CN101909038A - A blind signal bandwidth estimation method and device - Google Patents

Abstract

The invention discloses a blind signal bandwidth estimation method and device, comprising: transforming the received signal according to the preset discrete Fourier transform DFT points; obtaining the median bandwidth corresponding to the median bandwidth within the signal bandwidth range required by the system DFT points, and calculate the initial smoothing window length according to the median bandwidth DFT points; use the initial smoothing window length to smooth the DFT transformation results, and obtain the estimated value of the initial bandwidth DFT points; compare the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points Comparing, selecting the window length factor according to the comparison result; obtaining the variable smoothing window length processed by the window length factor; processing the DFT transformation result according to the variable smoothing window length; obtaining the signal bandwidth estimation value. The blind signal bandwidth estimation method disclosed in the embodiment of the present invention further smoothes the input signal by adding a variable window length adjusted according to the initial estimated bandwidth of the input signal, so that the estimation result is more accurate and the accuracy is improved.

Description

A blind signal bandwidth estimation method and device

technical field

The present invention relates to the field of cognitive radio technology, in particular to a blind signal bandwidth estimation method and device.

Background technique

In a cognitive radio system, the frequency range to be processed by the system is usually given when designing the system, that is, the bandwidth of the system receiver. Usually, there may be multiple signals entering the system receiver at the same time. The modulation methods of these signals are different, and the bandwidths of the signals are also different. The system usually only knows that the signal has been received, but does not know the bandwidth size, modulation type, carrier frequency and other information of the received signal. In order to better perform subsequent processing such as parameter estimation, modulation identification, and demodulation on the received signals, it is usually necessary to separate these simultaneously received signals and then estimate their respective bandwidths. When the prior art estimates the signal bandwidth, the discrete Fourier transform (DFT) is usually used to perform discrete Fourier transform (DFT) on the separated signal, and the signal is smoothed in the frequency domain with a preset smoothing window length, and then a certain threshold is set to intercept the obtained A method for estimating the bandwidth of a signal.

However, when smoothing with the preset smoothing window length, if the bandwidth of the processed signal is small, the bandwidth of the processed signal will be excessively widened, and if the bandwidth of the signal is large, the processed signal will have more Therefore, there is a large error in the estimated bandwidth, and the accuracy is low.

Contents of the invention

In view of this, the present invention provides a blind signal bandwidth estimation method and device to solve the problem of large error and low accuracy in the bandwidth estimated by using the prior art. The specific plan is as follows:

A blind signal bandwidth estimation method, comprising:

Transform the received signal according to the preset discrete Fourier transform DFT points;

Obtain the median bandwidth DFT points corresponding to the bandwidth median within the signal bandwidth range required by the system, and calculate the initial smoothing window length according to the median bandwidth DFT points;

smoothing the DFT transformation result by using the initial smoothing window length to obtain an estimated value of initial bandwidth DFT points;

Comparing the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points, selecting a window length factor according to the comparison result, the value range of the window length factor is the reciprocal greater than the estimated value of the initial bandwidth DFT points, less than 1;

Acquiring the variable smoothing window length corresponding to the estimated value of the initial bandwidth DFT points after the processing of the window length factor;

processing the DFT transformation result according to the variable smoothing window length;

An estimated signal bandwidth value corresponding to the processed result of the variable smoothing window length is obtained.

Preferably, the initial smoothing window length is calculated according to the median bandwidth DFT points:

Determine one-sixth of the median bandwidth DFT points as the initial smoothing window length.

Preferably, the process of smoothing the DFT transformation result by using the initial smoothing window length to obtain the estimated value of the initial bandwidth DFT points includes:

Normalizing the smoothed results;

Select the threshold to intercept the normalized data to obtain the estimated value of the initial bandwidth DFT points.

Preferably, the estimated value of the initial bandwidth DFT points is compared with the median bandwidth DFT points, and the process of selecting the window length factor according to the comparison result includes:

Multiplying the estimated value of the initial bandwidth DFT points by the reciprocal of the numerical value after subtracting the threshold from 1;

Compare whether the multiplication result is greater than the median bandwidth DFT points, if so, select a value greater than the median value in the value range of the window length factor, if not, select a value smaller than the value range of the window length factor The value of the middle value.

Preferably, the process of obtaining the estimated signal bandwidth corresponding to the processed result of the variable smoothing window includes:

Obtain the DFT points after the variable smoothing window length processing;

Calculate its corresponding signal bandwidth estimation value according to the number of DFT points.

A blind signal bandwidth estimation device, comprising:

The DFT transformation unit is used to transform the received signal according to the preset discrete Fourier transform DFT points;

The initial smoothing window length acquisition unit is used to obtain the median bandwidth DFT points corresponding to the bandwidth median within the signal bandwidth range required by the system, and calculate the initial smoothing window length according to the median bandwidth DFT points;

An initial bandwidth DFT point estimate acquisition unit, configured to use the initial smoothing window length to smooth process the DFT transformation result to obtain an initial bandwidth DFT point estimate;

The window length factor selection unit is used to compare the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points, select the window length factor according to the comparison result, and the value range of the window length factor is greater than the initial The reciprocal of the estimated value of bandwidth DFT points, less than 1;

A variable smoothing window length acquisition unit, configured to obtain a variable smoothing window length corresponding to the estimated value of the initial bandwidth DFT points after processing the window length factor;

A variable smoothing window length processing unit, configured to process the DFT transformation result according to the variable smoothing window length;

The estimated signal bandwidth acquisition unit is configured to obtain an estimated signal bandwidth corresponding to the processed result of the variable smoothing window length.

Preferably, the initial bandwidth DFT point estimation value acquisition unit includes:

A normalization unit, configured to normalize the smoothed result;

The threshold interception unit is used to select the threshold and intercept the normalized data to obtain the estimated value of the initial bandwidth DFT points.

Preferably, the window length factor selection unit includes:

A multiplying unit, configured to multiply the estimated value of the initial bandwidth DFT points with the reciprocal of the numerical value after subtracting the threshold value from 1;

The window length factor determination unit is used to compare whether the multiplication result is greater than the median bandwidth DFT point number, if so, select a value greater than the median value in the value range of the window length factor, if not, select a value smaller than the specified value The value of the middle value in the value range of the above-mentioned window length factor.

Preferably, the signal bandwidth estimation value acquisition unit includes:

A DFT point acquisition unit, configured to obtain the DFT points processed by the variable smoothing window length;

A signal bandwidth estimated value calculation unit, configured to calculate a corresponding signal bandwidth estimated value according to the number of DFT points.

It can be seen from the above technical solution that the blind signal bandwidth estimation method disclosed in the embodiment of the present invention further smoothes the input signal by increasing the variable window length adjusted according to the initial estimated bandwidth of the input signal, This makes the estimation result more accurate and improves the precision.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

FIG. 1 is a flowchart of a blind signal bandwidth estimation method disclosed in an embodiment of the present invention;

FIG. 2 is a flow chart of another blind signal bandwidth estimation method disclosed in an embodiment of the present invention;

Fig. 3 is a schematic structural diagram of an apparatus for estimating blind signal bandwidth disclosed by an embodiment of the present invention.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some, not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

In a blind signal bandwidth estimation method disclosed in the present invention, the process of further smoothing the input signal by using the variable window length adjusted according to the initial estimated bandwidth of the input signal is added, so that the estimation result is more accurate and the precision is improved. Spend. Its specific implementation is as follows:

The flow of a blind signal bandwidth estimation method disclosed in this embodiment is shown in Figure 1, including:

Step S11, transforming the received signal according to the preset discrete Fourier transform (DFT) points;

为使离散傅立叶变换频率分辨率满足系统需求，需使离散傅立叶变换频率分辨率小于等于系统估计精度需求d_i≤d_n，则N可由采样频率fs和离散傅立叶变换频率分辨率d_i来确定

当f_s选定以及系统估计精度需求d_n给定时，可由

即

计算出最小的N值

当N大于N_min时，可任意选择N的大小。The number of points for DFT transformation can be selected according to the principles described below. Usually the signal bandwidth range required by the cognitive radio system is [B _min , B _max ], the system estimation accuracy requirement is given as d _n , the sampling frequency is f _s , and the sampling frequency f _s is given by the processing bandwidth designed by the system, satisfying Nyquist sampling theorem, the number of discrete Fourier transform points is N, and the frequency resolution of discrete Fourier transform is

In order to make the discrete Fourier transform frequency resolution meet the system requirements, it is necessary to make the discrete Fourier transform frequency resolution less than or equal to the system estimation accuracy requirement d _i ≤ d _n , then N can be determined by the sampling frequency fs and the discrete Fourier transform frequency resolution d _i

When f _s is selected and the system estimation accuracy requirement d _n is given, it can be obtained by

Right now

Calculate the minimum N value

When N is greater than N _min , the size of N can be selected arbitrarily.

Step S12, obtaining the median bandwidth DFT points corresponding to the bandwidth median within the signal bandwidth range required by the system, and calculating the initial smoothing window length according to the median bandwidth DFT points;

When selecting the smoothing window length in the prior art, the principle that a small bandwidth signal is smoothed with a short window length and a large bandwidth signal is smoothed with a long window length is generally followed. Under the condition of known signal bandwidth, through a large number of simulations, it is found that when the length of the smoothing window is 1/4 to 1/8 of the number of DFT points N of the signal bandwidth, the smoothing of the signal with this window length in the frequency domain can achieve better results. smoothing effect. Since the currently processed signal is a blind signal and the signal bandwidth is unknown, the median B _mid of the system bandwidth range is selected to obtain the corresponding median bandwidth DFT points, and then the initial smoothing window length M ₁ is calculated based on the points.

Step S13, using the initial smoothing window length to smooth process the DFT transformation result to obtain an estimated value of initial bandwidth DFT points;

In this step, according to the method for calculating the estimated bandwidth in the prior art, and according to the result of DFT smoothing with the initial smoothing window length, an estimated value of DFT points of the initial bandwidth is obtained.

Step S14, comparing the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points, and selecting a window length factor according to the comparison result, and the value range of the window length factor is greater than the estimated value of the initial bandwidth DFT points The reciprocal of , less than 1;

In this step, the estimated value of the initial bandwidth DFT points is compared with the median bandwidth DFT points, and then it can be judged whether the bandwidth of the currently estimated signal is greater than the bandwidth median value within the signal bandwidth range required by the system or is smaller than the median value, according to the result. The window length factor is used to process the initial bandwidth DFT points according to the actual situation of the estimated signal.

It is set in the present invention that the variable window length _M2 is obtained by multiplying a certain window length factor and N _b1 , the theoretical minimum window length is 1, and its corresponding window length factor is 1/N _b1 ; the maximum window length is N _b1 , the corresponding window length factor is 1, and the theoretical range of the window length factor is (1/N _b1 , 1). In the above steps, when the length of the selected smoothing window is 1/4 to 1/8 of the number of DFT points of the signal bandwidth, a better smoothing effect can be obtained, so it can usually be in the range of (1/8, 1/4) Choose a window length factor.

Step S15. Obtain the variable smoothing window length corresponding to the estimated value of the initial bandwidth DFT point number processed by the window length factor;

The value of the initial bandwidth DFT point processed by the window length factor can reflect the characteristics of the bandwidth range of the currently processed signal.

Step S16, processing the DFT transformation result according to the variable smoothing window length;

Using a variable smoothing window length to process the DFT transformation result, the smoothing effect is better.

Step S17, obtaining an estimated signal bandwidth value corresponding to the processed result of the variable smoothing window length.

Signal bandwidth estimates obtained from this step are more accurate.

In the blind signal bandwidth estimation method disclosed in the present invention, the process of further smoothing the input signal with a variable window length adjusted according to the initial estimated bandwidth of the input signal is added, so that the estimation result is more accurate and the precision is improved.

The flow of another blind signal bandwidth estimation method disclosed in the present invention is shown in Figure 2, including:

Step S21, transforming the received signal according to the preset discrete Fourier transform (DFT) points;

Step S22. Obtain the median bandwidth DFT points corresponding to the bandwidth median within the signal bandwidth range required by the system, and calculate the initial smoothing window length according to the median bandwidth DFT points;

Calculate the number of points N _mid occupied by B _mid in the N-point DFT, and the calculation formula is That is, the number of points occupied by the median value of the signal bandwidth in the result of DFT transformation performed on the system processing bandwidth. Since the length of the smoothing window is 1/4 to 1/8 of the number of DFT points N of the signal bandwidth, a better smoothing effect can be obtained for signal smoothing with this window length in the frequency domain, so 1/6 is selected in this embodiment.

Step S23, smoothing the DFT transformation result by using the initial smoothing window length;

Step S24, normalizing the smoothed result;

Step S25, select the threshold to intercept the normalized data, and obtain the estimated value of the initial bandwidth DFT points;

After the signal is normalized, the theoretical range of its interception threshold is between 0 and 1, and the threshold value Thr is proportional to the interception value. Assume that the number of DFT points occupied by the real bandwidth of the signal is N _br . Through simulation, the relationship between the bandwidth intercept value N _b1 , the threshold Thr, and the number of points N _br of the real signal bandwidth can be approximately expressed as follows: N _br *(1-Thr)≈N _b1 . If the threshold is selected higher, the bandwidth value after interception is smaller and the error is larger; if the threshold is lower, the interception value is closer to the real value, but it is greatly affected by noise. Thr can be set according to actual conditions, and generally ranges from 0.4 to 0.7.

Step S26, multiplying the estimated value of the initial bandwidth DFT points by the reciprocal of the numerical value after subtracting the threshold value from 1;

The estimated value of the initial bandwidth DFT points in this step is an intercepted value N _b1 of DFT points obtained after smoothing and intercepting the fixed window length. This intercepted value cannot be directly compared with the median bandwidth DFT points corresponding to the median bandwidth. The value should reflect the signal bandwidth smoothed by the fixed initial smoothing window, obtained by N _b1 *1/(1-Thr). In step S25, the relationship between the bandwidth interception value N _b1 , the threshold Thr, and the number of real bandwidth points N _br of the signal is approximately expressed as follows: N _br *(1-Thr)≈N _b1 , in order to obtain a current estimated value of the DFT points occupied by the initial signal The approximate value of the number of DFT points N _br occupied by the signal bandwidth after window length smoothing, so the initial bandwidth estimate is multiplied by the reciprocal of the value obtained by subtracting the threshold value from 1. The corresponding formula is N _b1 *1/(1-Thr), and N _br ≈N _b1 *1/(1-Thr) is derived from the expression N _br *(1-Thr)≈N _b1 .

Step S27, comparing whether the multiplication result is greater than the number of median bandwidth DFT points, if yes, execute step S28a, if not, execute step S28b;

Step S28a, selecting a value greater than the median value within the value range of the window length factor;

Step S28b, selecting a value smaller than the median value within the value range of the window length factor;

Step S29, obtaining the variable smoothing window length corresponding to the estimated value of the initial bandwidth DFT points processed by the window length factor;

According to different situations, the selected value is multiplied by the result of step S26 to obtain a variable smoothing window length, which is variable because it has been adjusted accordingly according to the actual estimated signal bandwidth, and the window length is not fixed. In order to meet the needs of different signal bandwidths, to ensure the accuracy of the results.

Step S210, processing the DFT transformation result according to the variable smoothing window length;

The DFT value of the signal is smoothed twice by using the variable smoothing window length _M2 , and the smoothing result is normalized. Since the length of the variable smoothing window is different according to the actual signal bandwidth, the processed result will also be different according to the signal.

Step S211, obtaining the DFT points after the variable smoothing window length processing;

Select Thr as the threshold to intercept the normalized smoothing result, and obtain the number of DFT points N _b2 corresponding to the bandwidth estimation value of the signal.

Step S212. Calculate the corresponding signal bandwidth estimation value according to the number of DFT points.

According to the relationship of bandwidth value=number of DFT points occupied by the bandwidth×DFT resolution, the obtained DFT points are converted into a bandwidth estimation value, and the product of the DFT resolution is the final signal bandwidth estimation value.

The blind signal bandwidth estimation method disclosed in the present invention describes in detail the specific process of using the window length factor to process the estimated value of the initial bandwidth DFT points. According to the bandwidth characteristics of different processed signals, the result of secondary smoothing processing using the variable smoothing window length obtained from the estimated value of the initial bandwidth DFT points after processing is more accurate, the estimation accuracy is higher, and it can adapt to a variety of signal bandwidth ranges , expanding its scope of application.

The embodiment of the present invention does not limit the method of using the DFT points corresponding to the bandwidth for calculation and comparison in the intermediate processing process, and it can also be converted into a bandwidth according to the relationship of bandwidth value = DFT points occupied by the bandwidth × DFT resolution The value is processed, but because the value after the discrete Fourier transform is a discrete value, the bandwidth value needs to be converted to its corresponding DFT point number in the subsequent processing, so compared with the above-mentioned embodiment, the operation is more complicated.

The implementation process of the application disclosed in the present invention includes: first, according to the signal bandwidth range [B _min , B _max ] required by the system, take the median value of the bandwidth range and record it as B _mid , and calculate B _mid after performing N-point DFT on the received signal The number of points N _mid in the result of , the calculation formula is Set 1/6 of N _mid as the initial smoothing window length, and then perform N-point DFT transformation on the processed signal, and the transformed data X(n) performs sliding smoothing with a window length of M, X ₁ (k) is the smoothed data, the length is N-M+1, the smoothed data X ₁ (k) is divided by its maximum value max(X ₁ (k)), and the normalized data X ₂ (k) is obtained, set The threshold value is set to 0.5, and the number of X ₂ (k) values greater than 0.5 is taken as N _b1 . Divide the bandwidth range corresponding to N _b1 into two sections, the part greater than B _mid and the part smaller than B _mid ; calculate the number of DFT points N _mid of B _mid ; compare twice N _b1 with N _mid , and when it is greater than N _mid, two The secondary smoothing window length M ₂ is generated by the multiplication operation of N _b1 and the factor 1/6, and the secondary smoothing window length M ₂ is generated by the multiplication operation of N _b1 and the factor 1/8 when it is smaller than N _mid . Then set the variable smoothing window length, use the variable smoothing window length to smooth again, and finally get a more accurate bandwidth estimation value.

The present invention discloses a device for estimating blind signal bandwidth. Its structure is shown in FIG. 3, including: DFT transformation unit 31, initial smoothing window length acquisition unit 32, initial bandwidth DFT points estimation value acquisition unit 33, window length factor selection unit 34. Variable smoothing window length acquisition unit 35, variable smoothing window length processing unit 36 and signal bandwidth estimation value acquisition unit 37, wherein:

The DFT transformation unit 31 is used to transform the received signal according to the preset discrete Fourier transform DFT points; the initial smoothing window length acquisition unit 32 is used to obtain the median bandwidth DFT corresponding to the bandwidth median within the signal bandwidth range required by the system points, and calculate the initial smoothing window length according to the median bandwidth DFT points; the initial initial bandwidth DFT point estimate acquisition unit 33 is used to use the initial smoothing window length to smooth process the DFT transformation result to obtain the initial bandwidth DFT Estimated value of points; window length factor selection unit 34, used to compare the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points, select the window length factor according to the comparison result, and the value range of the window length factor is greater than the reciprocal of the estimated value of the initial bandwidth DFT points, and is less than 1; the variable smoothing window length obtaining unit 35 is used to obtain the variable smoothing window length corresponding to the initial bandwidth estimated value after the window length factor processing; The variable smoothing window processing unit 36 is used to process the DFT transformation result according to the variable smoothing window length; the signal bandwidth estimation value acquisition unit 37 is used to obtain the signal corresponding to the processed result of the variable smoothing window length Estimated bandwidth.

The initial bandwidth DFT point estimation value acquisition unit 33 includes: a normalization unit 331, which is used to normalize the smoothed result; a threshold interception unit 332, which is used to select the threshold and intercept the normalized data to obtain The initial bandwidth DFT point estimate.

The window length factor selection unit 34 includes: a multiplication unit 341, which is used to multiply the initial bandwidth DFT point number estimate with the reciprocal of the numerical value after subtracting the threshold value from 1; the window length factor determination unit 342 is used to compare the Whether the multiplication result is greater than the median bandwidth DFT point number, if so, then select a value greater than the median value in the value range of the window length factor, if not, select a value less than the median value in the value range of the window length factor value.

The specific operation process of each unit in this device is as follows:

The DFT transformation unit 31 transforms the signal received by the system according to the preset discrete Fourier transform DFT points to obtain a discrete signal, and the initial smoothing window length acquisition unit 32 acquires the median value according to the bandwidth median value within the signal bandwidth range required by the system bandwidth DFT points, and calculate the initial smoothing window length according to the median bandwidth DFT points, usually select one-eighth to one-fourth of the median bandwidth DFT points, and the initial bandwidth DFT point estimation value acquisition unit 33 uses the described The initial smoothing window length smoothes the discrete signal after the DFT transformation to obtain an estimated value of the initial bandwidth DFT points, and the window length factor selection unit 34 compares the estimated value of the initial bandwidth DFT points with the median bandwidth DFT points, and selects the window according to the comparison result The length factor, the value range of the window length factor is greater than the reciprocal of the estimated value of the initial bandwidth DFT points, less than 1, and utilizes the obtained window length factor to process the initial bandwidth DFT points, and the variable smoothing window length acquisition unit 35 according to The initial bandwidth estimation value processed by the window length factor obtains its corresponding variable smoothing window length; the variable smoothing window length processing unit 36 processes the DFT transformation result according to the variable smoothing window length, and finally, the signal bandwidth estimation value The obtaining unit 37 obtains an estimated signal bandwidth value corresponding to the processed result of the variable smoothing window length.

Each embodiment in this specification is described in a progressive manner, each embodiment focuses on the difference from other embodiments, and the same and similar parts of each embodiment can be referred to each other. As for the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and for the related information, please refer to the description of the method part.

Professionals can further realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software or a combination of the two. In order to clearly illustrate the possible For interchangeability, in the above description, the composition and steps of each example have been generally described according to their functions. Whether these functions are executed by hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be regarded as exceeding the scope of the present invention.

The steps of the methods or algorithms described in connection with the embodiments disclosed herein may be directly implemented by hardware, software modules executed by a processor, or a combination of both. Software modules can be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other Any other known storage medium.

The above description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (9)

1. a blind signal bandwidth method of estimation is characterized in that, comprising:

Received signal counted according to default discrete Fourier transform DFT carry out conversion;

The intermediate value bandwidth DFT that obtains the bandwidth intermediate value correspondence in the signal bandwidth scope of system requirements counts, and counts according to described intermediate value bandwidth DFT that to calculate initial smoothing windows long;

Utilize the described DFT transformation results of the long smoothing processing of described initial smoothing windows, obtain the initial bandwidth DFT estimated value of counting;

Count estimated value and described intermediate value bandwidth DFT of described initial bandwidth DFT counted and compare, select the long factor of window according to comparative result, the span of the long factor of described window is for the inverse of the estimated value of counting greater than described initial bandwidth DFT, less than 1;

It is long to obtain the count variable smoothing windows of estimated value correspondence of initial bandwidth DFT after the long factor of described window is handled;

According to the described DFT transformation results of described variable smoothing windows long process;

Obtain with described variable smoothing windows long process after the corresponding signal bandwidth estimated value of result.

2. method according to claim 1 is characterized in that, describedly counts according to described intermediate value bandwidth DFT that to calculate initial smoothing windows long:

Determine that the sixth that described intermediate value bandwidth DFT counts is that initial smoothing windows is long.

3. method according to claim 2 is characterized in that, the described described DFT transformation results of the long smoothing processing of described initial smoothing windows of utilizing obtains the initial bandwidth DFT estimated value process of counting and comprises:

Result after the described smoothing processing is carried out normalization;

Choose the data after the thresholding intercepting normalization, obtain the initial bandwidth DFT estimated value of counting.

4. method according to claim 3 is characterized in that, count estimated value and described intermediate value bandwidth DFT of described initial bandwidth DFT counted and compare, and selects the process of the long factor of window to comprise according to comparative result:

The described initial bandwidth DFT estimated value and 1 of counting is deducted numerical value reciprocal multiplication after the described threshold value;

Whether more described multiplied result counts greater than described intermediate value bandwidth DFT, if, then choose value greater than median in the long factor span of described window, if not, then select value less than median in the long factor span of described window.

5. according to any described method among the claim 1-4, it is characterized in that the process of the signal bandwidth estimated value that described acquisition is corresponding with the result after the described variable smoothing windows long process comprises:

The DFT that obtains after the described variable smoothing windows long process counts;

Count according to described DFT and to calculate its corresponding signal bandwidth estimated value.

6. a blind signal bandwidth estimation unit is characterized in that, comprising:

The DFT converter unit is used for received signal counted according to default discrete Fourier transform DFT and carries out conversion;

The initial long acquiring unit of smoothing windows, the intermediate value bandwidth DFT that is used to obtain the interior bandwidth intermediate value correspondence of the signal bandwidth scope of system requirements counts, and counts according to described intermediate value bandwidth DFT that to calculate initial smoothing windows long;

The initial bandwidth DFT estimated value acquiring unit of counting is used to utilize the described DFT transformation results of the long smoothing processing of described initial smoothing windows, obtains the initial bandwidth DFT estimated value of counting;

The long factor selected cell of window, be used for count estimated value and described intermediate value bandwidth DFT of described initial bandwidth DFT counted and compare, select the long factor of window according to comparative result, the span of the long factor of described window is for the inverse of the estimated value of counting greater than described initial bandwidth DFT, less than 1;

The long acquiring unit of variable smoothing windows, it is long to be used to obtain the count variable smoothing windows of estimated value correspondence of initial bandwidth DFT after the long factor of described window is handled;

Variable smoothing windows long process unit is used for according to the described DFT transformation results of described variable smoothing windows long process;

Signal bandwidth estimated value acquiring unit, be used to obtain with described variable smoothing windows long process after the corresponding signal bandwidth estimated value of result.

7. device according to claim 6 is characterized in that, the described initial bandwidth DFT estimated value acquiring unit of counting comprises:

The normalization unit is used for the result after the described smoothing processing is carried out normalization;

The thresholding interception unit is used to choose the data after the thresholding intercepting normalization, obtains the initial bandwidth DFT estimated value of counting.

8. device according to claim 7 is characterized in that, the long factor selected cell of described window comprises:

The unit that multiplies each other is used for the described initial bandwidth DFT estimated value and 1 of counting is deducted numerical value reciprocal multiplication after the described threshold value;

The long factor determining unit of window, be used for more described multiplied result and whether count, if then choose value greater than median in the long factor span of described window greater than described intermediate value bandwidth DFT, if not, then select to grow the value of median in the factor span less than described window.

9. according to any described device among the claim 6-8, it is characterized in that described signal bandwidth estimated value acquiring unit comprises:

The DFT acquiring unit of counting, the DFT that is used to obtain after the described variable smoothing windows long process counts;

Signal bandwidth estimated value computing unit is used for counting according to described DFT and calculates its corresponding signal bandwidth estimated value.

Images