal has become an important research direction to improve the utilization rate of spectrum.

[0004] For the problem of spectrum sensing of non-circular signals, researchers based on the local maximum efficacy invariant test framework, constructed a real-valued augmented covariance matrix to capture the second-order statistical properties of non-circular signals, and then proposed a sensing statistic based on the Frobenius norm of the coherence matrix. However, this statistic is expressed as the equal-gain sum of squares of the elements of the coherence matrix, which does not fully consider the numerical difference of the elements of the coherence matrix, limiting the further improvement of the detection performance. SUMMARY

[0005] In view of the above technical problems, the present application provides a non-circular signal sensing method based on weighted coherence matrix, which describes the second-order statistical properties of non-circular signals by using real-valued augmented covariance matrix, and uses data-driven weight coefficients to weight and square the elements of the coherence matrix with different numerical values, which can obtain better sensing performance.

[0006] The technical means adopted by the present application are as follows:

[0007] The non-circular signal sensing method based on weighted coherence matrix comprises:

[0008] S1, obtaining a complex-valued received signal, and calculating a real-valued augmented sample covariance matrix;

[0009] S2, based on the real-valued augmented sample covariance matrix, constructing a real-valued sample coherence matrix, and calculating the corresponding weight coefficients;

[0010] S3, based on the weighted sum of squares of the real-valued sample coherence matrix and the weight coefficients, constructing a test statistic.

[0011] S4. Calculate the sensing threshold based on the false alarm probability, and compare the magnitude of the sensing threshold with the test statistic to determine the result of spectrum sensing.

[0012] Further, step S1 includes:

[0013] S11. Obtain the complex received signal, as follows:

[0014]

[0015] in, This indicates that the spectrum is in an idle state. This indicates that the spectrum is occupied; Indicates at time The acquired complex received signal, express The source signal sent by the main user Indicates at time The noise vector follows a complex Gaussian distribution with a diagonal covariance matrix. Represents the channel gain coefficient matrix. Indicates the number of receiving antennas. Indicates the number of source signals. Indicates the sample length;

[0016] S12. Calculate the real-valued augmented sample covariance matrix as follows:

[0017]

[0018] in, Indicates received signal The real part, Indicates received signal The imaginary part, Represents the real part of the signal. Represents the imaginary part of the signal. Represents the real-valued augmented sample covariance matrix. Indicates the number of receiving antennas. Indicates sample length, superscript This represents the transpose symbol.

[0019] Further, step S2 specifically includes:

[0020] S21. Based on the real-valued augmented sample covariance matrix, calculate the real-valued sample coherence matrix as follows:

[0021]

[0022] in, Represents the real-valued augmented sample covariance matrix. Represented by matrix A diagonal matrix composed of the diagonal elements. Represents the real-valued sample coherence matrix;

[0023] S22. Calculate the corresponding weighting coefficients using the following formula:

[0024]

[0025]

[0026] in, Represents the real-valued sample coherence matrix element, This represents the unnormalized weighting coefficients. This represents the normalized weight coefficients. Indicates the number of receiving antennas. Indicates the sample length.

[0027] Furthermore, in step S3, a test statistic is constructed based on the weighted sum of squares of the real-valued sample coherence matrix and the weight coefficients, as follows:

[0028]

[0029] in, This represents the test statistic. Represents the real-valued sample coherence matrix element, Indicates correspondence Normalized weighting coefficients This represents the number of receiving antennas.

[0030] Further, step S4 specifically includes:

[0031] S41. Calculate the sensing threshold based on the false alarm probability, as follows:

[0032]

[0033] in, Indicates the perception threshold. This represents a given false alarm probability. The quantile function represents the standard normal distribution. Represents the normalized weighting coefficients. Indicates the number of receiving antennas. Indicates the sample length;

[0034] S42. Compare the magnitudes of the sensing threshold and the test statistic to determine the result of spectrum sensing, as follows:

[0035]

[0036] in, Indicates that the spectrum is idle. This indicates that the spectrum is occupied. This represents the test statistic. This represents the perception threshold.

[0037] Compared with the prior art, the present invention has the following advantages:

[0038] 1. This invention uses a real-valued augmented sample covariance matrix instead of the traditional standard sample covariance matrix, which can fully characterize the second-order statistical properties of non-circular signals (such as BPSK, MSK and other modulated signals). It overcomes the shortcomings of the standard sample covariance-based sensing method in terms of reduced performance in non-circular signal sensing, improves the recognition accuracy of non-circular signals, and significantly enhances the detection performance.

[0039] 2. This invention proposes a weighting coefficient based on sample data, which dynamically adjusts the weighting coefficient using the element values ​​of the sample coherence matrix, so that the contribution of strongly correlated signal data is greater, while suppressing the interference of noise and weakly correlated signal data.

[0040] 3. This invention uses the quantile function of the standardized normal distribution to adaptively calculate the sensing threshold and combines it with the weighted sum of squares statistic for decision-making, so that the sensing threshold matches the statistical characteristics of the signal. While ensuring a high detection probability, it can stably control the false alarm probability within the set value, thereby improving the reliability and stability of spectrum sensing.

[0041] 4. This invention achieves stronger adaptability in complex communication environments such as Rayleigh fading channels, multipath interference, and low signal-to-noise ratio (Low-SNR) by adaptive modeling of the real-valued augmented sample covariance matrix and dynamic weighting strategy. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a flowchart of the method of the present invention.

[0044] Figure 2 A comparison chart of the detection probabilities of the method of the present invention and the local maximum power invariance test method provided in the embodiments of the present invention. Detailed Implementation

[0045] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0046] It should be noted that the terms "comprising" and "having" and any variations thereof in the specification, claims and accompanying drawings of this invention are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such processes, methods, products or devices.

[0047] like Figure 1 As shown, this invention provides a non-circular signal sensing method based on a weighted coherence matrix, comprising:

[0048] S1. Obtain the complex-valued received signal and calculate the real-valued augmented sample covariance matrix;

[0049] S2. Based on the real-valued augmented sample covariance matrix, construct the real-valued sample coherence matrix and calculate the corresponding weight coefficients;

[0050] S3. Construct the test statistic based on the weighted sum of squares of the real-valued sample coherence matrix and the weight coefficients;

[0051] S4. Calculate the sensing threshold based on the false alarm probability, and compare the magnitude of the sensing threshold with the test statistic to determine the result of spectrum sensing.

[0052] In a specific implementation, as a preferred embodiment of the present invention, step S1 includes:

[0053] S11. Obtain the complex received signal, as follows:

[0054]

[0055] in, This indicates that the spectrum is in an idle state. This indicates that the spectrum is occupied; Indicates at time The acquired complex received signal, express The source signal sent by the main user Indicates at time The noise vector follows a complex Gaussian distribution with a diagonal covariance matrix. Represents the channel gain coefficient matrix. Indicates the number of receiving antennas. Indicates the number of source signals. Indicates the sample length;

[0056] S12. Calculate the real-valued augmented sample covariance matrix as follows:

[0057]

[0058] in, Indicates received signal The real part, Indicates received signal The imaginary part, Represents the real part of the signal. Represents the imaginary part of the signal. Represents the real-valued augmented sample covariance matrix. Indicates the number of receiving antennas. Indicates sample length, superscript This represents the transpose symbol.

[0059] In a specific implementation, as a preferred embodiment of the present invention, step S2 includes:

[0060] S21. Based on the real-valued augmented sample covariance matrix, calculate the real-valued sample coherence matrix as follows:

[0061]

[0062] in, Represents the real-valued augmented sample covariance matrix. Represented by matrix A diagonal matrix composed of the diagonal elements. Represents the real-valued sample coherence matrix;

[0063] S22. Calculate the corresponding weighting coefficients using the following formula:

[0064]

[0065]

[0066] in, Represents the real-valued sample coherence matrix element, This represents the unnormalized weighting coefficients. This represents the normalized weight coefficients. Indicates the number of receiving antennas. Indicates the sample length.

[0067] In a specific implementation, as a preferred embodiment of the present invention, in step S3, a test statistic is constructed based on the weighted sum of squares of the real-valued sample coherence matrix and the weight coefficients, as follows:

[0068]

[0069] in, This represents the test statistic. Represents the real-valued sample coherence matrix element, Indicates correspondence Normalized weighting coefficients This represents the number of receiving antennas.

[0070] In a specific implementation, as a preferred embodiment of the present invention, step S4 specifically includes:

[0071] S41. Calculate the sensing threshold based on the false alarm probability, as follows:

[0072]

[0073] in, Indicates the perception threshold. This represents a given false alarm probability. The quantile function represents the standard normal distribution. Represents the normalized weighting coefficients. Indicates the number of receiving antennas. Indicates the sample length;

[0074] S42. Compare the magnitudes of the sensing threshold and the test statistic to determine the result of spectrum sensing, as follows:

[0075]

[0076] in, Indicates that the spectrum is idle. This indicates that the spectrum is occupied. This represents the test statistic. This represents the perception threshold.

[0077] Example

[0078] The embodiments of the present invention will be described in detail below through simulation experiments.

[0079] Simulation parameter settings: The simulation experiment assumes a Rayleigh fading channel, i.e., the channel gain matrix... The column components follow a standard complex Gaussian distribution that is independent and identically distributed. Once generated in a Monte Carlo experiment, they remain unchanged, but are randomly generated again in subsequent Monte Carlo experiments. All simulation results were obtained through 50,000 Monte Carlo experiments.

[0080] like Figure 2 The figure shows a comparison of the detection probabilities of the method of this invention and the local maximum power invariance test method based on non-circular signals under different signal-to-noise ratio environments. The background noise variance follows a uniform distribution, and the sample length is... Number of receiving antennas The main user signal is a non-circular signal modulated by BPSK, and the number of signals... The corresponding power is dB, false alarm probability is It can be seen that the sensing method of this invention is superior to the local maximum power invariance test method based on non-circular signals, and exhibits better detection performance.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A non-circular signal sensing method based on weighted coherence matrix, characterized in that, By using real-valued augmented covariance matrix to describe the second-order statistical properties of non-circular signals, and using data-driven weight coefficients to weight and squarely fuse different numerical coherent matrix elements, better sensing performance is obtained, including: S1, obtaining a complex-valued received signal, and calculating a real-valued augmented sample covariance matrix; S2, based on the real-valued augmented sample covariance matrix, constructing a real-valued sample coherence matrix, and calculating the corresponding weight coefficients; the weight coefficients are dynamically adjusted by using the sample coherence matrix element values, so that the contribution of strong correlation signal data is greater, and the interference of noise and weak correlation signal data is suppressed, including: S21, based on the real-valued augmented sample covariance matrix, calculating the real-valued sample coherence matrix, as follows: in, Represents the real-valued augmented sample covariance matrix. Represented by matrix A diagonal matrix composed of the diagonal elements. Represents the real-valued sample coherence matrix; S22, calculating the corresponding weight coefficients, as follows: wherein, represents an element of a real-valued sample coherence matrix, represents a non-normalized weight coefficient, represents a normalized weight coefficient, represents a number of receive antennas, represents a sample length;​ S3, based on the weighted sum of the real-valued sample coherence matrix and the weight coefficients, constructing a test statistic; S4, calculating a sensing threshold based on a false alarm probability, and comparing the sensing threshold with the test statistic to determine the result of spectrum sensing.

2. The weighted coherence matrix based non-circular signal aware method of claim 1, wherein, Step S1 includes: S11, obtaining a complex-valued received signal, as follows: wherein, represents that the spectrum is in an idle state, represents that the spectrum is in an occupied state; represents the time instant the complex-valued received signal, represents the source signal transmitted by the represents the time instant noise vector, subject to a complex Gaussian distribution with a diagonal covariance matrix, represents the channel gain coefficient matrix, represents the number of receive antennas, represents the number of source signals, represents the sample length; S12, calculating a real-valued augmented sample covariance matrix, as follows: wherein denotes the real part of the received signal , denotes the imaginary part of the received signal , denotes the real part of the signal , denotes the real-valued augmented sample covariance matrix denotes the number of receive antennas denotes the sample length, the superscript denotes the transposition symbol.

3. The weighted coherence matrix based non-circular signal aware method of claim 1, wherein, In step S3, based on the weighted sum of the real-valued sample coherence matrix and the weight coefficients, a test statistic is constructed, as follows: wherein, denotes a test statistic, denotes the element of the real-valued sample coherence matrix, denotes the normalized weight coefficient corresponding to the is the number of receive antennas.

4. The weighted coherence matrix based non-circular signal aware method of claim 1, wherein, Step S4 specifically includes: S41, calculating a sensing threshold based on a false alarm probability, as follows: wherein denotes a perception threshold, denotes a given false alarm probability, denotes a quantile function of a standard normal distribution, denotes a normalization weight coefficient, denotes a number of receive antennas, denotes a sample length; S42, comparing the sensing threshold with the test statistic to determine the result of spectrum sensing, as follows: wherein, represents a spectrum free state, represents a spectrum occupied state, represents a test statistic, represents a sensing threshold.

Images