阅读说明：本技术 一种用于导航自定位的数字广播电视信号认知方法 (Digital broadcast television signal cognition method for navigation self-positioning ) 是由 王峰 吉丰 于 2019-09-19 设计创作，主要内容包括：本发明公开了一种用于导航自定位的数字广播电视信号认知方法,所述方法包括如下步骤：根据特征参数设计滤波器；通过所述滤波器对阵列数据进行滤波处理,获取第一数字信号；对所述第一数字信号进行分组合并,获取第二数字信号；判断每组所述第二数字信号是否存在同频台信号,若存在则采用盲源分离算法对第二数字信号进行分离,获取分离后的电台信号,若不存在则采用能量检测方法对第二数字信号进行检测,获取分离后的电台信号。(the invention discloses a digital broadcast television signal cognition method for navigation self-positioning, which comprises the following steps: designing a filter according to the characteristic parameters; filtering the array data through the filter to obtain a first digital signal; grouping and combining the first digital signals to obtain second digital signals; and judging whether the second digital signals of each group have the same-frequency station signals, if so, separating the second digital signals by adopting a blind source separation algorithm to obtain separated station signals, and if not, detecting the second digital signals by adopting an energy detection method to obtain the separated station signals.)

1. a method for navigating self-positioned digital broadcast television signal cognition, the method comprising the steps of:

designing a filter according to the characteristic parameters;

Filtering the array data through the filter to obtain a first digital signal;

Grouping and combining the first digital signals to obtain second digital signals;

and judging whether the second digital signals of each group have the same-frequency station signals, if so, separating the second digital signals by adopting a blind source separation algorithm to obtain separated station signals, and if not, detecting the second digital signals by adopting an energy detection method to obtain the separated station signals.

2. the method of claim 1, wherein the characteristic parameters include type, number, frequency point and bandwidth of the digital television broadcasting signals.

3. the method of claim 1, wherein the filter is designed by a method comprising:

Establishing a frequency response function;

acquiring unit impulse response according to the frequency response function;

And windowing the unit impulse response to obtain the filter.

4. the method of claim 3, wherein the filter is expressed as follows:

h(n)＝h_d(n)ω(n)，

Where H (n) is the filter, ω (n) is the window function, H_d(e^jω) As a function of the frequency response, h_d(n) is the unit impulse response, n is the length of the filter, ω is the angular frequency, e is an exponential representation, and j is an imaginary representation.

5. the method of claim 1, wherein the blind source separation algorithm separation process comprises:

carrying out zero mean value and whitening pretreatment on the observation signal matrix;

Constructing a fourth-order cumulant function of the processed observation matrix;

Calculating a fourth-order cumulant matrix according to the fourth-order cumulant function;

Performing eigenvalue decomposition on the fourth-order cumulant matrix to obtain an estimation of a separation matrix;

And obtaining the source signal according to the estimation of the separation matrix.

6. The method of claim 5, wherein the fourth order cumulant function is:

In the formula, H_Y(i, j, p, q) is a fourth-order cumulant function, y_i、y_pThe ith and p-th beams of Y (k) respectively,the j and q beams of Y (k) respectively take conjugate operation, and cum (·, ·,) represents the symbol of fourth-order cumulant operation.

7. the method of claim 5, wherein the fourth order cumulant matrix is calculated as follows:

let m_pqfor the p-th row and q-th column of the matrix M, the i-th row and j-th column of the fourth-order cumulant matrix are defined as:

From the correlation properties of the cumulative quantities:

in the formula, H_Y(M) is a fourth order cumulant matrix, k_mis the fourth order cumulant of the signal source, M is any N order matrix, X is orthogonal matrix, X_mIs the m-th column of the orthogonal matrix X,diag[·]representing the diagonal matrix.

8. The method of claim 5, wherein the eigenvalue decomposition is formulated as follows:

in the formula, H_Y(M) is a fourth-order cumulant matrix, V is a diagonal matrix composed of eigenvalues,For the estimation of the separation matrix, matrixAnd momentThe relationship of the matrix X is:

in the formula, X is an orthogonal matrix, K is a permutation matrix, D is a matrix with the diagonal elements of +/-1 and other elements of 0.

9. the method of claim 5, wherein the source signal is expressed as follows:

In the formula (I), the compound is shown in the specification,For the estimation of the separation matrix, Y (k) is the observed signal matrix, [. cndot]^His a conjugate transpose operation.

10. The method of claim 1, wherein the energy detection method comprises the following steps:

carrying out Fourier transform on the second digital signal to obtain a frequency domain signal;

Performing modular squaring on the frequency domain signal to obtain a third digital signal;

and judging whether the peak value of the third digital signal is greater than a threshold, if so, the signal is a separated radio station signal.

Technical Field

The invention relates to the field of cognitive radio, in particular to a digital broadcast television signal cognitive method for navigation self-positioning.

Background

Through the development of years, digital television broadcast signals (DTMB and DVB-T, FM) have wider coverage, longer range of action, dense stations and high power, and become an ideal choice for navigation and self-positioning. However, there is a problem that the selection of the radiation source and the frequency band thereof by the system platform has a great influence on the positioning effect, mainly because the broadcast television signals have strong noise and interference when being transmitted to the receiving end, thereby causing the signal quality to be reduced. Signals doped with various interferences and noises cannot be used for accurately evaluating the signals, so that the measurement of various parameters is influenced, and the navigation self-positioning precision of a system is influenced finally.

disclosure of Invention

In view of the defects of the prior art, the present invention aims to provide a method for recognizing a digital broadcast television signal for navigation and self-positioning, so as to solve the problem that the signal in the prior art has strong noise and interference.

In order to solve the technical problems, the technical scheme adopted by the invention is as follows:

A digital broadcast television signal recognition method for navigation self-localization, the method comprising the steps of:

Designing a filter according to the characteristic parameters;

Filtering the array data through the filter to obtain a first digital signal;

Grouping and combining the first digital signals to obtain second digital signals;

and judging whether the second digital signals of each group have the same-frequency station signals, if so, separating the second digital signals by adopting a blind source separation algorithm to obtain separated station signals, and if not, detecting the second digital signals by adopting an energy detection method to obtain the separated station signals.

further, the characteristic parameters include the type, the number, the frequency points and the bandwidth of the digital television broadcasting signals.

further, the design method of the filter comprises the following steps:

establishing a frequency response function;

Acquiring unit impulse response according to the frequency response function;

And windowing the unit impulse response to obtain the filter.

Further, the expression of the filter is as follows:

h(n)＝h_d(n)ω(n)，

where H (n) is the filter, ω (n) is the window function, H_d(e^jω) As a function of the frequency response, h_d(n) is the unit impulse response, n is the length of the filter, ω is the angular frequency, e is an exponential representation, and j is an imaginary representation.

Further, the separation process of the blind source separation algorithm includes:

carrying out zero mean value and whitening pretreatment on the observation signal matrix;

Constructing a fourth-order cumulant function of the processed observation matrix;

calculating a fourth-order cumulant matrix according to the fourth-order cumulant function;

Performing eigenvalue decomposition on the fourth-order cumulant matrix to obtain an estimation of a separation matrix;

And obtaining the source signal according to the estimation of the separation matrix.

Further, the fourth order cumulant function is:

in the formula, H_Y(i, j, p, q) is a fourth-order cumulant function, y_i、y_pI, p beams of the Y (k) observation signal vector matrix respectively,the j and q beams of Y (k) respectively take conjugate operation, and cum (·, ·,) represents the symbol of fourth-order cumulant operation.

Further, the fourth-order cumulant matrix is calculated as follows:

Let m_pqFor the p-th row and q-th column of the matrix M, the i-th row and j-th column of the fourth-order cumulant matrix are defined as:

From the correlation properties of the cumulative quantities:

In the formula, H_Y(M) is a fourth order cumulant matrix, k_mis the fourth order cumulant of the signal source, M is any N order matrix, X is orthogonal matrix, X_mis the m-th column of the orthogonal matrix X,diag[·]representing the diagonal matrix.

Further, the formula of the eigenvalue decomposition is as follows:

In the formula, H_Y(M) is a fourth-order cumulant matrix, V is a diagonal matrix composed of eigenvalues,for the estimation of the separation matrix, matrixThe relationship to matrix X is:

In the formula, X is an orthogonal matrix, K is a permutation matrix, D is a matrix with the diagonal elements of +/-1 and other elements of 0.

further, the expression formula of the source signal is as follows:

in the formula (I), the compound is shown in the specification,For the estimation of the separation matrix, Y (k) is the observed signal matrix, [. cndot]^HIs a conjugate transpose operation.

further, the detection process of the energy detection method is as follows:

carrying out Fourier transform on the second digital signal to obtain a frequency domain signal;

performing modular squaring on the frequency domain signal to obtain a third digital signal;

And judging whether the peak value of the third digital signal is greater than a threshold, if so, the signal is a separated radio station signal.

Compared with the prior art, the invention has the following beneficial effects:

The invention designs a plurality of groups of filters by adopting prior information such as digital television signal characteristic parameters and the like, extracts each signal in the received signals and then carries out energy detection to obtain excellent radio signals

Drawings

FIG. 1 is a prior knowledge base based digital television signal recognition method;

FIG. 2 is a simplified schematic diagram of a FIR filter bank;

FIG. 3 is a real part and a spectrogram of a received signal;

FIG. 4 is a DTMB (506MHz) filter amplitude frequency response;

FIG. 5 shows the DTMB (506MHz) filtered signal;

FIG. 6 is a frequency modulated stereo real part and spectrum diagram;

FIG. 7 is a real part and a frequency spectrum of a blind source separation signal 1;

Fig. 8 is a real part and frequency spectrum of the blind source separation signal 2.

Detailed Description

The invention is further described below with reference to the accompanying drawings. The following examples are only for illustrating the technical solutions of the present invention more clearly, and the protection scope of the present invention is not limited thereby.

The digital video broadcasting-terrestrial (DVB-T) is a set of transmission system developed by the european digital video broadcasting organization, which was established in 1996 and is the current international terrestrial broadcasting standard for digital television with the widest coverage rate, the signal bandwidth is 8MHz, and the frequency coverage range is from 177.5MHz to 858 MHz. The standard employs a Coded Orthogonal Frequency Division Multiplexing (COFDM) modulation technique that combines inner code coding with orthogonal frequency division multiplexing modulation. The system transmission frame adopts a layered frame structure, and is respectively a superframe, a signal frame and an OFDM symbol from top to bottom, and each symbol consists of a guard interval and a symbol effective part. Terrestrial digital multimedia broadcasting (DTMB) is a China digital video broadcasting standard and is determined as the only mandatory standard of China television signals in 2007, the frequency coverage range is 470 MHz-860 MHz, and the signal bandwidth is 8 MHz. The standard adopts an orthogonal frequency division multiplexing modulation technology of time domain synchronization, and has a layered multi-frame structure which is absolutely synchronous with natural time, a system data frame is respectively a daily frame, a sub-frame, a super-frame and a signal frame from top to bottom, and a basic data unit signal frame consists of a PN frame head and a frame body data. Frequency modulation-amplitude modulation (pilot frequency system) is adopted in China for frequency modulation-amplitude modulation (FM), the frequency range is 87-108 MHz, and the channel width of each radio station is 200 kHz. The signal is to encode and modulate the left and right sound channels separately, the baseband modulation signal is mainly composed of a sum signal, a difference signal, a pilot signal, a subcarrier modulation signal, etc.

1. Digital broadcast television signal cognition model

The invention provides a digital television broadcast signal cognitive method for navigation self-positioning, which combines a priori knowledge base with a cognitive radio method and adds a blind source separation algorithm to realize separation of digital television broadcast signals. The method comprises the steps of setting 6 groups of digital television broadcasting signals comprising FM signals and DVB-T, DTMB signals, wherein the frequency points are different, the sampling rate is set to be 1.6GHz, a receiving antenna is an even quinary circular array, the five-ary circular array is used for receiving the television broadcasting signals to generate five groups of channel array data, then a group of filters are designed for each group of channels according to priori knowledge such as signal characteristic parameters and the like to filter and extract the television broadcasting signals, then corresponding signals extracted from each group are combined respectively, whether common-frequency signals exist in each group of signals is judged, blind source separation is carried out on the channel signals if the common-frequency signals exist, and energy detection is carried out on the signals if the common-frequency signals do not exist. The cognitive model is shown in fig. 1.

Fig. 1 shows a prior knowledge base-based digital television signal cognition method, which includes the following steps:

(1) Setting a filter bank according to the type, the number, the frequency points, the bandwidth and other characteristic parameters of the digital television broadcasting signals existing in a certain area in a priori knowledge base;

(2) the method comprises the steps that a set filter bank is used for filtering array data to obtain a first digital signal, each filter filters out a digital television broadcast signal, and each filter is required to filter out a complete signal and not contain signals of other frequency points;

(3) Grouping and combining the extracted signals to obtain a second digital signal, and judging whether the signals extracted from each group have the same frequency station signals according to the prior information of the television signals in a prior knowledge base;

(4) If the grouped and combined signals have the same-frequency station signals, separating the signals by adopting a blind source separation algorithm to obtain separated station signals;

(5) if the grouped and combined signals do not have the same-frequency station signals, detecting the signals by adopting an energy detection method, firstly carrying out Fourier transform on the filtered signals to obtain frequency domain signals, and then carrying out modular squaring on the frequency domain signals to obtain third digital signals; setting a reasonable threshold, selecting the peak value of the frequency domain signal subjected to modular squaring to be compared with the threshold, judging that a good signal exists in the filter if the peak value is greater than the threshold so as to obtain a separated radio station signal, and judging that only noise exists in the filter if the peak value is lower than the threshold.

the steps of the prior knowledge base-based digital television signal cognition method are completely finished, and the prior knowledge base, the filter bank design and the blind source separation algorithm are elaborated in detail below. 2.1 a priori knowledge base

the parameters of the filter bank are set according to a priori knowledge base. The priori knowledge base stores information of all digital television broadcast signals in a certain geographic range, and can update data at any time according to a certain mechanism, and the specific structure of the priori knowledge base is shown in table 1.

TABLE 1 priori knowledge base structure of digital television broadcast signals

The specific structure of the prior knowledge base of the digital television broadcast signals is shown in table 1, and the types of the digital television broadcast signals existing in each region, the number of the signals corresponding to each type of signals, the characteristic parameters corresponding to each type of signals, and the radio coordinates are listed in the prior knowledge base one by one. When the aircraft flies to a certain region, the digital television broadcast signal characteristic parameters of the region can timely interact with system equipment through the prior knowledge base, so that the access equipment can reasonably utilize frequency spectrum resources conveniently, and interference among signals is avoided.

Because the prior knowledge base is responsible for interacting with the system equipment and providing various related data information for the system equipment, the prior knowledge base has the following basic functions:

(1) according to the request information and the geographical position information uploaded by the system equipment, the priori knowledge base provides various data information of all digital television broadcasting signals existing in the region to the system;

(2) the prior knowledge base simultaneously calculates and provides a channel list which can be used by system equipment according to the data information of the television signals, and reasonably divides and utilizes frequency spectrum resources.

2.2 FIR Filter Bank design

Finite Impulse Response (FIR) filters have strict linear phase characteristics and good stability and are therefore often designed as multi-passband filter banks to achieve signal subband decomposition. The invention adopts the FIR filter group to realize the filter decomposition of the digital television broadcast receiving signal, and simultaneously sets the parameters of the FIR filter by combining the priori knowledge of the digital television broadcast signal in the priori knowledge base.

FIR filters generally have three design methods, which are: window function design method, frequency sampling method and optimal approximation algorithm. The design of the FIR filter is realized by combining a window function design method with prior information, and the specific realization steps are as follows:

(1) Constructing a frequency response function H expected to be approximated according to prior information_d(e^jω)；

(2) Calculate h_d(n) value, setting the frequency response function of the filter to be solved as H_d(e^jω) Then the unit impulse response is:

In the formula, H_d(e^jω) As a function of the frequency response, h_d(n) is the unit impulse response, n is the length of the filter, ω is the angular frequency, e is an exponential representation, and j is an imaginary representation.

(3) selecting a form of a window function according to various index requirements on a transition band and a stop band, and estimating a window length N;

(4) and windowing the filter to obtain the design result of the filter:

h(n)＝h_d(n)ω(n)

In the formula, h_d(n) is the unit impulse response and ω (n) is the window function.

Specific parameters such as the analysis frequency range, the cut-off frequency, the passband length, the order and the like of the finite impulse response filter need to be set according to prior information in a prior knowledge base, a system interacts with the prior knowledge base to obtain the type of a digital television broadcast signal existing in a certain area and information such as the number, the bandwidth, the frequency point and the like corresponding to the signal, so as to carry out design, and the following two points need to be noticed during design:

(1) the filter bank design requires knowledge of the types of digital television broadcast signals present in the region and the bandwidth associated with each signal. The bandwidth of digital television signals is large, 8MHz, but the bandwidth of FM broadcast signals is small, only 200kHz, the same parameters cannot be used when designing filters for the signals with different bandwidths, and the passband length of the filter should be dynamically set according to the bandwidth of each signal.

(2) meanwhile, when the filter needs to be designed, the frequency point positions of various signals also need to be known from the priori knowledge base, if the frequency point information of each signal is not noticed, two signals with close frequency points can be contained in one filter when the filter is designed, or one filter cannot filter out a complete signal and a part of the signal is leaked, so that serious interference is generated and the detection of subsequent steps is influenced.

the FIR filter bank is shown in fig. 2, and according to the difference of the regions, the filter bank knows the type, number, bandwidth, frequency point and other prior information of each regional signal from the prior knowledge base, so that when designing the filter, each filter can consider the bandwidth of the signal, reasonably design the length of each filter, completely contain each signal in the filter, consider the frequency point information of the signal, and timely change the passband range of the filter. If the above two points are not considered, the filter is designed to cause serious waste of frequency spectrum, and serious interference of two signals can be generated.

Suppose that there are 2 signals in a certain area, which are frequency modulation broadcast signal and DTMB signal, respectively, the bandwidth of the frequency modulation broadcast signal is 200kHz, the frequency point is set to 95.2MHz, the bandwidth of the DTMB signal is 8MHz, the frequency point is set to 506MHz, and the sampling rate is set to 1.6 GHz. The system calls the information of the signals in the area from the prior knowledge base and then designs the filter parameters according to the prior information. Setting the attenuation of the stopband to 60dB, selecting a Kessel window by a window function, setting the frequency spectrum ranges of the frequency modulation broadcast signals and the television signals, and calculating the related parameters of the two filters.

Setting the analysis frequency range of the DTMB signal to be 496 MHZ-516 MHZ, and then setting the filter parameters for filtering the DTMB signal as follows:

Ideal low-pass filter cutoff frequency:

f_c＝(516-496)/2＝10MHz

the center frequency is:

f_k＝496+10＝506MHz

The value of the parameter β is:

β＝0.1102(A_s-8.7)＝5.6533

The transition bandwidth is:

ΔF＝Δf×T_s＜2f_c×T_s＝20e⁶/1.6e⁹＝1/80

The order of the filter is:

M＝(A_s-7.95)/(14.36ΔF)+1＝290

the impulse response of the band-pass filter is then

h(n)＝ω(n_d)h_d(n-n_d)

And finally, performing discrete Fourier transform on the impulse response to obtain the frequency characteristic of the DTMB signal band-pass filter.

2.3 Blind Source separation Algorithm

blind source separation refers to a process of separating and recovering relatively independent source signals from a received mixed signal, and has good separation capability on same-frequency signals. The system of the invention mutually learns the channel with the same frequency signal by the prior knowledge base and separates and recovers the signal by adopting a blind source separation algorithm.

the system antenna array is assumed to be composed of M array elements, N source signals are provided, and the incidence directions of the signals are respectively set to be theta₁,θ₂,…….,θ_Noutput signal matrix Y (k) of array signals ═ Y₁(k),y₂(k),.....,y_M(k)]^Tthe source signal is S (k) ═ S₁(k),s₂(k),.......s_N(k)]^Teach receiving channel has gaussian noise, which can be expressed as N (k) ═ N₁(k),n₂(k),.....n_M(k)]^T. The output observation data of the array model at the time k is:

Y(k)＝AS(k)+N(k)

the instantaneous mixture model of blind source separation can also be expressed as

Y(k)＝AS(k)+N(k)

In the formula, Y (k) is an observation signal vector matrix, A is a mixed matrix, S (k) is an information source signal, and N (k) is a noise vector matrix.

The content of blind source separation is that in case the mixing matrix a and the source signal S (k) are unknown, the separation matrix Q is determined only from the observation signal matrix Y (k), so that the transformed output

S(k)＝QY(k)

An estimate of the signal from each station is obtained.

The method adopts a feature matrix approximate joint diagonalization (JADE) blind source separation algorithm, the algorithm firstly utilizes an observation matrix to construct a group of fourth-order cumulant matrixes, and then further obtains a separation matrix Q through the joint diagonalization of the cumulant matrixes, and the specific implementation steps are as follows:

(1) Carrying out zero-mean and whitening preprocessing on the observation signal matrix Y (k) so as to eliminate the correlation among the wave beams of each channel;

(2) constructing a fourth-order cumulant function of the processed observation matrix Y (k):

In the formula, y_i、y_pThe ith and p-th beams of Y (k) respectively,the j and q beams of Y (k) respectively take conjugate operation, and cum (·, ·,) represents the symbol of fourth-order cumulant operation.

(3) calculating a fourth-order cumulant matrix according to the fourth-order cumulant function

for the N-order matrix M, an cumulant matrix H is defined_Y(M), provided that M_pqthe p-th row and q-th column elements of M, the i-th row and j-th column elements of the cumulant matrix are defined as

From the correlation properties of the accumulated quantities

In the formula, H_Y(M) is a fourth order cumulant matrix, k_mIs the fourth order cumulant of the signal source, X is the orthogonal matrix, X_mIs the mth column of the orthogonal matrix X, M is an N-order arbitrary matrix,diag[·]Representing the diagonal matrix.

(4) And (3) carrying out eigenvalue decomposition on the cumulant matrix to obtain the estimation of a separation matrix:

In the formula, H_Y(M) is a fourth-order cumulant matrix, V is a diagonal matrix composed of eigenvalues,for the estimation of the separation matrix, matrixIn relation to the matrix X of

in the formula, X is an orthogonal matrix, K is a permutation matrix, D is a matrix with the diagonal elements of +/-1 and other elements of 0.

(5) Estimating the source signal:

wherein S (k) is a source signal,for the estimation of the separation matrix Q, Y (k) is the observation signal matrix, [. cndot]^His a conjugate transpose operation.

And determining the channel of the same frequency station signal according to the prior knowledge in the prior knowledge base, and then completing the separation and recovery of the same frequency station signal by using the blind source separation algorithm.

the simulation setting is that six groups of digital television broadcast signals exist in a certain area, wherein the frequency points of three groups of frequency modulation stereo broadcast signals are 95.2MHz, 95.8MHz and 101MHz respectively; two groups are digital terrestrial multimedia broadcasting signals (DTMB), the frequency points are 506MHz and 634 Mhz; one group is digital video terrestrial broadcast signals (DVB-T), the frequency point is 671MHz, the sampling rate is set to 1.6GH, and the receiving array is a uniform quinary circular array. One of the antennas is selected for observation in the simulation, and fig. 3 is a real part diagram and a frequency spectrum diagram of a received signal of the 1 st antenna, and the frequency spectrum of the received signal is divided at the same time.

Fig. 3 shows the real part and the frequency spectrum of the received signal of the first antenna in the array, and it can be seen from the frequency spectrum that six groups of digital television broadcast signals are all at the set frequency point positions. Meanwhile, the system divides the frequency spectrum of the signal according to the prior information in the prior knowledge base to obtain the passband range and the length of the filter to be designed.

Meanwhile, it can be seen from fig. 3 that the passband lengths of the fm stereo broadcast signal and the digital television signal are not the same because the bandwidths of the fm broadcast signal and the digital television signal are different, and the different bandwidths correspond to different frequency passband lengths. The filter bank divides the frequency spectrum according to the signal characteristic priori knowledge in the priori knowledge base and reasonably sets the parameters of the filters, so that each filter contains a complete signal as much as possible and different frequency point signals can be distributed in different filters. The filter bank extracts each signal and then carries out energy detection to obtain a station signal, thereby determining a good radiation source signal for subsequent processing.

The parameters associated with each group of filters are determined according to the spectral length of the signal division, and one group is selected for observation, fig. 4 shows the amplitude and frequency response of the filter of 506MHz of the DTMB signal, and fig. 5 shows the signal after the received signal is filtered by the filter.

Fig. 4 and 5 show the filter amplitude frequency response of the DTMB signal at 506MHz and the filtered signal, respectively, and it can be seen from fig. 5 that the received signal is filtered to obtain a single DTMB signal. For other signals, corresponding filters can be designed and extracted according to the divided frequency band range, so that the condition that the frequency spectrum range is divided according to the priori knowledge signal information so as to determine the parameters of each group of filters can be verified, and each single television broadcast signal can be effectively extracted.

it can be seen from the spectrum diagram of fig. 6 that there are two signals in one channel because the frequency points of the two fm broadcast signals are very close to each other, and are close to the same frequency station signal. The two signals are separated by a blind source separation algorithm. The real part and the frequency spectrum of the signal obtained after the signal in the channel is subjected to blind source separation are shown in fig. 7 and 8.

Fig. 7 and 8 show signals of two frequency-modulated broadcast signals with close frequency points separated by a blind source separation algorithm, and it can be seen from a spectrogram that both signals are effectively separated, the right-side signal in fig. 7 drops by nearly 16dB, and the left-side signal in fig. 8 drops by nearly 14dB, so that the effectiveness of blind source separation is verified, and the signal quality is effectively enhanced.

the invention combines a priori knowledge base and cognitive radio, designs a plurality of groups of filters by adopting the prior information such as the characteristic parameters of digital television signals in the prior knowledge base, extracts each signal in the received signals and then carries out energy detection so as to obtain excellent radio station signals. The prior knowledge base stores relevant data information of digital television signals in all regions, interacts with the access equipment and provides the digital television signal information of the region according to the geographical position of the equipment and the request information. The cognitive radio can calculate the idle frequency spectrum better, realizes the optimization to cognitive environment according to the position information of main user and perception user, from the information of the wireless environment of its work, selects suitable frequency spectrum and working parameter to confirm good radiation source signal, promote system's precision.

The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It should be understood by those skilled in the art that the above embodiments do not limit the present invention in any way, and all technical solutions obtained by using equivalent alternatives or equivalent variations fall within the scope of the present invention.