Signal interception method and device, electronic equipment and storage medium

Abstract

The invention provides a signal interception method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the signal filtering of an obtained to-be-intercepted signal through a first deep learning model, and obtaining a target to-be-intercepted signal and target time-frequency feature information corresponding to the target to-be-intercepted signal; performing protocol identification on the target to-be-monitored signal by using a second deep learning model to obtain a communication protocol of the target to-be-monitored signal; performing information reduction on the target to-be-monitored signal based on the communication protocol to obtain target information corresponding to the target to-be-monitored signal; and the target to-be-monitored signal, the target time-frequency characteristic information, the communication protocol and the target information are stored, so that the efficiency and accuracy of signal monitoring are improved. Besides, dynamic fusion among the time-frequency characteristics, the communication protocol and the target information of the signals is realized, complete data support is provided for signal traceability, and a full-process intelligent solution is provided for the fields of public safety, emergency management, radio monitoring management and the like.

Description

Technical Field

[0001] The present invention relates to the field of signal monitoring technology, and in particular to a signal monitoring method, device, electronic equipment and storage medium. Background Art

[0002] Intercom signal interception is widely used in public safety, emergency management, radio monitoring and management, and other fields. Its core goal is to obtain, analyze and utilize information from intercom communications. Currently, the common intercom interception methods are as follows: Direct frequency scanning method: Walkie-talkies typically operate in specific frequency bands (such as VHF 136-174MHz and UHF 400-470MHz). By scanning a specific frequency band, active frequencies can be quickly discovered and analog voice signals can be received, completing the walkie-talkie signal detection task.

[0003] Based on the modulation and protocol identification method: by analyzing the modulation characteristics of the signal (modulation frequency, symbol rate, etc.), the frame header information is cracked, and then the communication protocol is extracted from the frame header information through multi-channel parallel processing, and the walkie-talkie signal is monitored based on the communication protocol.

[0004] Machine learning-based listening methods: By learning signal characteristics (such as spectral characteristics and time series characteristics), they are identified using traditional machine learning classifiers.

[0005] However, the above-mentioned intercom monitoring method has the following problems: 1. The direct frequency scanning method is only suitable for monitoring traditional analog walkie-talkies. If the walkie-talkie signal is in digital mode or encrypted mode, the direct frequency scanning method can only receive noise and cannot decode the content. In other words, the direct frequency scanning method cannot monitor digital or encrypted walkie-talkie signals.

[0006] 2. When using modulation and protocol recognition methods for intercom interception, due to the redundant behavior of multi-channel parallel processing, the computing power and resource space consumption are large, resulting in low recognition efficiency of intercom signals based on modulation and protocol recognition methods.

[0007] 3. The listening method based on machine learning has high requirements for the signal-to-noise ratio environment. In a complex electromagnetic environment, the listening method based on machine learning is easily affected by noise, resulting in a low accuracy rate of walkie-talkie signal listening.

[0008] Therefore, how to improve the efficiency and accuracy of signal monitoring is an urgent problem to be solved. Summary of the Invention

[0009] The present invention provides a signal monitoring method, device, electronic equipment and storage medium, which can solve the problem of low efficiency and accuracy of signal monitoring.

[0010] According to a first aspect of the present invention, a signal monitoring method is provided, the method comprising: Get the signal to be listened; Performing signal filtering on the signal to be intercepted using a first deep learning model to obtain a target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted, wherein the first deep learning model is a signal filtering model; Performing protocol identification on the target signal to be intercepted using a second deep learning model to obtain a communication protocol of the target signal to be intercepted, where the second deep learning model is a protocol identification model; Performing information restoration on the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted; The target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information are stored.

[0011] According to a second aspect of the present invention, a signal monitoring device is provided, the device comprising: A signal acquisition module is used to obtain the signal to be monitored; a signal filtering module, configured to perform signal filtering on the signal to be detected using a first deep learning model to obtain a target signal to be detected and target time-frequency feature information corresponding to the target signal to be detected, wherein the first deep learning model is a signal filtering model; a protocol identification module, configured to perform protocol identification on the target signal to be intercepted using a second deep learning model to obtain a communication protocol of the target signal to be intercepted, wherein the second deep learning model is a protocol identification model; An information restoration module, configured to restore information of the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted; The information storage module is used to store the target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information.

[0012] According to a third aspect of the present invention, there is provided an electronic device comprising a processor and a memory, The memory is used to store codes and related data; The processor is configured to execute the code in the memory to implement the signal monitoring method as described in any one of the embodiments of the present invention.

[0013] According to a fourth aspect of the present invention, a storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the signal monitoring method as described in any one of the embodiments of the present invention is implemented.

[0014] Compared with the prior art, the signal monitoring method provided by the present invention has the following beneficial effects: A signal to be intercepted (e.g., a walkie-talkie signal) is obtained and filtered using a first deep learning model to obtain a target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted. The first deep learning model is a signal filtering model. Specifically, through signal filtering, the frequencies corresponding to interfering signals and noise are suppressed or attenuated, reducing their impact and preserving the valid signal segments of the target signal to be intercepted. Furthermore, because the deep learning model has excellent feature extraction capabilities in low signal-to-noise ratio environments, it can extract more reliable target signals in these environments, thereby improving the signal quality of the target signals to be intercepted.

[0015] Then, a second deep learning model is used to perform protocol identification on the target signal to obtain the communication protocol of the target signal. The second deep learning model serves as a protocol identification model. Since only the target signal needs to be input into the second deep learning model for protocol identification, there is no need to use multi-path parallel processing to identify the communication protocol of the target signal, reducing redundant behavior. Furthermore, the deep learning model has good feature extraction capabilities in low signal-to-noise ratio environments, and the signal quality of the target signal is high. Therefore, the efficiency and accuracy of communication protocol recognition are improved, and the error rate of communication protocol misjudgment is reduced.

[0016] Finally, based on the high-accuracy communication protocol, the high-quality target signal to be intercepted is restored to obtain the target information corresponding to the target signal to be intercepted, thereby improving the restoration degree of the target information; storing the target signal to be intercepted, the target time-frequency feature information, the communication protocol and the target information is equivalent to constructing the correlation relationship between the time-frequency features of the signal, the communication protocol and the target information, realizing the dynamic fusion between the time-frequency features of the signal, the communication protocol and the target information, and forming a standardized structured storage scheme for the correlation information of the signal (the time-frequency features of the signal, the communication protocol and the target information), which can provide complete data support for signal tracing, and thus provide a full-process intelligent solution for the fields of public safety, emergency management, radio monitoring management, etc. In addition, the signal interception method provided by the present invention, compared with the traditional walkie-talkie signal interception method, not only improves the signal quality of the target signal to be intercepted, but also improves the recognition efficiency and accuracy of the communication protocol of the target signal to be intercepted. Therefore, signal interception based on the high-quality target signal to be intercepted greatly improves the signal interception efficiency and signal interception accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 This is a flow chart of a signal monitoring method provided by an embodiment of the present invention; Figure 2 is a schematic diagram of a signal monitoring method provided by an embodiment of the present invention; Figure 3 is another flowchart of the signal monitoring method provided by an embodiment of the present invention; Figure 4 is a structural diagram of a signal monitoring device provided by an embodiment of the present invention; Figure 5 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] The terms "first," "second," "third," "fourth," and so on (if any) in the description and claims of the present invention and in the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate, so that the embodiments of the present invention described herein can be implemented in orders other than those illustrated or described herein. In addition, the terms "including" and "having," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus that includes a series of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, product, or apparatus.

[0021] The following specific embodiments are used to describe the technical solution of the present invention in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0022] Figure 1This is a flow chart of a signal interception method provided by an embodiment of the present invention. The method can be executed by a signal interception device, which can be implemented in software and / or hardware. In a specific embodiment, the device can be integrated into an electronic device, such as a computer, a server, etc. The following embodiments will be described using the device integrated into an electronic device as an example. Figure 1 , the method may specifically include the following steps: Step 101: Acquire a signal to be monitored.

[0023] The signal to be intercepted may be a walkie-talkie signal. In other embodiments, the signal to be intercepted may also be other radio voice communication signals.

[0024] In one embodiment, obtaining the signal to be listened to may include: using a signal detection device to obtain an original signal, and filtering the original signal according to a spectrum analysis method to obtain the signal to be listened to. In this way, the interference signal or noise in the original signal can be filtered out based on the spectrum analysis method, that is, the existence of the signal to be listened to is determined, and the signal quality of the signal to be listened to is improved.

[0025] For example, if the original signal is a walkie-talkie signal, the original signal can be filtered using a spectrum analysis method to obtain a signal to be detected (a walkie-talkie signal to be detected). Specifically, the spectrum analysis method is used to first determine whether a walkie-talkie signal is present. If a walkie-talkie signal is determined to be present, the identified walkie-talkie signal is determined as the signal to be detected.

[0026] Specifically, screening the original signal according to the spectrum analysis method to obtain the signal to be monitored may include: marking the original time-frequency feature map of the original signal according to the spectrum analysis method to obtain the signal to be monitored.

[0027] The marking method may be a mark box mark or other marking methods, which are not specifically limited here.

[0028] For example, the signal to be intercepted is a walkie-talkie signal. Figure 2 As shown, the original signal of the intercom is obtained by using a signal detection device, and the original signal is filtered according to a spectrum analysis method to obtain a signal to be listened (intercom signal to be listened).

[0029] Step 102: Filter the signal to be intercepted using the first deep learning model to obtain the target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted.

[0030] The first deep learning model can be understood as a deep learning model used to process and filter the initial signal. That is, the first deep learning model is a signal filtering model. The first deep learning model may include a time-frequency feature map extraction network, a signal detection network, and a signal filtering network. The signal filtering network may be a two-dimensional time-frequency filter. Specifically, the corresponding two-dimensional time-frequency filter may be pre-designed based on the signal's time-frequency feature information. The target signal to be detected can be understood as the signal obtained through filtering by the first deep learning model. The time-frequency feature information can be understood as the characteristic representation of the signal in both time and frequency dimensions. The time-frequency feature information may include, but is not limited to, one or more of the signal's bandwidth, time slot information, signal center frequency, and frequency hopping pattern information. The time slot information may include the signal's start time, end time, and time slot, and the time slot information can be determined based on the signal's start time and end time. The frequency hopping pattern information may include the signal's start frequency, end frequency, and frequency hopping pattern, and the frequency hopping pattern information can be determined based on the signal's start frequency and end frequency. The target time-frequency feature information can be understood as the signal's time-frequency feature information obtained through filtering by the first deep learning model.

[0031] In one embodiment, using a first deep learning model to perform signal filtering on a signal to be listened to, and obtaining a target signal to be listened to and target time-frequency feature information corresponding to the target signal to be listened to, can include: inputting the signal to be listened to into a time-frequency feature map extraction network to extract the time-frequency feature map to obtain a time-frequency feature map of the signal to be listened to; inputting the time-frequency feature map into a signal detection network to extract the time-frequency features to obtain the time-frequency feature information of the signal to be listened to; inputting the signal to be listened to and the time-frequency feature information into a signal filtering network for filtering to obtain the target signal to be listened to and the target time-frequency feature information corresponding to the target signal to be listened to.

[0032] In this embodiment, the time-frequency feature map extraction network, signal detection network, and signal filtering network are all based on deep learning network structures. The deep learning model has a strong time-frequency feature extraction capability and can more comprehensively and quickly extract the time-frequency feature information of the target signal, thereby increasing the comprehensiveness of the time-frequency feature information. Furthermore, the signal filtering network filters the time-frequency feature information extracted by the signal detection network, further reducing interfering signals or noise in the target signal, thereby improving the signal quality of the target signal.

[0033] Specifically, the time-frequency feature map is input into the signal detection network for time-frequency feature extraction to obtain the time-frequency feature information of the signal to be listened to. It can be understood as inputting the time-frequency feature map into the signal detection network, marking the suspected signal to be listened to (for example, a walkie-talkie signal) through the signal detection network and outputting the time-frequency feature information of the marked signal.

[0034] The signal marking method may be a marking box or other marking methods, which are not specifically limited here.

[0035] Specifically, filtering the time-frequency feature information through the signal filtering network may include: filtering the target time-frequency feature information that meets preset conditions from the time-frequency feature information based on the bandwidth, time slot, and frequency hopping pattern in the time-frequency feature information through the signal filtering network. The preset conditions may include: 1. The bandwidth falls within a preset bandwidth range; 2. The time slot falls within a preset time slot range; 3. The frequency hopping pattern meets a preset frequency hopping pattern; and 4. The signal center frequency falls within a preset center frequency range.

[0036] It should be noted that if the time-frequency characteristic information includes other information in addition to bandwidth, time slot, frequency hopping pattern, and signal center frequency, conditions corresponding to other information in addition to bandwidth, time slot, frequency hopping pattern, and signal center frequency can be added to the preset conditions to improve the accuracy of signal filtering.

[0037] For example, the signal to be intercepted is a walkie-talkie signal. Figure 2 As shown, a signal to be detected can be obtained and input into a time-frequency feature map extraction network for time-frequency feature map extraction to obtain a time-frequency feature map of the signal to be detected. The time-frequency feature map is then input into a signal detection network for time-frequency feature extraction to obtain time-frequency feature information of the signal to be detected (the time-frequency feature information may include: start time, end time, start frequency, cutoff frequency, time slot, bandwidth, and frequency hopping pattern). Finally, the signal to be detected and the time-frequency feature information are input into a signal filtering network for filtering to obtain a target signal to be detected and target time-frequency feature information corresponding to the target signal to be detected.

[0038] Step 103: Use the second deep learning model to perform protocol identification on the target signal to be intercepted to obtain the communication protocol of the target signal to be intercepted.

[0039] The second deep learning model can be understood as a deep learning model for identifying the communication protocol of the signal. The second deep learning model is a protocol recognition model.

[0040] In one embodiment, a second deep learning model is used to perform protocol identification on the target signal to be listened to, and a communication protocol of the target signal to be listened to is obtained. The second deep learning is a protocol identification model, which may include: using the second deep learning model to perform time slot structure identification on the target signal to be listened to, and obtaining the time slot structure of the target signal to be listened to; using the second deep learning model to determine the communication protocol of the target signal to be listened to based on the time slot structure.

[0041] Specifically, using the second deep learning model to identify the time slot structure of the target signal to be listened to and obtaining the time slot structure of the target signal to be listened to can include: using the second deep learning model to obtain the signal frame of the target signal to be listened to and determining the time slot structure of the target signal to be listened to based on the signal frame.

[0042] The signal frame can be understood as a form of structured segmentation of the signal carried by the signal. The signal frame may include multiple frame synchronization signals of the target signal to be monitored. Since the time slot result of the signal can be determined based on the frame synchronization signal, in a specific embodiment, using the second deep learning model to determine the time slot structure of the target signal to be monitored based on the signal frame may include: using the second deep learning model to determine the time slot structure of the target signal to be monitored based on the multiple frame synchronization signals.

[0043] In this embodiment, the second deep learning model is used to identify the time slot structure of the target signal to be listened to, and the time slot structure of the target signal to be listened to is obtained; the second deep learning model is used to determine the communication protocol of the target signal to be listened to based on the time slot structure, that is, the target signal to be listened to only needs to be input into the second deep learning model for protocol identification to obtain the communication protocol, and there is no need to use multi-channel parallel processing to identify the communication protocol of the target signal to be listened to, thereby reducing redundant behavior; and the deep learning model has better feature extraction capabilities in a low signal-to-noise ratio environment, and the signal quality of the target signal to be listened to is high, thereby improving the recognition efficiency and accuracy of the communication protocol and reducing the misjudgment rate of the communication protocol. In addition, since the second deep learning model contains more prior knowledge, the protocol compatibility of the signal listening method is improved.

[0044] For example, Figure 2 As shown, the target signal to be intercepted is input into the second deep learning model, so as to use the second deep learning model to perform protocol recognition on the target signal to be intercepted to obtain the communication protocol of the target signal to be intercepted.

[0045] Step 104 : performing information restoration on the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted.

[0046] The target information can be understood as the information carried by the target signal to be intercepted. The target information can include target text information and target voice data. The target text information can include control instructions and / or text information.

[0047] In one embodiment, information restoration of a target signal to be intercepted based on a communication protocol to obtain target information corresponding to the target signal to be intercepted may include: determining a protocol encoding and decoding rule corresponding to the communication protocol; bit-level decoding of the target signal to be intercepted according to the protocol encoding and decoding rule to obtain decoded information; determining a signal type of the target signal to be intercepted based on the decoded information; and restoring information corresponding to the target signal to be intercepted based on the signal type to obtain the target information.

[0048] The protocol encoding and decoding rules can be understood as the encoding and decoding rules of the communication protocol. The decoding information can be understood as the information obtained by bit-level decoding the target intercepted signal according to the protocol encoding and decoding rules. The signal types can include voice signals and text signals.

[0049] Specifically, based on the decoded information, determining the signal type of the target signal to be listened to may include: performing protocol layer verification and data structure verification on the decoded information to obtain a verification result; when the verification result is verification passed, extracting the frame header identifier of the decoded information; when the frame header identifier is a first identifier, determining that the signal type of the target signal to be listened to is a voice signal; when the frame header identifier is a second identifier, determining that the signal type of the target signal to be listened to is a text signal.

[0050] The first identifier is a frame header identifier corresponding to the voice signal, and the second identifier is a frame header identifier corresponding to the text signal.

[0051] In this embodiment, the decoded information undergoes protocol layer verification and data structure verification to obtain a verification result. If the verification result indicates a pass, the frame header identifier of the decoded information is extracted, and the signal type is determined based on the frame header identifier. This allows the accuracy of the decoded information to be determined through a two-step verification process of protocol layer verification and data structure verification, thereby avoiding a reduction in the accuracy of signal type determination due to incorrect decoding of the target signal to be intercepted. Furthermore, determining the signal type of the target signal to be intercepted based on the frame header identifier can improve the accuracy of the signal type.

[0052] Specifically, based on the signal type, restoring information corresponding to the target signal to be intercepted to obtain target information may include: if the signal type is a voice signal, determining the voice codec rules corresponding to the communication protocol; performing voice decoding on the target signal to be intercepted according to the voice codec rules to obtain decoded voice data; determining target voice data based on the decoded voice data; and if the signal type is a text signal, performing application layer decoding according to the communication protocol to obtain target text information. This is equivalent to establishing a dual-modal processing channel for the target signal to be intercepted, simultaneously achieving text information decoding and / or voice data decoding of the target signal to be intercepted.

[0053] In a specific embodiment, determining the target voice data based on the decoded voice data may include: using a voice recognition model to identify the decoded voice data for data validity to obtain a recognition result; if the recognition result is valid, performing noise reduction on the decoded voice data to obtain noise-reduced voice data; and performing voice enhancement on the noise-reduced voice data to obtain the target voice data. Specifically, through noise reduction and voice enhancement, the target voice data carried by the target signal to be intercepted is highly restored, thereby enhancing the authenticity of the target voice data carried by the target signal to be intercepted.

[0054] In one embodiment, when the recognition result is invalid, the decoded voice data can be deleted, which is equivalent to providing an adaptive fault-tolerant mechanism. When the decoded voice data is an interference signal or noise, the decoded voice data can be automatically deleted to avoid determining the invalid decoded voice data as the target voice data, thereby improving the data quality of the target voice data of the target signal to be listened to.

[0055] For example, Figure 2 As shown, when the signal type is a voice signal, the voice codec rules corresponding to the communication protocol are determined; the target signal to be monitored is voice-decoded according to the voice codec rules to obtain decoded voice data; the decoded voice data is validated using a voice recognition model to obtain a recognition result; if the recognition result is valid, the decoded voice data is denoised to obtain denoised voice data; the denoised voice data is voice-enhanced to obtain target voice data; if the recognition result is invalid, the decoded voice data is deleted. When the signal type is a text signal, application-layer decoding is performed according to the communication protocol to obtain the target text information.

[0056] Step 105: store the target signal to be intercepted, target time-frequency characteristic information, communication protocol and target information.

[0057] In one embodiment, the target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information can be stored correspondingly in the database, which is equivalent to constructing the correlation relationship between the time-frequency characteristics of the signal, the communication protocol and the target information, realizing the dynamic fusion between the time-frequency characteristics of the signal, the communication protocol and the target information, and forming a standardized structured storage scheme for the associated information of the signal (the time-frequency characteristics of the signal, the communication protocol and the target information); providing complete data support for signal tracing, and then providing a full-process intelligent solution for public safety, emergency management, radio monitoring management and other fields.

[0058] Compared with the prior art, the signal monitoring method provided by the present invention has the following beneficial effects: A signal to be intercepted (e.g., a walkie-talkie signal) is obtained and filtered using a first deep learning model to obtain a target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted. The first deep learning model is a signal filtering model. Specifically, through signal filtering, the frequencies corresponding to interfering signals and noise are suppressed or attenuated, reducing their impact and preserving the valid signal segments of the target signal to be intercepted. Furthermore, because the deep learning model has excellent feature extraction capabilities in low signal-to-noise ratio environments, it can extract more reliable target signals in these environments, thereby improving the signal quality of the target signals to be intercepted.

[0059] Then, a second deep learning model is used to perform protocol identification on the target signal to obtain the communication protocol of the target signal. The second deep learning model serves as a protocol identification model. Since only the target signal needs to be input into the second deep learning model for protocol identification, there is no need to use multi-path parallel processing to identify the communication protocol of the target signal, reducing redundant behavior. Furthermore, the deep learning model has good feature extraction capabilities in low signal-to-noise ratio environments, and the signal quality of the target signal is high. Therefore, the efficiency and accuracy of communication protocol recognition are improved, and the error rate of communication protocol misjudgment is reduced.

[0060] Finally, based on the high-accuracy communication protocol, the high-quality target signal to be intercepted is restored to obtain the target information corresponding to the target signal to be intercepted, thereby improving the restoration degree of the target information; storing the target signal to be intercepted, the target time-frequency feature information, the communication protocol and the target information is equivalent to constructing the correlation relationship between the time-frequency features of the signal, the communication protocol and the target information, realizing the dynamic fusion between the time-frequency features of the signal, the communication protocol and the target information, and forming a standardized structured storage scheme for the correlation information of the signal (the time-frequency features of the signal, the communication protocol and the target information), which can provide complete data support for signal tracing, and thus provide a full-process intelligent solution for the fields of public safety, emergency management, radio monitoring management, etc. In addition, the signal interception method provided by the present invention, compared with the traditional walkie-talkie signal interception method, not only improves the signal quality of the target signal to be intercepted, but also improves the recognition efficiency and accuracy of the communication protocol of the target signal to be intercepted. Therefore, signal interception based on the high-quality target signal to be intercepted greatly improves the signal interception efficiency and signal interception accuracy.

[0061] The signal monitoring method provided by the embodiment of the present invention is further described below. Figure 3 As shown, Figure 3 FIG. 5 is another flow chart of the signal interception method provided by an embodiment of the present invention, which may specifically include the following steps: Step 201: Acquire an original signal using a signal detection device.

[0062] Step 202: Filter the original signal according to a spectrum analysis method to obtain a signal to be monitored.

[0063] Step 203: Input the signal to be monitored into a time-frequency feature map extraction network to extract the time-frequency feature map, and obtain the time-frequency feature map of the signal to be monitored.

[0064] Step 204: Input the time-frequency feature graph into a signal detection network to extract the time-frequency features, and obtain the time-frequency feature information of the signal to be monitored.

[0065] Step 205: Input the signal to be detected and the time-frequency characteristic information into a signal filtering network for filtering to obtain the target signal to be detected and the target time-frequency characteristic information corresponding to the target signal to be detected.

[0066] Step 206: Use the second deep learning model to perform time slot structure recognition on the target signal to be monitored to obtain the time slot structure of the target signal to be monitored.

[0067] Step 207: Determine the communication protocol of the target signal to be monitored according to the time slot structure using the second deep learning model.

[0068] Step 208: Determine the protocol encoding and decoding rules corresponding to the communication protocol.

[0069] Step 209: Decode the target signal to be intercepted at the bit level according to the protocol encoding and decoding rules to obtain decoding information.

[0070] Step 210: Determine the signal type of the target signal to be intercepted based on the decoded information.

[0071] Step 211 , determining whether the signal type is a text signal, if so, executing steps 212 to 213 ; if not, executing steps 214 to 221 .

[0072] Step 212: Perform application layer decoding according to the communication protocol to obtain target text information.

[0073] Step 213: store the target signal to be intercepted, target time-frequency characteristic information, communication protocol and target text information.

[0074] Step 214: Determine the speech encoding and decoding rules corresponding to the communication protocol.

[0075] Step 215: Speech decoding is performed on the target signal to be intercepted according to speech coding and decoding rules to obtain decoded speech data.

[0076] Step 216: Use the speech recognition model to perform data validity recognition on the decoded speech data to obtain a recognition result.

[0077] Step 217 , determining whether the recognition result is valid, if so, executing steps 218 to 220 , if not, executing step 221 .

[0078] Step 218: De-noise the decoded speech data to obtain de-noised speech data.

[0079] Step 219: Perform speech enhancement on the noise reduction speech data to obtain target speech data.

[0080] Step 220: store the target signal to be intercepted, target time-frequency feature information, communication protocol and target voice data.

[0081] Step 221: Delete the decoded voice data.

[0082] The signal monitoring method provided by the embodiment of the present invention suppresses or attenuates the frequencies corresponding to the interference signal and noise by means of signal filtering, reduces the influence of the interference signal and noise, and thereby retains the effective signal segment of the target signal to be monitored. In addition, since the deep learning model has a good feature extraction capability in a low signal-to-noise ratio environment, it is possible to extract a target signal to be monitored with higher reliability in a low signal-to-noise ratio environment, thereby improving the signal quality of the target signal to be monitored. Then, since only the target signal to be monitored needs to be input into the second deep learning model for protocol identification to obtain the communication protocol, there is no need to use a multi-channel parallel processing method to identify the communication protocol of the target signal to be monitored, thereby reducing redundant behavior; and, the deep learning model has a good feature extraction capability in a low signal-to-noise ratio environment, and the signal quality of the target signal to be monitored is high, thereby improving the recognition efficiency and accuracy of the communication protocol and reducing the misjudgment rate of the communication protocol. Finally, the correlation between the time-frequency characteristics of the signal, the communication protocol and the target information is constructed, the dynamic fusion between the time-frequency characteristics of the signal, the communication protocol and the target information is realized, and a structured storage scheme of the correlation information of the standardized signal (the time-frequency characteristics of the signal, the communication protocol and the target information) is formed. This can provide complete data support for signal tracing, and when the target signal to be monitored is a walkie-talkie signal, it can provide a full-process intelligent solution for the field of radio monitoring and management. In addition, the signal monitoring method provided by the present invention, compared with the traditional walkie-talkie signal monitoring method, not only improves the signal quality of the target signal to be monitored, but also improves the recognition efficiency and accuracy of the communication protocol of the target signal to be monitored. Therefore, signal monitoring based on the target signal to be monitored with high signal quality greatly improves the signal monitoring efficiency and signal monitoring accuracy.

[0083] Figure 4 FIG. 1 is a schematic diagram of a structure of a signal interception device provided by an embodiment of the present invention, which is suitable for executing a signal interception method provided by an embodiment of the present invention. Figure 4As shown, the device may specifically include: The signal acquisition module 401 is used to acquire the signal to be monitored; a signal filtering module 402 configured to perform signal filtering on the signal to be intercepted using a first deep learning model to obtain a target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted, wherein the first deep learning model is a signal filtering model; A protocol identification module 403 is configured to perform protocol identification on the target signal to be intercepted using a second deep learning model to obtain a communication protocol of the target signal to be intercepted, where the second deep learning model is a protocol identification model; An information restoration module 404 is configured to restore information of the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted; Information storage module 405, used to store the target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information Optionally, the information restoration module 404 is specifically configured to: Determine the protocol encoding and decoding rules corresponding to the communication protocol; Decoding the target signal at the bit level according to the protocol encoding and decoding rules to obtain decoding information; Determining the signal type of the target signal to be intercepted based on the decoded information; Based on the signal type, information corresponding to the target signal to be intercepted is restored to obtain target information.

[0084] Optionally, the information restoration module 404 determines the signal type of the target signal to be intercepted based on the decoded information, including: Performing protocol layer verification and data structure verification on the decoded information to obtain a verification result; When the verification result is verification passed, extracting the frame header identifier of the decoding information; When the frame header identifier is the first identifier, determining that the signal type of the target signal to be monitored is a voice signal; When the frame header identifier is the second identifier, determining that the signal type of the target signal to be intercepted is a text signal; The first identifier is a frame header identifier corresponding to the voice signal, and the second identifier is a frame header identifier corresponding to the text signal.

[0085] Optionally, the signal type includes a voice signal and a text signal, and the target information includes target text information and target voice data. The information restoration module 404 restores information corresponding to the target to-be-intercepted signal based on the signal type to obtain the target information, including: When the signal type is the voice signal, determining a voice encoding and decoding rule corresponding to the communication protocol; Performing voice decoding on the target signal to be intercepted according to the voice coding and decoding rules to obtain decoded voice data; determining the target voice data based on the decoded voice data; When the signal type is the text signal, application layer decoding is performed according to the communication protocol to obtain the target text information.

[0086] Optionally, the information restoration module 404 determines the target voice data based on the decoded voice data, including: Performing data validity identification on the decoded speech data using a speech recognition model to obtain a recognition result; When the recognition result is valid, performing noise reduction on the decoded voice data to obtain noise-reduced voice data; Speech enhancement is performed on the noise-reduced speech data to obtain the target speech data.

[0087] Optionally, the protocol identification module 403 is specifically configured to: Using the second deep learning model to perform time slot structure identification on the target signal to be monitored, to obtain the time slot structure of the target signal to be monitored; The second deep learning model is used to determine the communication protocol of the target signal to be monitored according to the time slot structure.

[0088] Optionally, the first deep learning model includes a time-frequency feature map extraction network, a signal detection network, and a signal filtering network, and the signal filtering module 402 is specifically configured to: Inputting the signal to be monitored into the time-frequency feature map extraction network to extract the time-frequency feature map, thereby obtaining the time-frequency feature map of the signal to be monitored; Inputting the time-frequency feature graph into the signal detection network to extract time-frequency features to obtain time-frequency feature information of the signal to be monitored; The signal to be detected and the time-frequency characteristic information are input into the signal filtering network for filtering to obtain the target signal to be detected and the target time-frequency characteristic information corresponding to the target signal to be detected.

[0089] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional modules is used as an example for illustration. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the functional modules described above can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0090] The signal monitoring device provided by the embodiment of the present invention has the same beneficial effects as the signal monitoring method provided by the embodiment of the present invention.

[0091] Figure 5 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0092] Please refer to Figure 5 , provides an electronic device 50, comprising: processor 51; and a memory 52 for storing executable instructions of the processor; The processor 51 is configured to execute the above-mentioned method by executing the executable instructions.

[0093] The processor 51 can communicate with the memory 52 via a bus 53 .

[0094] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned method when executed by a processor.

[0095] Those skilled in the art will appreciate that all or part of the steps in the above-described method embodiments can be implemented using hardware associated with program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0096] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A signal monitoring method, characterized in that: The method comprises: Get the signal to be listened; Performing signal filtering on the signal to be intercepted using a first deep learning model to obtain a target signal to be intercepted and target time-frequency feature information corresponding to the target signal to be intercepted, wherein the first deep learning model is a signal filtering model; Performing protocol identification on the target signal to be intercepted using a second deep learning model to obtain a communication protocol of the target signal to be intercepted, where the second deep learning model is a protocol identification model; Performing information restoration on the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted; The target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information are stored.

2. The method according to claim 1, characterized in that The performing information restoration on the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted includes: Determine the protocol encoding and decoding rules corresponding to the communication protocol; Decoding the target signal at the bit level according to the protocol encoding and decoding rules to obtain decoding information; Determining the signal type of the target signal to be intercepted based on the decoded information; Based on the signal type, information corresponding to the target signal to be intercepted is restored to obtain target information.

3. The method according to claim 2, characterized in that The determining, based on the decoded information, the signal type of the target signal to be intercepted includes: Performing protocol layer verification and data structure verification on the decoded information to obtain a verification result; When the verification result is verification passed, extracting the frame header identifier of the decoding information; When the frame header identifier is the first identifier, determining that the signal type of the target signal to be monitored is a voice signal; When the frame header identifier is the second identifier, determining that the signal type of the target signal to be intercepted is a text signal; The first identifier is a frame header identifier corresponding to the voice signal, and the second identifier is a frame header identifier corresponding to the text signal.

4. The method according to claim 2, characterized in that The signal type includes a voice signal and a text signal, the target information includes target text information and target voice data, and the step of restoring information corresponding to the target signal to be intercepted based on the signal type to obtain the target information includes: When the signal type is the voice signal, determining a voice encoding and decoding rule corresponding to the communication protocol; Performing voice decoding on the target signal to be intercepted according to the voice coding and decoding rules to obtain decoded voice data; determining the target voice data based on the decoded voice data; When the signal type is the text signal, application layer decoding is performed according to the communication protocol to obtain the target text information.

5. The method according to claim 4, characterized in that The determining the target voice data based on the decoded voice data includes: Performing data validity identification on the decoded speech data using a speech recognition model to obtain a recognition result; When the recognition result is valid, performing noise reduction on the decoded voice data to obtain noise-reduced voice data; Speech enhancement is performed on the noise-reduced speech data to obtain the target speech data.

6. The method according to claim 1, characterized in that The method of using a second deep learning model to perform protocol identification on the target signal to be intercepted to obtain a communication protocol of the target signal to be intercepted, wherein the second deep learning model is a protocol identification model, including: Using the second deep learning model to perform time slot structure identification on the target signal to be monitored, to obtain the time slot structure of the target signal to be monitored; The second deep learning model is used to determine the communication protocol of the target signal to be monitored according to the time slot structure.

7. The method according to claim 1, characterized in that The first deep learning model includes a time-frequency feature map extraction network, a signal detection network, and a signal filtering network. The first deep learning model is used to perform signal filtering on the signal to be detected to obtain a target signal to be detected and target time-frequency feature information corresponding to the target signal to be detected, including: Inputting the signal to be monitored into the time-frequency feature map extraction network to extract the time-frequency feature map, thereby obtaining the time-frequency feature map of the signal to be monitored; Inputting the time-frequency feature graph into the signal detection network to extract time-frequency features to obtain time-frequency feature information of the signal to be monitored; The signal to be detected and the time-frequency characteristic information are input into the signal filtering network for filtering to obtain the target signal to be detected and the target time-frequency characteristic information corresponding to the target signal to be detected.

8. A signal monitoring device, characterized in that: The device comprises: A signal acquisition module is used to obtain the signal to be monitored; a signal filtering module, configured to perform signal filtering on the signal to be detected using a first deep learning model to obtain a target signal to be detected and target time-frequency feature information corresponding to the target signal to be detected, wherein the first deep learning model is a signal filtering model; a protocol identification module, configured to perform protocol identification on the target signal to be intercepted using a second deep learning model to obtain a communication protocol of the target signal to be intercepted, wherein the second deep learning model is a protocol identification model; An information restoration module, configured to restore information of the target signal to be intercepted based on the communication protocol to obtain target information corresponding to the target signal to be intercepted; The information storage module is used to store the target signal to be monitored, the target time-frequency characteristic information, the communication protocol and the target information.

9. An electronic device, characterized in that: Including processor and memory, The memory is used to store codes and related data; The processor is configured to execute the code in the memory to implement the signal monitoring method according to any one of claims 1 to 7.

10. A storage medium storing a computer program, wherein when the program is executed by a processor, the signal monitoring method according to any one of claims 1 to 7 is implemented.

Images