Noise reduction method based on selection of matched filter coefficients, electronic device and storage medium

Abstract

Embodiments of the present application relate to the field of signal processing, and disclose a noise reduction method based on selected matched filtering coefficients, an electronic device and a storage medium, wherein the method comprises: classifying channels according to time domain characteristics and / or frequency domain characteristics of the channels to obtain a plurality of channel categories; providing a plurality of sets of matched filtering coefficients; wherein the plurality of sets of matched filtering coefficients correspond to the plurality of channel categories one by one; matching the time domain characteristics and / or the frequency domain characteristics of the channels of each channel category with a set of matched filtering coefficients corresponding to the channel category respectively to obtain a set of matching values corresponding to each channel category; determining a matching value in the set of matching values corresponding to each channel category, and taking the matched filtering coefficient corresponding to the matching value as a target matched filtering coefficient corresponding to the channel category; and smoothing and reducing noise of the channels of each channel category by using the target matched filtering coefficient corresponding to each channel category. The present scheme is simple in calculation and can smooth and reduce noise of the channels.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of signal processing, in particular to a denoising method based on selection of matched filter coefficients, an electronic device and a storage medium. BACKGROUND

[0002] A channel refers to a signal path based on a transmission medium. In all communication systems, a signal will pass through a channel, and when the signal passes through the channel, the signal will be distorted or various noises will be added to the signal. If you want to correctly decode the received signal without too many errors, you need to remove the distortion and noise imposed by the channel from the received signal.

[0003] Due to the complexity of the channel in the wireless communication environment, the complexity is reflected in the difficulty of determining the smooth matched filter coefficient caused by the multipath complexity of the channel. In the prior art, there are two ways to select the matched filter coefficient, one of which needs to rely on historical measurement values and has poor window selection accuracy, or needs to perform FFT and iFFT transformation to estimate the power delay profile (PDP), which has a large amount of calculation. SUMMARY

[0004] The purpose of the embodiments of the present application is to provide a denoising method based on selection of matched filter coefficients, an electronic device and a storage medium, which can quickly select matched filter coefficients without relying on historical measurement information and PDP estimation, thereby realizing smooth denoising of the channel and simple calculation.

[0005] To solve the above technical problems, the embodiments of the present application provide a denoising method based on selection of matched filter coefficients, comprising:

[0006] classifying the channels according to the time domain characteristics and / or frequency domain characteristics of the channels to obtain a plurality of channel categories;

[0007] providing a plurality of sets of matched filter coefficients;

[0008] The plurality of sets of matched filter coefficients correspond one-to-one to the plurality of channel categories.

[0009] matching the time domain characteristics and / or frequency domain characteristics of the channels of each channel category with a set of matched filter coefficients corresponding to the channel category respectively to obtain a set of matched values corresponding to each channel category;

[0010] determining a matched value in the set of matched values corresponding to each channel category, and taking the matched filter coefficient corresponding to the matched value as the target matched filter coefficient corresponding to the channel category;

[0011] using the target matched filter coefficient corresponding to each channel category to perform smooth denoising on the channel of the channel category.

[0012] Embodiments of the present application also provide an electronic device, comprising:

[0013] at least one processor; and

[0014] a memory connected to the at least one processor in communication; wherein

[0015] the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the denoising method based on the selected matched filter coefficient as described above.

[0016] Embodiments of the present application also provide a computer-readable storage medium storing a computer program, and the computer program is executed by a processor to implement the denoising method based on the selected matched filter coefficient as described above.

[0017] Compared with the prior art, the embodiments of the present application classify channels according to time domain characteristics and / or frequency domain characteristics of the channels, provide a plurality of sets of matched filter coefficients for different types of channels according to the classification results of the channels, and each set of matched filter coefficients corresponds to one type of channel. The time domain characteristics and / or frequency domain characteristics of the channels of each type of channel are matched with the matched filter coefficients corresponding to the type of channel, to obtain a set of matched values corresponding to each type of channel. One matched value is determined in the set of matched values corresponding to each type of channel, and the matched filter coefficient corresponding to the matched value is used as the target matched filter coefficient corresponding to the type of channel. The target matched filter coefficient corresponding to the type of channel is used to perform smooth denoising on the channels of the type of channel. According to the embodiments of the present application, by classifying channels, a set of matched filter coefficients corresponding to each type of channel is determined according to the classification characteristics, and one matched filter coefficient is selected by matching the time domain characteristics and / or frequency domain characteristics of each type of channel with the corresponding set of matched filter coefficients. The matched filter coefficient is used as the target matched filter coefficient corresponding to the type of channel, and the target matched filter coefficient is used to perform smooth denoising on the channels of the type of channel. Therefore, without relying on historical measurement information and power delay spectrum estimation, the embodiments of the present application can quickly select matched filter coefficients for each type of channel, and then use the selected matched filter coefficients to achieve the effect of smooth denoising on the corresponding type of channel, and the calculation is simple. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a specific process of the denoising method based on the selected matched filter coefficient according to the embodiments of the present application Figure 1 ;

[0019] Figure 2 is a specific process of the denoising method based on the selected matched filter coefficient according to the embodiments of the present application Figure 2 ;

[0020] Figure 3 This is a detailed flowchart of the noise reduction method based on selected matched filter coefficients according to an embodiment of the present invention. Figure 3 ;

[0021] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings. However, those skilled in the art will understand that many technical details have been presented in the various embodiments of the present invention to enable the reader to better understand this application. However, the technical solutions claimed in this application can be implemented even without these technical details and various changes and modifications based on the following embodiments.

[0023] One embodiment of the present invention relates to a noise reduction method based on selected matched filter coefficients, such as... Figure 1 As shown, the noise reduction method based on selected matched filter coefficients provided in this embodiment includes the following steps.

[0024] Step 101: Classify the channel according to its time-domain and / or frequency-domain characteristics to obtain multiple channel categories.

[0025] In communication systems, due to the wide beamwidth of ground station antennas, and influenced by various factors such as terrain features, topography, and sea conditions, the receiver receives electromagnetic waves that arrive via several paths, including refraction, reflection, and direct transmission. This phenomenon is known as multipath effect. Multipath effect results in different characteristics of the channel in both the time and frequency domains.

[0026] The electromagnetic rays arriving via different paths have inconsistent phases and are time-varying, causing the received signal to fade. The varying arrival times of these rays also lead to inter-symbol interference. If the intensity of the multiple rays is high and the time delay difference cannot be ignored, bit errors will occur. This results in a distorted electromagnetic signal received by the receiver, which is a composite of multiple electromagnetic waves rather than the original electromagnetic wave. Increasing the transmission power cannot eliminate this type of bit error, and the fading caused by this multipath effect is called multipath fading.

[0027] Multipath fading caused by multipath effects can cause channel differences in both the time and frequency domains. Therefore, channels can be classified according to their time and / or frequency characteristics to obtain different types of channels. The classification results of channels are not limited here.

[0028] Step 102: Provide multiple sets of matched filter coefficients.

[0029] The multiple sets of matching filter coefficients correspond to the multiple channel categories one by one.

[0030] Specifically, in the signal processing process, an FFT (Fast Fourier Transform) transform is usually performed, and the FFT transform can only transform time domain data of a limited length, and thus, the time domain signal needs to be truncated. Even if the signal is periodic, if the truncated time length is not an integer multiple of the period (periodic truncation), the truncated signal will have leakage. In order to minimize this leakage error, a weighting function, i.e., a matching filter coefficient, needs to be used.

[0031] Since the channels are classified according to time domain characteristics and / or frequency domain characteristics, in order to select the most suitable matching filter coefficient for each channel, one or more sets of matching filter coefficients corresponding to the channel category of each channel can be provided according to the channel category. Each set of matching filter coefficients includes one or more matching filter coefficients. Different matching filter coefficients will process the signal differently and output different results. The matching filter coefficient can be a function such as a sine function or a gate function used in the signal processing process, which is not limited in the present application. Each set of matching filter coefficients includes a set of time delay spread widths and / or a set of Doppler shift sizes.

[0032] Step 103: Matching the time domain characteristics and / or frequency domain characteristics of the channels of each channel category with the set of matching filter coefficients corresponding to the channel category, respectively, to obtain a set of matching values corresponding to each channel category;

[0033] Specifically, since the channels are classified according to time domain characteristics and / or frequency domain characteristics, the channels of the same category have the same time domain characteristics and / or frequency domain characteristics. The time domain characteristics and / or frequency domain characteristics corresponding to each channel are matched with the set of matching filter coefficients corresponding to the channel. The matching calculation can be calculated by using a correlation coefficient or by using the variance of the residual error, which is not limited in the present application. Since a set of matching filter coefficients includes one or more matching filter coefficients, one or more matching values corresponding to each channel can be obtained after matching to form a set of matching values.

[0034] Step 104: Determining a matching value in the set of matching values corresponding to each channel category, and taking the matching filter coefficient corresponding to the matching value as the target matching filter coefficient corresponding to the channel category;

[0035] Specifically, after the time-domain feature and / or the frequency-domain feature of each channel is matched with each of the matching filter coefficients corresponding to the channel to obtain a set of matching values, one matching value is selected from all the matching values, the unique matching filter coefficient corresponding to the matching value is determined according to the matching value, and the matching filter coefficient is taken as the target matching filter coefficient of the channel. In the process of selecting one matching value from the set of matching values, the largest matching value can be selected from the set of matching values; or a threshold value can be set, and one matching value greater than the threshold value is selected from all the matching values. The specific selection method is not specifically limited here.

[0036] Step 105: smoothing and denoising the target channel based on the target matching filter coefficient.

[0037] Specifically, after the target matching filter coefficient corresponding to each type of channel is obtained, the target matching filter coefficient is applied to design a filter to filter the channel of the type, so as to achieve the purpose of smoothing and denoising each type of channel. The method of smoothing and denoising includes any one of the following methods: sliding average method, Savitzky-Golay method, method of processing outliers, wavelet denoising method, etc., which are not specifically limited herein.

[0038] Compared with the related art, the embodiment of the present application can select one matching filter coefficient by matching the time-domain feature and / or the frequency-domain feature of each channel with the corresponding set of matching filter coefficients, take the matching filter coefficient as the target matching filter coefficient corresponding to the channel, and smooth and denoise the channel of the type by the target matching coefficient, so as to quickly select the matching filter coefficient of each type of channel without relying on historical measurement information and power delay spectrum estimation, and then use the selected matching filter coefficient to achieve the effect of smoothing and denoising the corresponding type of channel, and the calculation is simple.

[0039] Another embodiment of the present application relates to a denoising method based on the selected matching filter coefficient, which is an improvement of the foregoing embodiment, and the improvement is that the specific execution process of step 101 is refined.

[0040] In one example, the time-domain feature of the channel at least includes the multipath time delay spread of the channel, and / or the frequency-domain feature of the channel at least includes the maximum Doppler shift of the channel.

[0041] Specifically, multipath fading caused by multipath effect can cause the channel to be different in time domain and frequency domain. Because the distance of the wave through each path is different, the arrival time of the transmitted wave in each path is different, causing the time domain characteristics of the channel to be different, which is usually manifested as different multipath time delay spread. Due to the difference in propagation distance, the phase and frequency are usually changed, and the frequency domain characteristics of the channel are also different, which is called Doppler shift. That is, in this process, the time domain characteristics of the channel at least include the multipath time delay spread of the channel, and the frequency domain characteristics of the channel at least include the maximum Doppler shift of the channel.

[0042] On this basis, step 101 of the above-mentioned embodiment corresponds to the following in this embodiment: according to the multipath time delay spread and / or the maximum Doppler shift of the channel, the channel is classified into different categories.

[0043] Specifically, when classifying the channel according to the multipath time delay spread and / or the maximum Doppler shift of the channel, the multipath time delay spread of the channel can be classified from narrow to wide and / or the maximum Doppler shift of the channel can be classified from small to large, to obtain multiple channel categories.

[0044] Specifically, the classification can be performed according to the multipath time delay spread, and channels with a specific width of time delay spread are classified into a category, and each category corresponds to a set of matched filtering coefficients. The classification can also be performed according to the maximum Doppler shift, and channels with a specific size of Doppler shift are classified into a category, and each category corresponds to a set of matched filtering coefficients. The classification can also be performed according to the multipath time delay spread and the maximum Doppler shift, and channels with a specific width of time delay spread and a specific size of Doppler shift are classified into a category. The specific classification method can be adjusted according to actual needs, and the present application does not specifically limit this.

[0045] Compared with the related art, the present embodiment classifies the multipath time delay spread of the channel from narrow to wide and / or the maximum Doppler shift of the channel from small to large, to obtain multiple channel categories. The present embodiment takes into account that the actual communication environment of the channel can be different, and adjusts the classification of the channel by setting different classification standards, so that the channel classification result is more reasonable.

[0046] Another embodiment of the present application relates to a noise reduction method based on selecting matched filtering coefficients. The present embodiment is an improvement on the foregoing embodiments, and the improvement is that the specific execution process of step 103 is refined, as shown in the following. Figure 2 As shown in the following, step 103 includes the following steps.

[0047] Step 1031: respectively convolve each channel's time domain feature and / or frequency domain feature with all the matched filtering coefficients in the matched filtering coefficient group corresponding to the channel to obtain a plurality of convolution values corresponding to the channel.

[0048] Specifically, since each matched filtering coefficient group contains at least one matched filtering coefficient, that is, each channel's time domain feature and / or frequency domain feature can match at least one matched value with each matched filtering coefficient group. Each channel's time domain feature and / or frequency domain feature is respectively convolved with all the matched filtering coefficients in the matched filtering coefficient group corresponding to the channel to obtain at least one convolution value.

[0049] Step 1032: take the plurality of convolution values as a matched value group, and the matched value group corresponds to the matched filtering coefficient group.

[0050] Specifically, the plurality of convolution values obtained by respectively convolving each channel's time domain feature and / or frequency domain feature with all the matched filtering coefficients in the matched filtering coefficient group corresponding to the channel are taken as a matched value group. Since the matched value group is calculated by each matched filtering coefficient in the matched filtering coefficient group and the channel's time domain feature and / or frequency domain feature, each matched value in the matched value group corresponds to each matched filtering coefficient in the matched filtering coefficient group.

[0051] In one example, step 103 can further include: intercepting part of the channel's time domain feature and / or frequency domain feature from the channel's time domain feature and / or frequency domain feature of each channel category, and matching the part of the channel's time domain feature and / or frequency domain feature with the matched filtering coefficient group corresponding to the channel category.

[0052] Specifically, before matching each channel's time domain feature and / or frequency domain feature with a matched filtering coefficient group, due to the uncertainty of channel length, each channel can be randomly sampled or intercepted to obtain a part of the channel, and each channel's time domain feature and / or frequency domain feature of the part of the channel is applied to match with a plurality of matched filtering coefficient groups.

[0053] Compared with the related art, the embodiment respectively convolves each channel's time domain feature and / or frequency domain feature with all the matched filtering coefficients in the matched filtering coefficient group corresponding to the channel to obtain a plurality of convolution values corresponding to the channel; and takes the plurality of convolution values as a matched value group. Meanwhile, the target channel is randomly sampled or intercepted to obtain a part of the channel, and each channel's time domain feature and / or frequency domain feature of the part of the channel is applied to match with each matched filtering coefficient group, which can greatly reduce the calculation amount of matching calculation and speed up the calculation speed.

[0054] Another embodiment of the present application relates to a noise reduction method based on selection of matched filter coefficients. The embodiment is an improvement of the foregoing embodiment, and the improvement is that the specific execution process of step 104 is specified as follows. Figure 3 As shown in the figure, step 104 includes the following steps.

[0055] Step 1041: Determine whether there is a matched value with a confidence greater than a preset threshold in the matched value set.

[0056] Specifically, after obtaining a matched value set corresponding to a matched filter coefficient set, it is determined whether there is a matched value with a confidence greater than a preset threshold in the matched value set. For example, the preset threshold is x, and the matched value set is a1, a2, a3, and the confidence of each matched value is p1=a1 / sum(a), p2=a2 / sum(a), and p3=a3 / sum(a), where sum(a)=a1+a2+a3. At this time, it is determined whether there is a value greater than x by comparing the values of p1, p2, and p3 with x.

[0057] Step 1042: If there is, select a matched value from all matched values with a confidence greater than the preset threshold, and use the matched filter coefficient corresponding to the matched value as the target matched filter coefficient of the channel category corresponding to the matched value set.

[0058] Specifically, if there is only one matched value with a confidence greater than the preset threshold in the matched value set, the matched filter coefficient corresponding to the matched value is used as the target matched filter coefficient of the channel category corresponding to the matched value set. If there are multiple matched values with a confidence greater than the preset threshold in the matched value set, the matched filter coefficient corresponding to the maximum value among the multiple matched values can be selected as the target matched filter coefficient of the channel category corresponding to the matched value set, or a matched value can be selected at will from the multiple matched values, and the matched filter coefficient corresponding to the matched value is used as the target matched filter coefficient of the channel category corresponding to the matched value set. The selection method of selecting one matched value from all matched values with a confidence greater than the preset threshold is not limited in the present application.

[0059] In one example, when it is determined whether there is a matched value with a confidence greater than a preset threshold in the matched value set, if there is not, the preset matched filter coefficient is used as the target matched filter coefficient.

[0060] In addition to the steps 1041-1042, the step 104 of determining a matching value in the set of matching values corresponding to each channel category, the selection method of the matching filter coefficient corresponding to the matching value as the target matching filter coefficient of the channel category corresponding to the set of matching values can further include another step completely different from the steps 1041-1042, that is, determining the maximum matching value in the set of matching values, and taking the matching filter coefficient corresponding to the maximum matching value as the target matching filter coefficient of the target channel corresponding to the set of matching values.

[0061] Compared with the related art, the embodiment of the present application determines whether there is a matching value with a confidence greater than a preset threshold in the set of matching values, and if there is, selects a matching value in all the matching values with a confidence greater than the preset threshold, and takes the matching filter coefficient corresponding to the matching value as the target matching filter coefficient of the channel category corresponding to the set of matching values, thereby most effectively ensuring the effectiveness of the matching filter coefficient in subsequent smoothing and noise reduction, and ensuring the smoothing and noise reduction effect.

[0062] Another embodiment of the present application relates to an electronic device, such as Figure 4 As shown in the figure, the electronic device includes at least one processor 202, and a memory 201 connected with the at least one processor 202; wherein the memory 201 stores instructions executable by the at least one processor 202, and the instructions are executed by the at least one processor 202 to enable the at least one processor 202 to execute any of the above method embodiments.

[0063] The memory 201 and the processor 202 are connected in a bus manner, the bus can include any number of interconnected buses and bridges, and the bus connects one or more processors 202 and various circuits of the memory 201 together. The bus can also connect various other circuits such as peripheral devices, voltage stabilizers, and power management circuits together, which are well known in the art, and thus, they will not be further described herein. The bus interface provides an interface between the bus and the transceiver. The transceiver can be one element or multiple elements such as multiple receivers and transmitters, which provide a unit for communicating with various other devices on the transmission medium. The data processed by the processor 202 is transmitted on the wireless medium through the antenna, and further, the antenna also receives data and transmits the data to the processor 202.

[0064] The processor 202 is responsible for managing the bus and general processing, and can also provide various functions including timing, peripheral interface, voltage regulation, power management, and other control functions. And the memory 201 can be used to store the data used by the processor 202 in the execution of the operation.

[0065] Another embodiment of the present application relates to a computer readable storage medium storing a computer program. The computer program, when executed by a processor, implements any of the above method embodiments.

[0066] That is, those skilled in the art can understand that all or part of the steps in the above method embodiments can be completed by programs instructing relevant hardware, the programs are stored in a storage medium, and include a plurality of instructions for causing a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the method described in various embodiments of the present application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0067] Those skilled in the art can understand that the above embodiments are specific embodiments for implementing the present application, and in actual applications, various changes can be made in form and details without departing from the spirit and scope of the present application.

Claims

1. A noise reduction method based on selected matched filter coefficients, characterized in that, include: Channels are classified according to their time-domain and / or frequency-domain characteristics to obtain multiple channel categories; Provides multiple sets of matched filter coefficients; The multiple sets of matched filter coefficients correspond one-to-one with the multiple channel categories; The time-domain and / or frequency-domain characteristics of the channel for each channel category are matched with a set of matched filter coefficients corresponding to that channel category to obtain a set of matched values ​​for each channel category. In each channel category, a matching value is determined from a set of matching values, and the matching filter coefficient corresponding to that matching value is used as the target matching filter coefficient for that channel category. The target matched filter coefficients corresponding to each channel category are used to smooth and reduce noise in the channel of that channel category; The step of matching the time-domain and / or frequency-domain features of each channel category with a set of matched filter coefficients corresponding to that channel category to obtain a set of matching values ​​for each channel category includes: The time-domain features and / or frequency-domain features of each channel are convolved with all the matched filter coefficients in a set of matched filter coefficients corresponding to that channel to obtain multiple convolution values ​​corresponding to that channel. The multiple convolutional values ​​are used as a set of matching values, and the set of matching values ​​corresponds to the set of matching filter coefficients.

2. The method according to claim 1, characterized in that, The time-domain characteristics of the channel include at least the multipath delay spread of the channel, and / or the frequency-domain characteristics of the channel include at least the maximum Doppler shift of the channel; The channel is classified according to its time-domain and / or frequency-domain characteristics, resulting in multiple channel categories, including: The channels are classified into different types based on the multipath delay spread and / or the maximum Doppler shift of the channel.

3. The method according to claim 2, characterized in that, Based on the multipath delay spread and / or maximum Doppler shift of the channel, the channel is classified into different types, including: The multipath delay spread of the channel is classified from narrow to wide and / or the maximum Doppler shift of the channel is classified from small to large to obtain multiple channel categories.

4. The method according to claim 1, characterized in that, The step of determining a matching value from a set of matching values ​​corresponding to each channel category, and using the matching filter coefficient corresponding to that matching value as the target matching filter coefficient for that channel category, includes: Determine whether there are any matching values ​​in this group with a confidence level greater than a preset threshold; If a match exists, select one match from all match values ​​with a confidence level greater than a preset threshold, and use the match filter coefficient corresponding to that match value as the target match filter coefficient for the channel category corresponding to that set of match values.

5. The method according to claim 4, characterized in that, The step of determining whether there are any matching values ​​in the group of matching values ​​with a confidence level greater than a preset threshold also includes: If it does not exist, the preset matched filter coefficients will be used as the target matched filter coefficients.

6. The method according to claim 1, characterized in that, The step of determining a matching value from a set of matching values ​​corresponding to each channel category, and using the matching filter coefficient corresponding to that matching value as the target matching filter coefficient for that channel category, includes: Determine the maximum matching value in a set of matching values, and use the matching filter coefficient corresponding to the maximum matching value as the target matching filter coefficient for the target channel corresponding to the set of matching values.

7. The method according to claim 1, characterized in that, The time-domain and / or frequency-domain characteristics of each channel category are matched with a set of matched filter coefficients corresponding to that channel category, including: Extract a portion of the time-domain and / or frequency-domain features of the channel from the time-domain and / or frequency-domain features of the channel for each channel category, and match the time-domain and / or frequency-domain features of the portion of the channel with a set of matched filter coefficients corresponding to that channel category.

8. An electronic device, characterized in that, include: At least one processor; as well as, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the noise reduction method based on selected matched filter coefficients as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the noise reduction method based on selected matched filter coefficients as described in any one of claims 1 to 7.

Images