Television station frequency modulation rebroadcasting system based on improved spectrum sensing method

Abstract

The invention discloses a television station frequency modulation rebroadcasting system based on an improved spectrum sensing method. In the MEHM method, a simple asymptotic substitution method is adopted to solve the calculation of a signal sensing judgment threshold, so that a large deviation is generated between a theoretical judgment threshold and a real judgment threshold, and the improvement of the sensing detection performance of the method is limited. Therefore, the invention provides a television station frequency modulation rebroadcasting system based on an improved spectrum sensing method. According to the system, firstly, a sample covariance matrix of a received signal is calculated by constructing a covariance matrix calculation module; constructing a characteristic value and detection statistic module, and calculating the maximum characteristic value of the sample covariance matrix and the harmonic average of the maximum characteristic value; and finally, constructing a spectrum sensing judgment module, thereby calculating a signal sensing judgment threshold, comparing whether the detection statistic is greater than the judgment threshold, if so, determining that the frequency band has a master user, and otherwise, determining that the frequency band does not have the master user. The system does not need priori knowledge of a main user signal and a wireless channel, and can effectively overcome the influence of noise uncertainty. Compared with some spectrum sensing methods based on characteristic value detection, the improved spectrum sensing method has the advantages that the sensing detection probability is higher, the sensing detection performance more excellent than that of a maximum characteristic value-harmonic average (MEHM) method is obtained, and the improved spectrum sensing method is more suitable for judging frequency modulation rebroadcasting signals in a television station; and the stability and the reliability of the television station frequency modulation rebroadcasting system are improved.

Description

Technical Field

[0001] This invention relates to the application of cognitive radio technology in television FM relay systems, specifically an FM relay system based on an improved spectrum sensing method. This method can provide a more efficient spectrum utilization solution for FM relay equipment and other wireless communication devices, improving overall communication system performance and resource utilization without interfering with normal FM relay services. Background Art

[0002] Cognitive radio (CR) technology, an innovation in wireless communications, has garnered significant attention in recent years for its ability to efficiently address the issue of limited signal channel resources. Spectrum sensing (SS), a core component of CR technology, is particularly crucial in FM broadcasting scenarios. The transmission of FM broadcast signals requires stable spectrum resources. However, in the complex environments of signal broadcasting, spectrum resource utilization is often unsatisfactory.

[0003] Among the large family of eigenvalues-based detection (EBD) methods, the Maximum Eigenvalue-Harmonic Mean (MEHM) detection method stands out for its unique decision metric and excellent detection performance. The MEHM method calculates the ratio of the maximum eigenvalue to the eigenvalue harmonic mean of the covariance matrix (SCM) of the TV station's FM broadcast signal as the perceptual decision metric. This decision metric can keenly capture the subtle differences in the presence or absence of the TV station's FM broadcast signal. This excellent detection capability enables the MEHM method to effectively identify idle frequency bands for TV station FM broadcast signals, providing reliable technical support for spectrum resource management and optimization of TV station FM broadcast systems.

[0004] However, although the MEHM method has achieved significant improvements in detection performance, the decision threshold of the TV station FM broadcast signal has become an urgent problem to be solved. In the original MEHM method, in order to simplify the calculation, a simple asymptotic substitution method is used to determine the decision threshold of the TV station FM broadcast signal, that is, the harmonic mean of the SCM eigenvalues ​​is simply replaced by the noise variance σ 2 However, with the latest progress in random matrix theory (RMT), in the detection of FM broadcast signals of TV stations, the maximum eigenvalue of SCM will significantly deviate from the noise variance σ when the number of samples is small. 2This results in significant deviations in the calculation of the threshold for determining FM relay signals in the MEHM method. This deviation not only limits the method's performance in detecting FM relay signals, but can also lead to false positives and missed positives, thus affecting the stability and reliability of the FM relay system.

[0005] To overcome this challenge, the present invention utilizes the distribution of SCM extreme eigenvalues ​​to design a TV station FM relay system based on an improved spectrum sensing method. This system more accurately calculates the decision threshold for TV station FM relay signals, resulting in a more reliable theoretical threshold calculation. By optimizing the decision threshold for FM relay signals, the TV station FM relay system based on the improved spectrum sensing method can maintain the superior detection performance of the MEHM method while further improving its detection accuracy under challenging conditions for TV station FM relay signal detection. This provides a more solid theoretical foundation and technical support for the application of CR technology in practical scenarios such as TV station FM. Summary of the Invention

[0006] In view of this, the technical problem solved by the present invention is that a TV station FM broadcasting system based on an improved spectrum sensing method can effectively overcome the influence of noise uncertainty in the spectrum broadcasting process, and obtain better detection performance than the maximum eigenvalue-harmonic mean (MEHM) method, and better detect the presence or absence of primary user (PU) signals, so that secondary users (SU) can opportunistically access idle spectrum resources, thereby improving the utilization rate of spectrum resources.

[0007] The technical solutions adopted by the present invention to solve the technical problems are as follows:

[0008] A TV station FM broadcasting system based on an improved spectrum sensing method, comprising:

[0009] Step 1) constructing a radio signal covariance matrix calculation module, which calculates the covariance matrix of the TV station FM broadcast signal sample;

[0010] Step 2) constructing a spectrum sensing eigenvalue and statistics calculation module, which obtains the maximum value and harmonic mean of the eigenvalue indicators of the covariance matrix of the TV station FM broadcast signal sample, and calculates the statistics used for spectrum sensing;

[0011] Step 3) Build a spectrum sensing decision threshold module to calculate the FM relay signal decision threshold.

[0012] Furthermore, the step 1) specifically includes the following steps:

[0013] Step 1.1) Establish a binary hypothesis testing model. Assume that the TV station FM broadcast receiving equipment is equipped with M antennas, and the TV FM broadcast signal received by multiple antennas is sampled N times. Then the signal received by the FM broadcast receiving equipment can be expressed as Here, n=1, 2, 3, ..., N, and the superscript T represents the transpose operation of the matrix. In order to detect the primary user signal, a binary hypothesis testing model is established: the primary user signal does not exist and the primary user signal exists.

[0014] Step 1.2) Construct the sample covariance matrix of the FM relay signal. Based on the above model, the FM relay signal of M antennas can form an M×N dimensional signal matrix X(N), and the sample covariance matrix of the signal is Express it.

[0015] Furthermore, the step 2) specifically includes the following steps:

[0016] Step 2.1) introduces the maximum eigenvalue metric of the power correlation matrix of the FM broadcast noise samples. If the noise is a real signal and the number of samples N approaches infinity, the maximum eigenvalue metric of the power correlation matrix of the noise samples can be calculated, and its limiting distribution and estimation results can be further analyzed.

[0017] Step 2.2) Calculate the harmonic mean of the FM broadcast signal power. M The minimum signal characteristic index is used as a substitute to obtain a new harmonic mean value of the FM relay signal power, which is substituted into the false detection probability calculation of the NMEHE method to obtain the final false detection probability.

[0018] Step 2.3) Calculate the test statistic. Covariance matrix The maximum eigenvalue λ max The harmonic mean of its eigenvalues The quotient of is constructed as the test statistic T MEHM , and finally calculate the detection statistic.

[0019] Furthermore, the step 3) specifically includes the following steps:

[0020] When N is large, we have Also note that λ min ≈σ 2 , substitute it into the final false alarm probability, and and Substitute the value of In the above equation, we can further find γ1 and γ MEHM , in order to obtain reliable detection results, use γ1 and γ MEHM The average value of is used as the final broadcast signal decision threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0022] Figure 1 This is the flow chart of this system;

[0023] Figure 2 This is the overall model framework diagram of this system;

[0024] Figure 3 The covariance matrix calculation module constructed for this method;

[0025] Figure 4 The eigenvalue and statistics calculation module constructed for this method;

[0026] Figure 5 The spectrum sensing decision threshold module constructed for this method;

[0027] Figure 6 is the detection performance of the method under different sample numbers;

[0028] Figure 7 The detection performance of the method under different antenna numbers. DETAILED DESCRIPTION

[0029] The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0030] The present invention provides a TV station FM broadcast system based on an improved spectrum sensing method, in order to obtain a more reliable signal theory decision threshold, thereby improving the detection performance. The flowchart and overall model framework diagram of the improved method are shown in the following figure. Figure 1 , as shown in 2, the method comprises the following steps:

[0031] Step 1) constructing a covariance matrix calculation module to calculate the sample covariance matrix of the FM broadcast signal;

[0032] Step 2) constructing an eigenvalue and statistics calculation module to calculate the maximum eigenvalue and harmonic mean of the eigenvalues ​​of the frequency sample covariance matrix of the FM broadcast signal, and taking the quotient of the two as the detection statistic;

[0033] Step 3) Build a spectrum sensing decision threshold module to calculate the decision threshold of the FM relay signal;

[0034] Step 4) Make a decision. If T NMEHM ≥γ NMEHM , then PU exists; if T NMEHM <γ NMEHM , then PU does not exist.

[0035] Step 5) Verify the detection performance of the method under different sample numbers and different antenna numbers.

[0036] Further, step 1) specifically includes the following steps:

[0037] Step 1.1) Figure 3 As shown in the figure, assuming that the TV station FM broadcast receiving equipment is equipped with M antennas and the multi-antenna received signal is sampled N times, the signal received by the cognitive user can be expressed as x(n) = [x1(n), x2(n)…x M (n)] T , where n = 1, 2, 3, ..., N, and the superscript T represents the matrix transpose operation. To detect the primary user signal, the following binary hypothesis testing model can be established.

[0038]

[0039] Wherein, H0 indicates that the primary user signal does not exist, H1 indicates that the primary user signal exists, s(n) indicates the primary user signal, and η(n) indicates M-dimensional Gaussian white noise.

[0040] Step 1.2) Based on the above model, the received signals of M antennas can form an M×N dimensional received signal matrix, as Figure 3 Covariance matrix calculation module:

[0041]

[0042] The sample covariance matrix of the received signal can be expressed as:

[0043]

[0044] in,(·) T Represents the transpose operation of a matrix.

[0045] Furthermore, step 2) specifically includes the following steps:

[0046] Step 2.1) Figure 4 To construct the eigenvalue and statistics calculation module, we first introduce the following estimation result of the maximum eigenvalue of the FM broadcast signal noise sampling covariance matrix. Assuming that the FM broadcast signal noise is a real signal, if Then the maximum eigenvalue of SCM has the following estimation results:

[0047]

[0048] Step 2.2) Perform the following equivalent transformation on the harmonic mean:

[0049]

[0050] Expand the above formula to get:

[0051]

[0052] The eigenvalues ​​λ2,λ3,...,λ M Substituting the minimum eigenvalue, we can get:

[0053]

[0054] Therefore, the false alarm probability of the proposed NMEHM method can be expressed as

[0055]

[0056] Substituting (7) into (8), the final false alarm probability can be expressed as

[0057]

[0058] Step 2.3) Using the FM broadcast signal covariance matrix The maximum eigenvalue λ max The harmonic mean of its eigenvalues Construct the detection statistic T NMEHM :

[0059]

[0060] Further, step 3) specifically includes the following steps:

[0061] like Figure 5 Construct spectrum sensing decision threshold module. When N is large, Also note that λ min ≈σ 2 , substituting both into formula (9) we can get:

[0062]

[0063] Simplifying, we can get:

[0064]

[0065] When the broadcast signal noise is a real signal and is defined as And mark:

[0066]

[0067] if but Converges to a Tracy-Widom distribution of order 1.

[0068] Therefore, combining the above we can get:

[0069]

[0070] From this we can get:

[0071]

[0072] Substituting the values ​​of α and β into the above formula, the FM broadcast signal decision threshold can be obtained as:

[0073]

[0074] In order to obtain reliable detection results, use γ1 and γ MEHM The average value of is used as the theoretical decision threshold:

[0075]

[0076] According to the above analysis, the FM broadcast signal decision threshold of the proposed method is finally obtained as

[0077]

[0078] Further, step 4) specifically includes the following steps:

[0079] According to formula (10) in step 2, the FM broadcast signal detection statistic T can be calculated NMEHM , the decision threshold γ of the designed NMEHM method can be calculated from formula (19) in step 3 NMEHM The decision rule of the proposed NMEHM method is as follows:

[0080]

[0081] It can be seen that if T NMEHM ≥γ NMEHM , then there is a primary user on this frequency band; if T NMEHM <γ NMEHM , then there is no primary user on this frequency band.

[0082] Further, step 5) specifically includes the following steps:

[0083] NMEHM and MEHM, MME, AGM, MEAM, MEGM and other classic EBD blind spectrum sensing methods, given the false alarm probability P fa = 0.1 when the detection performance is as follows Figure 6 and Figure 7 As shown in the figure, the results are based on 5000 Monte Carlo experiments.

[0084] Figure 6The success rate of FM broadcast signal capture of 6 EBD methods under different sample number conditions is shown, where the number of FM broadcast signal samples N is set to 500, 1000, 2000, 5000 respectively, and the number of antennas M = 4. It can be seen that: under different conditions of the number of sampling points of the FM broadcast signal, the perception performance of the NMEHM method for FM broadcast signals is significantly improved compared with the MEHM method, especially under low signal-to-noise ratio conditions; at the same time, it is not difficult to see that the NMEHM method has the best perception performance among the 6 EBD methods. For example, when N = 1000 and the signal-to-noise ratio is -15dB, the detection probabilities of the five methods MEHM, MEGM, MEAM, MME, and AGM are 14.04%, 11.20%, 8.54%, 5.64%, and 0% respectively, while that of the NMEHM algorithm is 49.08%. The detection probability of the NMEH M algorithm is improved by about 249% compared with the MEHM method. Figure 6 It can also be seen that the detection performance of the NMEHM method improves with the increase in the number of samples of the FM broadcast signal. For example, when the signal-to-noise ratio is -15dB and N = 500, 1000, 2000, and 5000, the detection probability of the NMEHM method is 32.96%, 49.08%, 70.84%, and 93.36%, respectively.

[0085] Figure 7 The signal recognition performance of 6 FM relay signal detection methods under different antenna numbers is demonstrated. The number of samples N=3000 during the simulation process. It can be seen that: under different transmission antenna numbers, the detection performance of the NMEHM method is significantly improved compared with the MEHM method, especially under low signal-to-noise ratio conditions. At the same time, it is not difficult to see that the spectrum perception ability of the NMEHM method for FM relay signals is better than that of the other 5 EBD methods. Figure 7 It can also be seen that the NMEHM method's detection probability increases with the number of station antennas. For example, when the signal-to-noise ratio is -20dB and M = 4, the NMEHM method's detection probability is 20.70%. When M = 12, the detection probability increases to 93.20%, an improvement of nearly 350%. This shows that increasing the number of station antennas can further improve NMEHM's spectrum perception of FM broadcast signals, providing a solution for effectively detecting FM broadcast signals under low signal-to-noise ratio conditions.

[0086] Finally, it should be noted that the above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although we have described specific embodiments of the present invention in detail, those skilled in the art will recognize that various adjustments and improvements can be made to the form and details without changing the core concept of the present invention. In short, such adjustments and improvements will not exceed the scope of protection defined by the claims of the present invention.

Claims

1. A TV station FM broadcasting system based on an improved spectrum sensing method, characterized in that: The following steps are involved: Step 1: Construct a covariance matrix calculation module and calculate the sample covariance matrix of the FM relay signal: Assume that the cognitive user is equipped with M antennas and samples the FM relay signal received by these multiple antennas N times. Then, the FM relay signal received by the cognitive user can be expressed as x(n) = [x1(n), x2(n)…x n (n)] T denoted by , where n = 1, 2, 3, ..., N. To detect the presence of the primary user's FM relay signal, this method constructs a binary hypothesis testing model encompassing both the absence and presence of the primary user's FM relay signal. Next, the sample covariance matrix of the FM relay received signal is constructed. Based on this model, the FM relay received signals from M antennas can be formed into an M×N dimensional received FM relay signal matrix, which is then represented by the sample covariance matrix of the received signal. Step 2: Construct an eigenvalue and statistic calculation module to calculate the maximum eigenvalue and harmonic mean of the eigenvalues ​​of the FM relay signal sample covariance matrix, and take the quotient of the two as the detection statistic: introduce the maximum eigenvalue of the signal noise sampling covariance matrix. If the noise is a real radio signal and the number of samples N tends to infinity, the maximum eigenvalue of the SCM can be calculated, and the limiting distribution of the SCM and the estimated result of the maximum eigenvalue can be further analyzed. The eigenvalue in the harmonic mean is replaced by the minimum eigenvalue to obtain a new harmonic mean. Substituting the new harmonic mean into the false alarm probability of the NMEHM method can obtain the final false alarm probability. The quotient of the maximum eigenvalue of the covariance matrix and the harmonic mean of its eigenvalues ​​is constructed as the detection statistic, and finally the FM relay signal detection statistic is calculated. Step 3: Build the spectrum sensing decision threshold module and calculate the signal decision threshold: when N is large, Also note that λ min ≈σ 2 , substitute it into the final false alarm probability, substitute the values ​​of α and β into In the above equation, we can further find γ1 and γ MEHM , in order to obtain reliable detection results, use γ1 and γ MEHM The average value of is used as the final signal decision threshold. Step 4: Make a decision: According to step 2, the detection statistic T can be calculated NMEHM , the decision threshold γ of the designed NMEHM method can be calculated from step 3 NMEHM , if T NMEHM ≥γ NMEHM , then there is a primary user on this frequency band; if T NMEHM <γ NMEHM , then there is no primary user on this frequency band. Step 5 verifies the detection performance of the methods under different sample numbers and antenna numbers. The detection performance of NMEHM and classic EBD blind spectrum sensing methods, such as MEHM, MME, AGM, MEAM, and MEGM, for a given false alarm probability is determined based on 5000 Monte Carlo experiments. The detection probabilities of the six EBD methods under different sample numbers are calculated with the sample number N set to 500, 1000, 2000, and 5000, respectively, and the number of antennas M = 4. It can be seen that the NMEHM method significantly improves detection performance over the MEHM method under different sample numbers, especially under low signal-to-noise ratio conditions. It is also clear that the NMEHM method performs best among the six EBD methods in sensing FM relay signals. The detection performance of the six EBD methods under different antenna numbers is shown with the sample number N = 3000. It can be seen that: under the conditions of different antenna numbers, compared with the MEHM method, the NMEHM method has significantly improved its perception ability of FM broadcast signals, especially under low signal-to-noise ratio conditions; at the same time, it is not difficult to see that the detection performance of the NMEHM method is better than that of the other five EBD methods.

2. The TV station FM broadcasting system based on the improved spectrum sensing method according to claim 1, characterized in that: The covariance matrix calculation module constructed in step 1, when the FM relay cognitive user is configured with M antennas and the FM relay signals received by these antennas are sampled N times. The received FM relay signals can be recorded and represented. In order to verify whether the primary user signal is present, this method constructs a binary hypothesis testing framework, which covers the two possibilities of the presence and absence of the primary user FM relay signal. Subsequently, a sample covariance matrix is ​​constructed based on these received FM relay signals. Based on this model, the FM relay signals received by M antennas can be organized into a received FM relay signal matrix with M rows and N columns, and then the sample covariance matrix of this received FM relay signal can be derived.

3. The TV station FM broadcasting system based on the improved spectrum sensing method according to claim 1, characterized in that: The eigenvalue and detection statistic module constructed in step 2, when considering the maximum eigenvalue of the noise sampling covariance matrix (SCM), if the noise is a real FM relay signal and the number of FM relay signal samples N tends to infinity, we can calculate the maximum eigenvalue of the SCM. Furthermore, this method analyzes the limit distribution of the SCM and estimates its maximum eigenvalue. By replacing the eigenvalue in the harmonic mean with the minimum eigenvalue, a new harmonic mean can be obtained. Substituting this new harmonic mean into the false alarm probability calculation of the NMEHM method, the final false alarm probability can be obtained. Then, a detection perception statistic can be constructed, which is composed of the quotient of the maximum eigenvalue of the covariance matrix and the harmonic mean of its eigenvalues. Finally, the value of this detection perception statistic can be calculated.

4. The TV station FM broadcasting system based on the improved spectrum sensing method according to claim 1, characterized in that: The spectrum sensing decision threshold module constructed in step 3 can obtain At the same time, and Substitute into the calculation formula of the final false alarm probability. Further, we can get γ1 and γ MEHM In order to ensure the reliability of the detection perception results, the average value of these two parameters is selected as the final perception decision threshold.

Images