Electromagnetic spectrum real-time monitoring system and method based on radio frequency SoC chip

Abstract

The invention provides an electromagnetic spectrum real-time monitoring system and method based on a radio frequency SoC chip, the electromagnetic spectrum real-time monitoring system comprises the radio frequency SoC chip, a control module and upper computer software, and the radio frequency SoC chip comprises a signal acquisition and conversion module. The upper computer software comprises a time-frequency diagram generation module, a signal detection module and a signal attribute determination module which are in communication connection in sequence; the control module is respectively in communication connection with the time-frequency diagram generation module and the signal acquisition and conversion module; and the sampling rate of the radio frequency SoC chip is 5Gsps. The broadband high-sampling-rate radio frequency SoC chip is used for collecting electromagnetic spectrum signals in real time, electromagnetic signals of a wide frequency band can be monitored in real time, and the frequency range of signal collection is widened; and a short-time Fourier transform method and a YOLO deep learning network algorithm are combined to synchronously realize large-bandwidth and real-time monitoring of the electromagnetic spectrum.

Description

Technical Field

[0001] The present invention belongs to the technical field of electromagnetic spectrum monitoring, and in particular relates to a system and method for real-time monitoring of electromagnetic spectrum based on a radio frequency SoC chip. Background Art

[0002] At present, electromagnetic spectrum monitoring mainly relies on traditional narrowband monitoring equipment and methods, which have limitations in many aspects. First, the frequency band coverage is limited: the existing spectrum monitoring equipment can usually only monitor within a specific frequency range and cannot effectively capture signals within a wide frequency band. This limitation makes it impossible to fully understand the use of the electromagnetic spectrum. Second, the level of intelligence is low: signal classification and recognition in existing technologies mostly rely on prior knowledge, lack efficient automated recognition methods, and have weak adaptability to complex environments; the monitoring system has a delayed response to unknown signals. Third, the management efficiency of spectrum resources is low: the management of spectrum resources often relies on manual statistics and analysis, which cannot timely and accurately reflect the actual use of the spectrum, resulting in waste of spectrum resources and inefficient management.

[0003] In order to solve these problems, in recent years, the use of RF SoC chips (System on Chip) for signal acquisition has gradually become a trend. However, a single RF SoC chip is still unable to realize intelligent processing of complex signals. The signal monitoring capability of a single RF SoC chip and the classification capability of electromagnetic spectrum monitoring equipment need to be improved urgently to improve the efficiency and intelligence level of electromagnetic spectrum management. Chinese patent CN117743258A discloses an intelligent broadband signal detection, identification and processing device, including FPGA, digital signal processor and SoC FPGA. FPGA is used to implement a sampling data receiving module, a digital down-conversion module, and a real-time FFT processing module. The digital signal processor is used to implement a broadband signal preprocessing module, and SoC FPGA is used to implement a signal processing intelligent reasoning module. When the radio receiving device receives a broadband spectrum and performs digital sampling, the present invention detects and identifies electromagnetic signals in the broadband spectrum in real time based on the intelligent stream processing method of SoC FPGA. Although CN117743258A can achieve real-time monitoring, its sampling rate is 102.4Msps, the signal coverage range is narrow, the signal information contained is less, and the amount of data generated is also small, which cannot fully understand the usage status of the electromagnetic spectrum; and its use of Gaussian filtered time-frequency grayscale image will result in less useful information and lower image feature space dimension, which will be unfavorable for subsequent electromagnetic spectrum monitoring and identification.

[0004] Therefore, in order to break through the limitation of "limited frequency band coverage", it is necessary to adopt high sampling rate RF SoC chips. Although high sampling rate RF SoC chips bring large bandwidth, they also bring high data volume. It takes a long time to do a lot of complex processing under high data volume, which affects real-time performance, which is not conducive to users making timely decisions. How to achieve real-time monitoring of the electromagnetic spectrum based on large bandwidth and high data volume has become a technical problem that must be solved. In view of this, the development of a real-time monitoring system for the electromagnetic spectrum has become a problem that needs to be solved urgently by those skilled in the art. Summary of the invention

[0005] The purpose of the present invention is to solve the problems existing in the above-mentioned prior art, and provides an electromagnetic spectrum real-time monitoring system and method based on RF SoC chip. The electromagnetic spectrum real-time monitoring system utilizes broadband and high-sampling rate RF SoC chip to perform real-time acquisition of electromagnetic spectrum signals, realize real-time monitoring of wide-band signals, and detect signals in combination with deep learning algorithms. At the same time, the signals are feature matched, intelligently classified and identified through host computer software and prior signal feature library; in addition, the early warning module is used to realize the early warning of unknown signals, and the statistical module is used to realize the statistical analysis of the monitoring results.

[0006] One of the inventive objectives of the present invention is to provide an electromagnetic spectrum real-time monitoring system based on a radio frequency SoC chip, comprising:

[0007] A radio frequency SoC chip, comprising a signal acquisition and conversion module and a control module connected in communication, wherein the signal acquisition and conversion module is used to acquire electromagnetic signals and convert the electromagnetic signals into intermediate frequency digital signals; the sampling rate of the radio frequency SoC chip is 5Gsps; the control module is used to configure the parameters of the signal acquisition and conversion module, receive the intermediate frequency digital signal sent by the signal acquisition and conversion module, and send it to the host computer software;

[0008] The upper computer software includes a time-frequency diagram generation module, a signal detection module, and a signal attribute determination module which are communicatively connected in sequence; the time-frequency diagram generation module is communicatively connected to the control module, and is used to receive the intermediate frequency digital signal sent by the control module and convert the intermediate frequency digital signal into a time-frequency diagram; the signal detection module is used to receive the time-frequency diagram sent by the time-frequency diagram generation module and obtain the parameters of the electromagnetic signal based on a deep learning network algorithm; the signal attribute determination module is used to receive the parameters of the electromagnetic signal sent by the signal detection module and call a priori signal feature library to obtain the attributes of the electromagnetic signal.

[0009] In a preferred embodiment of the present invention, the RF SoC chip also includes multiple filtering modules, each of which is respectively communicatively connected to the signal acquisition and conversion module and the control module; each of the filtering modules is used to receive the intermediate frequency digital signal sent by the signal acquisition and conversion module and send the intermediate frequency digital signal of one of the preset key monitoring frequency bands to the control module.

[0010] In a preferred implementation manner of the present invention, the number of the filtering modules is the same as the number of the preset key monitoring frequency bands.

[0011] In a preferred implementation of the present invention, in the time-frequency diagram generation module, a short-time Fourier transform method is used to convert the intermediate frequency digital signal into a time-frequency diagram.

[0012] In a preferred embodiment of the present invention, in the signal detection module, the parameters of the electromagnetic signal are obtained based on and using the YOLO deep learning network algorithm time-frequency diagram;

[0013] Preferably, the parameters of the electromagnetic signal include but are not limited to the bandwidth, duration, carrier frequency and signal modulation method of the electromagnetic signal.

[0014] In a preferred embodiment of the present invention, the attributes of the electromagnetic signal include but are not limited to category attributes and inherent attributes; the category attributes include but are not limited to unknown signals, known signals, threatening signals, and non-threatening signals; the inherent attributes include but are not limited to bandwidth, duration, carrier frequency, and signal modulation method.

[0015] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system also includes an early warning module, which is connected to the signal attribute determination module; the early warning module is used to receive the attributes of the electromagnetic signal sent by the signal attribute determination module and generate an early warning message based on the attributes of the electromagnetic signal.

[0016] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system further comprises a statistical module, which is connected to the signal detection module; the statistical module is used to receive the parameters of the electromagnetic signal sent by the signal detection module, and obtain the occupancy status based on the parameters of the electromagnetic signal;

[0017] Preferably, the occupancy status includes but is not limited to frequency occupancy rate and channel occupancy rate.

[0018] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system also includes a storage module, which is connected to the RF SoC chip, the time-frequency graph generation module, the signal detection module, and the signal attribute determination module, and is used to receive and store the intermediate frequency digital signal sent by the RF SoC chip, receive and store the time-frequency graph sent by the time-frequency graph generation module, receive and store the parameters of the electromagnetic signal sent by the signal detection module, and receive and store the attributes of the electromagnetic signal sent by the signal attribute determination module.

[0019] The second object of the present invention is to provide a method for real-time monitoring of electromagnetic spectrum based on radio frequency SoC chip, which is performed by using the real-time monitoring system for electromagnetic spectrum described in one of the objects of the present invention, and comprises the following steps:

[0020] S1: collects electromagnetic signals and converts them into intermediate frequency digital signals;

[0021] S2: Convert the intermediate frequency digital signal into a time-frequency diagram;

[0022] S3: Based on the time-frequency diagram and deep learning network algorithm, the parameters of the electromagnetic signal are obtained;

[0023] S4: obtaining the properties of the electromagnetic signal based on the parameters of the electromagnetic signal and the prior signal feature library;

[0024] Preferably, the electromagnetic spectrum real-time monitoring method based on the radio frequency SoC chip further includes the following steps:

[0025] generating a warning message based on the properties of the electromagnetic signal; and / or,

[0026] Obtaining occupancy status based on parameters of the electromagnetic signal; and / or,

[0027] Receive and store the intermediate frequency digital signal sent by the RF SoC chip, receive and store the time-frequency diagram sent by the time-frequency diagram generation module, receive and store the parameters of the electromagnetic signal sent by the signal detection module, and receive and store the properties of the electromagnetic signal sent by the signal property determination module.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The electromagnetic spectrum real-time monitoring system based on the radio frequency SoC chip of the present invention utilizes a broadband high sampling rate radio frequency SoC chip to perform real-time acquisition of electromagnetic spectrum signals, and can monitor electromagnetic signals in a wide frequency band in real time, covering a range from several kilohertz to several gigahertz, thereby improving the frequency range of signal acquisition. The short-time Fourier transform method is used in the time-frequency graph generation module of the present invention to convert the intermediate frequency digital signal into a time-frequency graph, which is convenient for detection and reduces the amount of data; at the same time, the signal detection module of the present invention uses the YOLO (You Only Look Once) deep learning network algorithm for signal detection, thereby achieving the purpose of real-time monitoring to ensure effective monitoring of electromagnetic signals in a wide frequency band, and simultaneously realizing the large bandwidth and real-time monitoring of the electromagnetic spectrum.

[0030] 2. The real-time electromagnetic spectrum monitoring system based on the RF SoC chip of the present invention combines the RF SoC chip and the deep learning algorithm to realize the intelligent classification and recognition of signals, and the monitoring of electromagnetic signals is more accurate (especially the automatic monitoring and early warning of unknown signals), thereby reducing the false alarm rate.

[0031] 3. The electromagnetic spectrum real-time monitoring system based on the radio frequency SoC chip of the present invention combines the host computer software with the prior signal feature library to achieve in-depth analysis of signals and intelligent early warning of unknown signals, thereby improving the intelligence level of the system.

[0032] 4. The electromagnetic spectrum real-time monitoring system based on the radio frequency SoC chip of the present invention provides a friendly host computer software interface, which is convenient for users to perform data visualization, analysis and report generation.

[0033] 5. The electromagnetic spectrum real-time monitoring system based on the radio frequency SoC chip of the present invention can provide a scientific basis for the management of spectrum resources through frequency occupancy statistics, support optimal configuration and scheduling, and promote the rational use of resources.

[0034] 6. The electromagnetic spectrum real-time monitoring system based on the radio frequency SoC chip of the present invention can save monitoring data, provide a basis for subsequent analysis, and help understand the long-term trend of spectrum usage. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of a real-time monitoring system of electromagnetic spectrum based on radio frequency SoC chip;

[0036] Figure 2 An exemplary time-frequency diagram generated by a real-time monitoring system of the electromagnetic spectrum of a radio frequency SoC chip;

[0037] Figure 3 A schematic diagram of an exemplary electromagnetic signal detection result generated by a real-time monitoring system of an electromagnetic spectrum of a radio frequency SoC chip;

[0038] Figure 4 Schematic diagram of the real-time monitoring method of the electromagnetic spectrum based on the RF SoC chip. DETAILED DESCRIPTION

[0039] In the present invention, FPGA refers to Field-Programmable Gate Array, which is the full name of Field-Programmable Gate Array.

[0040] In the present invention, ADC refers to an analog / digital converter, and its full name is Analog-to-Digital Converter.

[0041] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0042] like Figure 1 As shown, the present invention provides an electromagnetic spectrum real-time monitoring system based on a radio frequency SoC chip, including a radio frequency SoC chip and a host computer software. The radio frequency SoC chip and the host computer software are described in detail below.

[0043] The radio frequency SoC chip includes a signal acquisition and conversion module and a control module connected in communication; the signal acquisition and conversion module is used to collect electromagnetic signals and convert the electromagnetic signals into intermediate frequency digital signals. The signal acquisition and conversion module of the radio frequency SoC chip has a high sampling rate of 10MHz to 10GHz, and can collect electromagnetic signals in a wide frequency band in real time, covering a wide spectrum from low frequency (several thousand hertz) to high frequency (several gigahertz). Specifically, the signal acquisition and conversion module of the radio frequency SoC chip is responsible for receiving and processing the radio frequency signal from the antenna, first converting the radio frequency signal into an intermediate frequency signal, and then converting the intermediate frequency signal into an intermediate frequency digital signal. Among them, "converting the radio frequency signal into an intermediate frequency signal" belongs to the conventional technical means in this field, and will not be repeated here; "converting the intermediate frequency signal into an intermediate frequency digital signal" is because the radio frequency SoC chip has a built-in ADC to ensure the accuracy of the collected signal. The ADC quantizes and encodes the intermediate frequency signal, and finally obtains the intermediate frequency digital represented by the digital representation, that is, the intermediate frequency digital signal. In addition, the radio frequency SoC chip can set the intermediate frequency by configuring the register; exemplary, the chip model is Zynq UltraScale+RFSoC ZU47DR.

[0044] In the present invention, the acquisition and conversion processing of electromagnetic signals are directly completed on the RF SoC chip without the need for additional data transmission, which can reduce the delay in data processing and improve the processing speed; in addition, since the acquisition and conversion processing of electromagnetic signals are completed on the same RF SoC chip, compared with the discrete device solution, this integrated solution reduces the need for data transmission between different chips or modules, reduces the data transmission communication process, and thus reduces delay and power consumption.

[0045] In a preferred embodiment of the present invention, the RF SoC chip also includes a plurality of filtering modules, each of which is respectively connected to the signal acquisition and conversion module and the control module in communication; each of which is used to receive the intermediate frequency digital signal sent by the signal acquisition and conversion module and send the intermediate frequency digital signal of one of the preset key monitoring frequency bands to the control module. Specifically in this embodiment, the RF SoC chip has a built-in FPGA, and the FPGA can perform parallel processing of multiple data channels. The RF SoC chip includes a plurality of digital filters (i.e., filtering modules) set by FPGA, which are used to filter the collected signals respectively. After filtering, only one of the signals of the preset key monitoring frequency bands is left in the frequency domain, and the signals of other frequency bands are attenuated, and then processed and analyzed. Preferably, the number of the filtering modules is the same as the number of the key monitoring frequency bands. Since each digital filter allows the signal of one of the preset key monitoring frequency bands to pass, after filtering by multiple digital filters, the intermediate frequency digital signals of multiple preset key monitoring frequency bands are obtained, which are then processed and analyzed. Exemplarily, there are five preset key monitoring frequency bands, namely the first key monitoring frequency band, the second key monitoring frequency band, the third key monitoring frequency band, the fourth key monitoring frequency band, and the fifth key monitoring frequency band. Then five filters need to be configured. The five digital filters respectively allow the intermediate frequency digital signals of the first key monitoring frequency band, the second key monitoring frequency band, the third key monitoring frequency band, the fourth key monitoring frequency band, and the fifth key monitoring frequency band to pass through, and finally obtain the intermediate frequency digital signals corresponding to the five key monitoring frequency bands.

[0046] The control module is connected to the signal acquisition and conversion module for configuring the parameters of the signal acquisition and conversion module, receiving the intermediate frequency digital signal sent by the signal acquisition and conversion module, and sending it to the host computer software. When the RF SoC chip also includes multiple filter modules, in this embodiment, the control module is also used to configure the parameters of the multiple filter modules, receive the multiple intermediate frequency digital signals sent by the multiple filter modules, and send them to the host computer software.

[0047] The host computer software is connected to the control module for receiving the intermediate frequency digital signal sent by the control module, converting the intermediate frequency digital signal into a time-frequency diagram, obtaining the parameters of the electromagnetic signal based on the time-frequency diagram, and obtaining the properties of the electromagnetic signal based on the parameters of the electromagnetic signal and a priori signal feature library. Preferably, the host computer software includes a time-frequency diagram generation module, a signal detection module, and a signal property determination module.

[0048] The time-frequency diagram generation module is communicatively connected with the control module, and is used to receive the intermediate frequency digital signal sent by the control module and convert the intermediate frequency digital signal into a time-frequency diagram. The time-frequency diagram can clearly reflect the spectral characteristics of the signal, which is helpful for the deep learning algorithm to perform subsequent analysis. Preferably, in the time-frequency diagram generation module, the short-time Fourier transform (STFT) method is used to convert the intermediate frequency digital signal into a time-frequency diagram, showing the changes of the intermediate frequency digital signal in the time and frequency domain. STFT (Short-time Fourier transform) is a method of localizing a signal in time and frequency, and is commonly used to analyze non-stationary signals. STFT performs Fourier transform on each segment of the original signal by segmenting the original signal, thereby obtaining the frequency components of the signal at different times. This localized analysis can show the frequency characteristics of the signal changing over time.

[0049] Exemplarily, the process of STFT processing the intermediate frequency digital signal is as follows:

[0050] Window segmentation: First, select a suitable window function (such as Hamming window, Hanning window, etc.) and window length to cut the long-term intermediate frequency digital signal into multiple small time periods. Window length and overlap rate (i.e., the overlapping part of adjacent windows) are important parameters of STFT, which determine the balance between time and frequency resolution.

[0051] Fourier transform: Perform Fourier transform on the intermediate frequency digital signal in each window to obtain the frequency component at that moment. In this way, the spectrum of the signal at different moments can be obtained.

[0052] Result superposition: The results of each Fourier transform are stitched together to form a two-dimensional time-frequency diagram. The value of each point in the diagram represents the signal amplitude at that time and frequency.

[0053] The results of STFT are usually presented as a time-frequency diagram, where the horizontal axis is time and the vertical axis is frequency. The color or brightness represents the power of the frequency component (e.g. Figure 2 The changes of different frequency components over time will appear as different color blocks or textures in the figure, reflecting the dynamic spectrum characteristics of the signal.

[0054] The short-time Fourier transform method is used in the time-frequency diagram generation module of the present invention to convert the intermediate frequency digital signal into a time-frequency diagram, so that the characteristics of the intermediate frequency digital signal are more obvious and easy to detect, while reducing the amount of data. STFT decomposes the signal into multiple short time segments, and these segments can be processed in parallel, thereby reducing the computational complexity and improving the processing speed. In the time-frequency diagram, the energy distribution of the intermediate frequency digital signal is more concentrated, especially in the case of a known frequency band, only the area where the energy is concentrated in the intermediate frequency digital signal is concerned, thereby reducing the amount of data to be processed.

[0055] The signal detection module is connected to the time-frequency graph generation module, and is used to receive the time-frequency graph sent by the time-frequency graph generation module and obtain the parameters of the electromagnetic signal based on the deep learning network algorithm. Preferably, the parameters of the electromagnetic signal include but are not limited to the bandwidth, duration, carrier frequency and signal modulation method of the electromagnetic signal. More preferably, the signal detection module uses the YOLO (You Only Look Once) deep learning network algorithm for signal detection. The YOLO deep learning network algorithm has the ability of fast real-time target detection. Exemplarily, the process of YOLO receiving the time-frequency graph and obtaining the parameters of the electromagnetic signal is as follows:

[0056] A: Model input.

[0057] The received frequency spectrum is input into the YOLO model. YOLO will process the frequency spectrum as a normal two-dimensional image without modifying the input layer. YOLO will capture the different frequency and time variation patterns in the frequency spectrum through the feature extraction layer, and map these patterns to different feature layers to identify the key parts of the signal.

[0058] B: Detection process.

[0059] Grid division: YOLO divides the time-frequency map into several grids, each of which is responsible for detecting certain target areas.

[0060] Prediction box and confidence: Each grid predicts multiple bounding boxes and calculates the confidence of each box. For time-frequency graphs, the prediction box can correspond to a specific time and frequency interval, as well as the strength of the signal in the area.

[0061] Class Probability: YOLO also predicts the probability value of each bounding box belonging to each class of signal.

[0062] C: Post-processing.

[0063] Non-maximum suppression (NMS): After detection is completed, YOLO will use the non-maximum suppression method to remove overlapping boxes and only retain bounding boxes with high confidence to ensure that the recognition results are not repeated.

[0064] D: Parameters of output electromagnetic signal.

[0065] Detection box: YOLO will frame the identified signal events in the time-frequency graph. Each detection box corresponds to a time-frequency region, indicating that a specific signal pattern exists at that time and frequency range.

[0066] Category label: Each detection box is accompanied by a category label, indicating the type of signal detected. For example, in the interference signal detection scenario, different types of interference signals can be identified; in the modulation signal analysis, different modulation modes can be distinguished.

[0067] like Figure 3 As shown, the final output detection result includes information such as the bounding box, category, and confidence of each signal target. Among them, the length of the bounding box on the time axis represents the duration of the intermediate frequency digital signal, which is one of the output results of the signal detection module; the width of the bounding box on the frequency axis represents the range of occupied frequencies, also known as bandwidth, which is the second output result of the signal detection module; the midpoint of the edge of the bounding box in the vertical direction is the carrier frequency, which is the third output result of the signal detection module. Combining the category and confidence, the category corresponding to the maximum value in the confidence vector is the predicted category of the electromagnetic signal (i.e., the signal modulation method), which is the fourth output result of the signal detection module. It should be noted that, first, the network structure of YOLO can refer to the journal entitled "YOLOv9: Learning What You Want to Learn UsingProgrammable Gradient Information" (published in 2024), which will not be repeated in the present invention. Second, the signal detection module ultimately obtains the properties of the signal itself. Third, in order to increase the robustness and generalization of the YOLO network, a variety of color maps are used to map the time-frequency matrix into a time-frequency graph during network training.

[0068] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system also includes a priori signal feature library. The signal attribute determination module is used to receive the parameters of the electromagnetic signal sent by the signal detection module and call the priori signal feature library to obtain the attributes of the electromagnetic signal. The "attribute of the electromagnetic signal" here is used to determine whether it is an unknown signal and whether the signal is threatening. Among them, the priori signal feature library contains the current business division of the frequency band, which stores the known communication, radar, navigation signal patterns and signal frequency bands, which is used to train the deep learning network algorithm to identify the attributes of known signals and to judge the signals. It should be noted that in the priori signal feature library, signals with different "bandwidth, duration, carrier frequency and signal modulation mode" are located in different attribute partitions. Exemplarily, different frequency bands are divided into different services, and the signal modulation methods on different services will also be different. FM frequency modulation broadcasting is mainly concentrated in 88MHz~108MHz, and aviation frequencies are mainly concentrated in 118MHz~136MHz. Therefore, when signals other than AM amplitude modulation signals appear in the aviation frequency band, it is abnormal, and these signals are threatening signals here. The prior signal feature library helps to more accurately monitor the frequency band. The signal attribute determination module matches the parameters of the electromagnetic signal with the prior signal feature library, and can determine whether the electromagnetic signal should appear in the frequency band based on the bandwidth, duration, carrier frequency and signal modulation method sent by the prior signal feature library and the signal detection module, thereby determining the attributes of the electromagnetic signal.

[0069] The properties of electromagnetic signals include but are not limited to category attributes and inherent attributes; category attributes include but are not limited to unknown signals, known signals, threatening signals, and non-threatening signals; inherent attributes include but are not limited to bandwidth, duration, carrier frequency, and signal modulation method.

[0070] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system also includes a statistical module, which is connected to the signal detection module, and is used to receive the parameters of the electromagnetic signal sent by the signal detection module, and obtain the occupancy status based on the parameters of the electromagnetic signal, and can also output an occupancy report. The occupancy report can help users understand the signal distribution. The occupancy status includes but is not limited to frequency occupancy and channel occupancy. Among them, the frequency occupancy rate refers to the proportion of the frequency bandwidth occupied by the signal within the entire spectrum range, and the channel occupancy rate refers to the proportion of the channel occupied by the signal in the entire time period. Exemplarily, if the signal only occupies a part of the entire spectrum range, then the frequency occupancy can be expressed by the ratio of the occupied bandwidth to the total bandwidth. Exemplarily, within a 2-second observation time, if the signal only sends data within 1 second, the time domain occupancy of the channel is 50%.

[0071] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system further includes an early warning module, which is connected to the signal attribute determination module and is used to receive the attributes of the electromagnetic signal sent by the signal attribute determination module and generate an early warning message based on the attributes of the electromagnetic signal. When a signal that cannot be matched with the prior signal feature library is detected, the attribute of the electromagnetic signal is "unknown signal" and / or "threat signal", and the early warning module automatically generates an early warning to indicate the possible occurrence of unknown or abnormal signals.

[0072] In a preferred embodiment of the present invention, the electromagnetic spectrum real-time monitoring system also includes a storage module, which is connected to the RF SoC chip, the time-frequency graph generation module, the signal detection module, and the signal attribute determination module, and is used to receive the intermediate frequency digital signal sent by the RF SoC chip, receive the time-frequency graph sent by the time-frequency graph generation module, receive the parameters of the electromagnetic signal sent by the signal detection module, receive the attributes of the electromagnetic signal sent by the signal attribute determination module, and save the intermediate frequency digital signal, time-frequency graph, electromagnetic signal parameters, and electromagnetic signal attributes as historical monitoring data to facilitate subsequent analysis and backtracking. Users can view the spectrum usage in a certain period of time as needed.

[0073] In the present invention, the electromagnetic spectrum real-time monitoring system includes both real-time monitoring of the full frequency band and real-time detection of the preset key monitoring frequency band. For the real-time monitoring of the full frequency band, since the signal acquisition and conversion module and the control module are connected in communication, the control module can receive the intermediate frequency digital signal of the full frequency band, and then send it to the host computer software. The host computer software converts the intermediate frequency digital signal of the full frequency band into a time-frequency diagram, obtains the parameters of the electromagnetic signal, and obtains the properties of the electromagnetic signal, so as to realize the real-time monitoring of the electromagnetic spectrum of the full frequency band. Secondly, the statistical module can also count the occupancy of the full frequency band. Finally, because there may be short-term burst signals in the channel, and their frequency and start and end times are uncertain, the real-time monitoring of the full frequency band can detect such burst signals with a higher probability, and the early warning module can report these signal information to the monitoring personnel. For real-time detection of preset key monitoring frequency bands, since each of the filter modules is connected to the control module in communication, the control module can receive the intermediate frequency digital signals of multiple preset key monitoring frequency bands, and then send them to the host computer software respectively. The host computer software converts them into time-frequency diagrams, obtains the parameters of electromagnetic signals, and obtains the properties of electromagnetic signals, thereby realizing real-time monitoring of the electromagnetic spectrum of the preset key monitoring frequency bands. Since multiple preset key monitoring frequency bands are processed in parallel and the preset key monitoring frequency bands are much narrower than the full frequency band, the amount of data is much smaller, so the accuracy when converted into time-frequency diagrams will be higher, and the accuracy of obtaining the properties of electromagnetic signals will also be higher.

[0074] like Figure 4 As shown, the present invention also provides an electromagnetic spectrum monitoring method based on a radio frequency SoC chip, comprising the following steps:

[0075] S1: Collect electromagnetic signals and convert them into intermediate frequency digital signals.

[0076] The RF SoC chip samples the electromagnetic signals in the target frequency band in real time and converts them into intermediate frequency digital signals, and transmits the data to the control module in digital form. The I / Q data (intermediate frequency digital signals) collected by the RF SoC chip are organized into arrays and transmitted through optical fibers.

[0077] Preferably, step S1 includes collecting electromagnetic signals of the full frequency band and electromagnetic signals of multiple preset key monitoring frequency bands, and converting them into intermediate frequency digital signals respectively;

[0078] S2: Convert the intermediate frequency digital signal into a time-frequency diagram.

[0079] The intermediate frequency digital signal is transformed into a time-frequency diagram through short-time Fourier transform (STFT) to show the time-frequency characteristics of the signal.

[0080] Preferably, step S2 includes converting the intermediate frequency digital signals of the entire frequency band and the intermediate frequency digital signals of multiple preset key monitoring frequency bands into time-frequency diagrams respectively;

[0081] S3: Based on the time-frequency diagram and deep learning network algorithm, the parameters of the electromagnetic signal are obtained.

[0082] The time-frequency graph is input into the YOLO deep learning network algorithm for detection and classification. The YOLO deep learning network algorithm obtains the parameters of the electromagnetic signal in real time through its convolutional neural network structure.

[0083] Preferably, step S3 includes detecting the time-frequency diagram of the full frequency band and the time-frequency diagram of multiple preset key monitoring frequency bands respectively to obtain the parameters of the electromagnetic signal of the full frequency band and the parameters of the electromagnetic signal of multiple preset key monitoring frequency bands;

[0084] S4: Based on the parameters of the electromagnetic signal and a priori signal feature library, the properties of the electromagnetic signal are obtained.

[0085] The host computer software calls the prior signal feature library, compares the parameters of the electromagnetic signal obtained by the YOLO deep learning network algorithm with the prior signal feature library, and obtains the properties of the electromagnetic signal.

[0086] Preferably, step S4 includes comparing the parameters of the electromagnetic signal of the entire frequency band with the prior signal feature library to obtain the properties of the electromagnetic signal of the entire frequency band, and comparing the parameters of the electromagnetic signal of multiple preset key monitoring frequency bands with the prior signal feature library to obtain the properties of the electromagnetic signal of multiple preset key monitoring frequency bands.

[0087] In a preferred embodiment of the present invention, the electromagnetic spectrum monitoring method based on the radio frequency SoC chip comprises the following steps:

[0088] Generate warning messages based on the properties of electromagnetic signals. When a signal that cannot be matched with the prior signal feature library is detected, the properties of the electromagnetic signal are "unknown signal" and / or "threat signal". The warning module automatically generates a warning to prompt the possible unknown or abnormal signal and provide it to the user for decision-making; and / or,

[0089] Obtaining occupancy information based on the parameters of the electromagnetic signal and providing it to the user for decision making; and / or,

[0090] Receive and store the intermediate frequency digital signal sent by the RF SoC chip, receive and store the time-frequency diagram sent by the time-frequency diagram generation module, receive and store the parameters of the electromagnetic signal sent by the signal detection module, and receive and store the properties of the electromagnetic signal sent by the signal property determination module.

[0091] Example 1: Radio monitoring application

[0092] In a certain radio monitoring system, the aforementioned "real-time monitoring system for electromagnetic spectrum based on RF SoC chip" is used for real-time monitoring of electromagnetic spectrum. The RF SoC chip collects electromagnetic signals, converts them into intermediate frequency digital signals, and then converts them into intermediate frequency digital signals; the host computer software receives the intermediate frequency digital signals, generates time-frequency graphs, and then inputs them into the YOLO network, and combines the prior signal feature library to perform signal classification and frequency occupancy statistics. Through this system, unknown signals can be discovered and alarmed in a timely manner, and the monitoring results significantly improve the efficiency of spectrum resource management.

[0093] Example 2: Military spectrum monitoring

[0094] In the military field, the aforementioned "real-time monitoring system of electromagnetic spectrum based on RF SoC chip" is used for real-time monitoring of electromagnetic spectrum. The RF SoC chip collects electromagnetic signals, converts them into intermediate frequency digital signals, and then converts them into intermediate frequency digital signals (including electromagnetic signals of the full frequency band and electromagnetic signals of the preset key monitoring frequency band). The system analyzes and classifies signals in real time, and uses the upper computer software to collect statistics and warn the monitoring results, providing decision support for the command center. The system effectively improves the ability to control the spectrum and enhances the security of military communications.

[0095] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "connected" and "connection" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0096] In the description of the present invention, unless otherwise specified, the terms "upper", "lower", "left", "right", "inside", "outside", etc. indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the present invention.

[0097] The above technical solution is only one implementation mode of the present invention. For those skilled in the art, it is easy to make various types of improvements or modifications based on the principles disclosed in the present invention, and it is not limited to the technical solution described in the above specific embodiments of the present invention. Therefore, the above description is only preferred and does not have a restrictive meaning.

Claims

1. A real-time monitoring system for electromagnetic spectrum based on radio frequency SoC chip, characterized in that: include: A radio frequency SoC chip, comprising a signal acquisition and conversion module and a control module connected in communication, wherein the signal acquisition and conversion module is used to acquire electromagnetic signals and convert the electromagnetic signals into intermediate frequency digital signals; the sampling rate of the radio frequency SoC chip is 5Gsps; the control module is used to configure the parameters of the signal acquisition and conversion module, receive the intermediate frequency digital signal sent by the signal acquisition and conversion module, and send it to the host computer software; The upper computer software includes a time-frequency diagram generation module, a signal detection module, and a signal attribute determination module which are communicatively connected in sequence; the time-frequency diagram generation module is communicatively connected to the control module, and is used to receive the intermediate frequency digital signal sent by the control module and convert the intermediate frequency digital signal into a time-frequency diagram; the signal detection module is used to receive the time-frequency diagram sent by the time-frequency diagram generation module and obtain the parameters of the electromagnetic signal based on a deep learning network algorithm; the signal attribute determination module is used to receive the parameters of the electromagnetic signal sent by the signal detection module and call a priori signal feature library to obtain the attributes of the electromagnetic signal.

2. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: The RF SoC chip also includes multiple filtering modules, each of which is respectively communicated with the signal acquisition and conversion module and the control module; each of the filtering modules is used to receive the intermediate frequency digital signal sent by the signal acquisition and conversion module and send the intermediate frequency digital signal of one of the preset key monitoring frequency bands to the control module.

3. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 2 is characterized in that: The number of the filtering modules is the same as the number of the preset key monitoring frequency bands.

4. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: In the time-frequency diagram generation module, the short-time Fourier transform method is used to convert the intermediate frequency digital signal into a time-frequency diagram.

5. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: In the signal detection module, the parameters of the electromagnetic signal are obtained based on the YOLO deep learning network algorithm time-frequency diagram; Preferably, the parameters of the electromagnetic signal include but are not limited to the bandwidth, duration, carrier frequency and signal modulation method of the electromagnetic signal.

6. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: The properties of the electromagnetic signal include but are not limited to category attributes and inherent attributes; the category attributes include but are not limited to unknown signals, known signals, threatening signals, and non-threatening signals; the inherent attributes include but are not limited to bandwidth, duration, carrier frequency, and signal modulation method.

7. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: The electromagnetic spectrum real-time monitoring system also includes an early warning module, which is connected to the signal attribute determination module; the early warning module is used to receive the attributes of the electromagnetic signal sent by the signal attribute determination module and generate an early warning message based on the attributes of the electromagnetic signal.

8. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: The electromagnetic spectrum real-time monitoring system further includes a statistical module, which is connected to the signal detection module; the statistical module is used to receive the parameters of the electromagnetic signal sent by the signal detection module, and obtain the occupancy status based on the parameters of the electromagnetic signal; Preferably, the occupancy status includes but is not limited to frequency occupancy rate and channel occupancy rate.

9. The electromagnetic spectrum real-time monitoring system based on radio frequency SoC chip according to claim 1 is characterized in that: The electromagnetic spectrum real-time monitoring system also includes a storage module, which is connected to the RF SoC chip, the time-frequency graph generation module, the signal detection module, and the signal attribute determination module, and is used to receive and store the intermediate frequency digital signal sent by the RF SoC chip, receive and store the time-frequency graph sent by the time-frequency graph generation module, receive and store the parameters of the electromagnetic signal sent by the signal detection module, and receive and store the attributes of the electromagnetic signal sent by the signal attribute determination module.

10. A method for real-time monitoring of electromagnetic spectrum based on radio frequency SoC chip, using the real-time monitoring system of electromagnetic spectrum according to any one of claims 1 to 9, characterized in that: The steps include: S1: collects electromagnetic signals and converts them into intermediate frequency digital signals; S2: Convert the intermediate frequency digital signal into a time-frequency diagram; S3: Based on the time-frequency diagram and deep learning network algorithm, the parameters of the electromagnetic signal are obtained; S4: obtaining the properties of the electromagnetic signal based on the parameters of the electromagnetic signal and the prior signal feature library; Preferably, the electromagnetic spectrum real-time monitoring method based on the radio frequency SoC chip further includes the following steps: generating warning messages based on the properties of electromagnetic signals; and / or, Obtaining occupancy status based on parameters of the electromagnetic signal; and / or, Receive and store the intermediate frequency digital signal sent by the RF SoC chip, receive and store the time-frequency diagram sent by the time-frequency diagram generation module, receive and store the parameters of the electromagnetic signal sent by the signal detection module, and receive and store the properties of the electromagnetic signal sent by the signal property determination module.

Images