A software-defined intelligent forwarding jamming system

Abstract

The present application relates to a software-defined intelligent forwarding jamming system. By designing a communication unit, a forwarding jamming unit, an jamming effectiveness evaluation unit and a multi-layer neural network unit, a multi-layer neural network method is used to construct a functional relationship between time-frequency characteristics and a sampling window and replay jamming intensity based on a known signal sent by the communication unit. This enables the system to select the optimal sampling window size and replay jamming intensity to implement jamming when jamming an unknown signal in an actual environment, effectively reducing the storage space for electromagnetic signal acquisition, lowering the power consumption of the forwarding jamming system and saving preparation costs.

Description

Technical Field

[0001] The present application relates to the field of electronic countermeasure technology, and in particular to a software-defined intelligent forwarding jamming system. Background Art

[0002] In the context of electromagnetic warfare between opposing sides, jamming is common and necessary. With the development of electronic countermeasures, numerous jamming techniques have emerged. Among them, forwarding jamming has garnered significant attention in recent years. This method utilizes all or part of a local transmitted signal, forwarding it directly or after appropriate processing, to achieve the desired jamming effect. Forwarding jamming offers the advantages of automatic frequency overlap, high targeting accuracy, and excellent real-time performance. Spectrum sensing and data acquisition are essential prerequisites for implementing forwarding jamming. However, system storage resources are often limited, and collecting large amounts of data for subsequent processing is a challenge, especially when the time-frequency characteristics of the communication are unknown.

[0003] However, existing forwarding jamming systems, when the time-frequency characteristics of environmental communications are unknown, often simply apply maximum power to the replayed jamming signal without fully considering the size of the sampling window. This results in the power supply capacity being too large and the size being correspondingly increased when making forwarding jamming systems. This leads to high costs and difficulty in portability, making it difficult to adapt to field operations. This greatly limits the scope of application of forwarding jamming. Summary of the Invention

[0004] Based on this, it is necessary to provide a software-defined intelligent forwarding jamming system to address the above technical issues and optimize the cost and power consumption of forwarding jamming.

[0005] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0006] On the one hand, an embodiment of the present invention provides a software-defined intelligent forwarding jamming system, comprising a communication unit, a forwarding jamming unit, a jamming effectiveness evaluation unit, and a multi-layer neural network unit;

[0007] The communication unit is used to transmit data messages;

[0008] The forwarding interference unit includes a spectrum situation awareness module, a replay interference module and a storage module. The spectrum situation awareness module is used to receive and parse data messages, obtain the current time-frequency characteristics of the communication unit, and sample the data messages by setting a sampling window to obtain IQ signal data. The replay interference module is used to set the replay interference intensity and transmit the interference signal. The storage module is used to store the current time-frequency characteristics and IQ signal data.

[0009] The interference effectiveness evaluation unit is used to evaluate the interference signal and send the evaluation result to the multi-layer neural network unit;

[0010] The multi-layer neural network unit is used to select sample parameters based on the evaluation results to construct a sample parameter set. The sample parameters include time-frequency characteristics, sampling window, and replay interference intensity. Based on the sample parameter set, the multi-layer neural network performs supervised training to obtain a relationship function between the time-frequency characteristics, sampling window, and replay interference intensity. In an actual environment, the current sampling window and replay interference intensity are calculated based on the relationship function and the current time-frequency characteristics, and sent to the forwarding interference unit. The forwarding interference unit sends an actual interference signal based on the current sampling window and replay interference intensity.

[0011] The communication unit, forwarding interference unit, interference effectiveness evaluation unit and multi-layer neural network unit are all software-defined.

[0012] In one embodiment, the working mode of the communication unit includes a regular communication mode and an irregular communication mode. The data message sending mode in the regular communication mode is periodic sending, and the data message sending mode in the irregular communication mode is random sending.

[0013] In one embodiment, the communication unit includes a source coding module, a channel coding module and a modulation and demodulation module. The source coding module is communicatively connected to the channel coding module, and the channel coding module is communicatively connected to the modulation and demodulation module.

[0014] In one embodiment, the spectrum situation awareness module includes a time-frequency feature calculation component and an IQ signal data sampling component;

[0015] The time-frequency feature calculation component is used to receive the electromagnetic spectrum of the data message and calculate and measure the time-frequency features of the communication unit in real time based on the electromagnetic spectrum;

[0016] The IQ signal data sampling component is used to sample the electromagnetic spectrum according to a set sampling window to obtain IQ signal data.

[0017] In one embodiment, the replay interference module includes a filtering component, a frequency conversion component, a power amplifier circuit, and a directional antenna;

[0018] The filter component is used to filter and purify the IQ signal to obtain noise reduction interference sources;

[0019] The frequency conversion component is used to perform up-conversion processing on the noise reduction interference source;

[0020] The power amplifier circuit is used to set the playback interference intensity to amplify the power of the noise reduction interference source after up-conversion;

[0021] The directional antenna is used to transmit the noise reduction interference source after power amplification.

[0022] In one embodiment, the interference effectiveness evaluation unit includes an evaluation module and a feedback module;

[0023] The evaluation module evaluates interference signals through communication failure rate, storage resource consumption space and system power consumption;

[0024] The feedback module sends the communication failure rate, storage resource consumption space and system power consumption to the multi-layer neural network unit.

[0025] In one embodiment, the communication failure rate is a function of the bit error rate and the packet loss rate:

[0026] P f =w1·P e +w2·P l

[0027] Among them, P f is the communication failure rate, P e is the bit error rate, P l is the packet loss rate, w1 and w2 are weight coefficients.

[0028] In one embodiment, the multi-layer neural network unit includes an input module, a sample parameter set preparation module, a training module and a calculation module;

[0029] The input module is used to receive the evaluation results, time-frequency characteristics, sampling window value and replay interference intensity;

[0030] The sample parameter set production module is used to rate the interference signals according to the evaluation results, select the best-level interference signal as a sample, and record its sample parameters, which include time-frequency characteristics, sampling window value, and replay interference intensity. The sample parameter set is produced based on the sample parameters.

[0031] The training module is used to iteratively train the multi-layer neural network according to the sample parameter set to obtain the relationship function between time-frequency characteristics, sampling window and replay interference intensity;

[0032] The calculation module is used to calculate the current sampling window and replay interference intensity according to the relationship function and the current time-frequency characteristics in an actual environment, and send them to the forwarding interference unit. The forwarding interference unit sends the actual interference signal according to the current sampling window and replay interference intensity.

[0033] In one embodiment, the computing module provides feedback on the cost function related to the communication failure rate, storage resource consumption space, and system power consumption based on the current time-frequency characteristics in the actual environment, adjusts the weight coefficients of each layer of the neural network unit, and calculates the current sampling window and replay interference intensity.

[0034] In one embodiment, the forwarding interference unit obtains the current time-frequency characteristics of the communication unit through short-time Fourier transform.

[0035] One of the above technical solutions has the following advantages and beneficial effects:

[0036] The above-mentioned software-defined intelligent forwarding interference system designs a new structure of the forwarding interference system composed of a communication unit, a forwarding interference unit, an interference effectiveness evaluation unit and a multi-layer neural network unit. According to the known signal sent by the communication unit, a multi-layer neural network method is used to construct a functional relationship between the time-frequency characteristics and the sampling window and the replay interference intensity. The forwarding interference system calculates the current sampling window and replay interference intensity through this functional relationship and the time-frequency characteristics of the signal with unknown parameters in the actual environment, and then implements the interference, thereby realizing the use of the optimal sampling window size and replay interference intensity to implement interference in the actual environment, effectively reducing the storage space of electromagnetic signal acquisition, reducing the power consumption of the forwarding interference system, and saving preparation costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The figure is a schematic diagram of the structure of a software-defined intelligent forwarding jamming system in one embodiment. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application pertains. The terms used in this specification are for the purpose of describing specific embodiments only and are not intended to limit this application.

[0040] In addition, the technical solutions between the various embodiments of the present invention can be combined with each other, but it must be based on the fact that ordinary technicians in this field can implement it. When the combination of technical solutions is mutually contradictory or cannot be implemented, it should be deemed that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0041] This application provides a software-defined intelligent forwarding interference system, such as Figure 1 As shown, it includes a communication unit 11, a forwarding interference unit 12, an interference effectiveness evaluation unit 13 and a multi-layer neural network unit 14.

[0042] The communication unit 11 is used to transmit data messages. The forwarding interference unit 12 includes a spectrum situation awareness module 121, a replay interference module 122 and a storage module 123. The spectrum situation awareness module 121 is used to receive and parse data messages, obtain the current time-frequency characteristics WF of the communication unit 11, and sample the data message by setting the sampling window T to obtain IQ signal data. The replay interference module 122 is used to set the replay interference intensity S and transmit the interference signal. The storage module 123 is used to store the current time-frequency characteristics and IQ signal data. The interference effectiveness evaluation unit 13 is used to evaluate the interference signal and send the evaluation results to the multi-layer neural network unit. The multi-layer neural network unit 14 is used to select sample parameters according to the evaluation results to construct a sample parameter set. The sample parameters include time-frequency features WF, sampling window T and replay interference intensity S. According to the sample parameter set, the multi-layer neural network performs supervised training to obtain the relationship function (T, S) = Φ(WF) among the time-frequency features WF, sampling window T and replay interference intensity S. In an actual environment, the current sampling window and replay interference intensity are calculated according to the relationship function and the current time-frequency features, and sent to the forwarding interference unit. The forwarding interference unit sends the actual interference signal according to the current sampling window and replay interference intensity. The communication unit, forwarding interference unit, interference effectiveness evaluation unit and multi-layer neural network unit are all defined and constructed by software.

[0043] It can be understood that the communication unit 11 can adjust the data message signal by setting parameters such as the modulation mode and the working frequency band. The data message contains the complete data information to be sent. As a communication signal, it can carry multiple types of digital data, such as pictures, videos, HTML documents, software applications, etc.

[0044] It can be understood that the spectrum situation awareness module 121 calculates and measures the surrounding electromagnetic spectrum in real time according to certain rules, obtains the time-frequency feature WF through methods such as short-time Fourier transform, and collects and records the electromagnetic environment by setting the sampling window T, samples IQ data for caching, and the replay interference intensity S set by the replay interference module 122 can adjust the strength of the transmitted interference signal.

[0045] It can be understood that the interference effectiveness evaluation unit 13 can evaluate the impact of the interference signal on the communication link through some preset parameters or standards.

[0046] It can be understood that the multi-layer neural network unit 14 inputs the time-frequency feature WF of the communication unit and outputs the sampling window T and the replay interference strength S of the replay interference module. The multi-layer neural network unit determines the optimal state value after comprehensive weighing based on the evaluation results fed back by the interference effectiveness evaluation unit 13, and records the sample parameters (WF N ,T N,k ,S N,k), where N is the number of times the communication unit sets the modulation parameters and the working frequency band, k is the serial number corresponding to the determined optimal state value, and N is greater than 1. Therefore, the obtained sample parameters are not unique. All sample parameters are summarized and made into a sample parameter set. Then, a multi-layer neural network is used for iteration to obtain the functional relationship between the time-frequency characteristics and the sampling window T and the replay interference intensity S: (T, S) = Φ(WF).

[0047] It can be understood that the time-frequency characteristics in the actual environment belong to the communication signal of unknown parameters. In this embodiment, the communication unit 11 (communication signal with known parameters) is used to train to obtain the functional relationship between the time-frequency characteristics and the sampling window and the replay interference intensity. The functional relationship is further used to calculate the time-frequency characteristics of the communication signal with unknown parameters in the actual environment to obtain the current sampling window and replay interference intensity suitable for the actual environment. These two values are sent to the forwarding interference unit 12, and the forwarding interference unit 12 sends the actual interference signal based on these two values. Compared with the traditional method of simply piling up power consumption to send interference signals, the system adopted in this embodiment can achieve better interference effects in the actual environment while saving system memory and power consumption.

[0048] It can be understood that this system adopts a general software radio architecture, programmable RF hardware and supporting drivers, and combines CPU, GPU, and storage media to build an overall system. To facilitate the quantitative analysis of the overall system, the scope of this system is limited to digital communication.

[0049] It can be understood that software-defined means using software to define the functions of the system, using software to empower hardware, and maximizing the operating efficiency and energy efficiency of the system.

[0050] In one embodiment, the working mode of the communication unit includes a regular communication mode and an irregular communication mode. The data message sending mode in the regular communication mode is periodic sending, and the data message sending mode in the irregular communication mode is random sending.

[0051] It can be understood that the communication unit can switch to different modes, and the data message it sends will contain different parameters. In this embodiment, when the neural network unit is trained, the communication unit can be set to the normal communication mode first, and the modulation mode M of the communication unit is set in this mode. N and operating frequency band f N , obtain and record sample parameters (WF N ,T N,k ,S N,k ), then set the communication unit to an unconventional communication mode, in which the modulation mode M' of the communication unit is set N and operating frequency band f′ N , obtain and record the sample parameters (WF′ N ,T′ N,k ,S′N,k ).

[0052] In one embodiment, the communication unit includes a source coding module, a channel coding module and a modulation and demodulation module. The source coding module is communicatively connected to the channel coding module, and the channel coding module is communicatively connected to the modulation and demodulation module.

[0053] It can be understood that the communication unit implements the complete communication process of source coding → channel coding → modulation module → demodulation module → channel decoding → source decoding through the source coding and decoding module, the channel coding and decoding module and the modulation and demodulation module.

[0054] It can be understood that the functions of the source coding and decoding module include source coding and source decoding. Source coding refers to the method of converting analog signals into digital signals, and source decoding refers to the method of converting digital signals into analog signals. The functions of the channel coding and decoding module include channel coding and channel decoding. Channel coding is the process of error correction and detection coding of digital signals transmitted in the channel. Error correction code types include RS coding, convolutional code, Turbo code, etc. The modulation and demodulation module can realize the modulation process and demodulation process. The modulation process is to use the baseband signal to control the change of one or several parameters of the carrier signal, and load the information on it to form a modulated signal for transmission. The demodulation process is the reverse process of the modulation process, and the original baseband signal will be restored from the parameter change of the modulated signal.

[0055] In one embodiment, the spectrum situation awareness module includes a time-frequency feature calculation component and an IQ signal data sampling component; the time-frequency feature calculation component is used to receive the electromagnetic spectrum of the data message, and obtain the time-frequency feature WF of the communication unit in real time based on the electromagnetic spectrum; the IQ signal data sampling component is used to set a sampling window T to sample the electromagnetic spectrum and obtain IQ signal data.

[0056] It can be understood that the time-frequency feature calculation component has the function of real-time detection, reception, and calculation of the electromagnetic spectrum of data packets. It can analyze the acquired electromagnetic spectrum and calculate the time-frequency characteristics of the communication unit. The sampling window T refers to the time when the sampling component samples the electromagnetic spectrum. The larger the sampling window, the longer the sampling time. Conversely, the smaller the sampling window, the shorter the sampling time.

[0057] In one embodiment, the replay interference module includes a filtering component, a frequency conversion component, a power amplifier circuit and a directional antenna; the filtering component is used to filter and purify the data message to obtain a noise reduction interference source; the frequency conversion component is used to up-convert the noise reduction interference source; the power amplifier circuit is used to set the replay interference intensity S to power amplify the noise reduction interference source after up-conversion; the directional antenna is used to transmit the power-amplified noise reduction interference source to the outside.

[0058] It can be understood that the role of the filtering and purification process is to eliminate the noise of the data message during the transmission process, obtain functional data messages with data information value, and retransmit the functional data message after filtering and purification, which can more effectively interfere with the original data message information; the power amplifier circuit can adjust the strength of the interference signal. By setting the interference intensity S, the signal can be made into an interference signal of the specified interference intensity through the power amplifier circuit.

[0059] In one embodiment, the interference effectiveness evaluation unit includes an evaluation module and a feedback module; the evaluation module uses the communication failure rate P f , storage resource consumption space SP and system power consumption PW to evaluate the interference signal; the feedback module will calculate the communication failure rate P f , storage resource consumption space SP and system power consumption PW are sent to the multi-layer neural network unit.

[0060] It can be understood that in this embodiment, the communication failure rate P f The effect of interference signals is evaluated by three parameters: storage resource consumption space SP and system power consumption PW. Under normal circumstances, the communication failure rate is required to be as high as possible, the storage resource consumption space is as small as possible, and the system power consumption is as low as possible.

[0061] It can be understood that in the multi-layer neural network unit in this embodiment, the cost function C used in the iteration is related to the communication failure rate P of the interference effectiveness evaluation unit. f , Storage resource consumption space SP is related to system power consumption PW: C = w i ·P f +w j SP+w k ·PW

[0062] And the reasonable cost function weight coefficient w can be determined according to the application scenario i , w j , w k .

[0063] In one embodiment, the communication failure rate is a function of the bit error rate and the packet loss rate:

[0064] P f =w1·P e +w2·P l

[0065] Among them, P f is the communication failure rate, P e is the bit error rate, P l is the packet loss rate, w1 and w2 are weight coefficients.

[0066] It can be understood that the bit error rate = bit errors in transmission / total number of transmitted codes * 100%, is an indicator to measure the accuracy of data transmission within the specified time; the packet loss rate = [(input message - output message) / input message] * 100%, which refers to the ratio of the number of lost data packets in the test to the total number of data groups sent.

[0067] In one embodiment, the multi-layer neural network unit includes an input module, a sample parameter set production module, a training module and a calculation module; the input module is used to receive evaluation results, time-frequency features WF, sampling window value T and replay interference intensity S; the sample parameter set production module is used to rate the interference signal according to the evaluation results, select the best-level interference signal as a sample, and record its sample parameters, the sample parameters include time-frequency features WF, sampling window value T and replay interference intensity S, and produce a sample parameter set according to the sample parameters; the training module is used to iteratively train the multi-layer neural network according to the sample parameter set to obtain the relationship function (T, S) = Φ(WF) of the time-frequency features WF, sampling window T and replay interference intensity S; the calculation module is used to calculate the current sampling window and replay interference intensity based on the relationship function and the current time-frequency features, and send them to the forwarding interference unit, and the forwarding interference unit sends the actual interference signal according to the current sampling window and replay interference intensity.

[0068] It can be understood that in this embodiment, the multi-layer neural network unit is trained in the following manner:

[0069] The communication unit selects the conventional communication mode, given the modulation mode M1 and the working frequency band f1, the forwarding interference unit obtains the time-frequency feature WF1, and selects the sampling window T 1,1 , replay interference intensity S 1,1 , the interference effectiveness evaluation unit calculates the communication failure rate And record the average power consumption and storage space consumed during this operation; step the sampling window ΔT, replay the interference intensity step ΔS, and select the sampling window T 1,2 , replay interference intensity S 1,2 , calculate the communication failure rate Record the average power consumption and storage space consumed during this operation; and so on, select the sampling window T 1,N , replay interference intensity S 1,N , calculate the communication failure rate Record the average power consumption and storage space consumed during this operation. Weigh the above communication failure rate, average power consumption, and storage space consumed, determine the optimal state value, and record the sample parameters (WF1, T 1,k ,S 1,k ).

[0070] Given the modulation mode M2 and working frequency band f2, the sampling window step value ΔT, the replay interference intensity step ΔS, the forwarding interference unit obtains the time-frequency feature WF2. According to the above steps, the sample parameters (WF2, T 2,k ,S 2,k Given the modulation mode M3 and the working frequency band f3, record the sample parameters (WF3,T 3,k ,S 3,k ).

[0071] Similarly, given the modulation mode M N and operating frequency band f N , record sample parameters (WF N ,T N,k ,S N,k ).

[0072] The communication unit selects an unconventional communication mode, gives the modulation mode M′1 and the working frequency band f′1, and records the sample parameters (WF′1, T′ 1，k ,S′ 1,k ). In turn, given the modulation mode M′ N and operating frequency band f′ N , record sample parameters (WF′ N ,T′ N,k ,S′ N,k ).

[0073] According to the recorded sample parameters, a training sample set is prepared. By iteratively training the multi-layer neural network, the functional relationship between the time-frequency characteristics and the sampling window and the replay interference intensity is obtained: (T, S) = Φ(WF).

[0074] In one embodiment, the calculation module provides feedback on the cost function related to the communication failure rate, storage resource consumption space, and system power consumption based on the current time-frequency characteristics in the actual environment, adjusts the weight coefficients of each layer of the neural network unit, and calculates the current sampling window and replay interference intensity.

[0075] It can be understood that in order to improve the reasoning and generalization ability of neural networks, in actual environments, such as special outdoor environments, the parameters of the communication unit are unknown. Based on the actual time-frequency characteristics obtained, feedback can be provided through cost functions related to communication failure rate, storage resource consumption space SP, and system power consumption PW to fine-tune and optimize the weight coefficients of each layer of the neural network unit to obtain the ability to adapt to actual scenarios.

[0076] In one embodiment, the forwarding interference unit obtains the current time-frequency characteristics of the communication unit through short-time Fourier transform.

[0077] It can be understood that the short-time Fourier transform (STFT) divides the signal into many small time intervals through a window function, and performs Fourier transform on each time interval to determine the frequency existing in the time interval.

[0078] It can be understood that the method for the forwarding interference unit to obtain the current time-frequency characteristics of the communication unit may also be continuous wavelet transform, constant-Q Gabor transform, Hilbert-Huang transform and other methods.

[0079] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0080] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A software-defined intelligent forwarding jamming system, characterized in that: It includes a communication unit, a forwarding jamming unit, a jamming effectiveness evaluation unit, and a multi-layer neural network unit; The communication unit is used to transmit data messages; The forwarding interference unit includes a spectrum situation awareness module, a replay interference module and a storage module. The spectrum situation awareness module is used to receive and parse the data message, obtain the current time-frequency characteristics of the communication unit, and sample the data message by setting a sampling window to obtain IQ signal data. The replay interference module is used to set the replay interference intensity and transmit the interference signal. The storage module is used to store the current time-frequency characteristics and the IQ signal data. The interference effectiveness evaluation unit is used to evaluate the interference signal and send the evaluation result to the multi-layer neural network unit; The multi-layer neural network unit is used to select sample parameters according to the evaluation result to construct a sample parameter set, where the sample parameters include the time-frequency characteristics, sampling window, and replay interference intensity. Based on the sample parameter set, the multi-layer neural network performs supervised training to obtain a relationship function between the time-frequency characteristics, sampling window, and replay interference intensity. In an actual environment, the current sampling window and replay interference intensity are calculated based on the relationship function and the current time-frequency characteristics, and are sent to the forwarding interference unit. The forwarding interference unit sends an actual interference signal based on the current sampling window and replay interference intensity. The communication unit, forwarding interference unit, interference effectiveness evaluation unit and multi-layer neural network unit are all constructed by software definition.

2. A software-defined intelligent forwarding jamming system according to claim 1, characterized in that: The working mode of the communication unit includes a regular communication mode and an irregular communication mode. The data message sending mode in the regular communication mode is periodic sending, and the data message sending mode in the irregular communication mode is random sending.

3. A software-defined intelligent forwarding jamming system according to claim 2, characterized in that: The communication unit includes a source coding and decoding module, a channel coding and decoding module and a modulation and demodulation module. The source coding and decoding module is communicatively connected to the channel coding and decoding module, and the channel coding and decoding module is communicatively connected to the modulation and demodulation module.

4. The software-defined intelligent forwarding jamming system according to claim 1, characterized in that: The spectrum situation awareness module includes a time-frequency feature calculation component and an IQ signal data sampling component; The time-frequency feature calculation component is used to receive the electromagnetic spectrum of the data message, and obtain the time-frequency features of the communication unit by real-time calculation and measurement based on the electromagnetic spectrum; The IQ signal data sampling component is used to sample the electromagnetic spectrum according to a set sampling window to obtain IQ signal data.

5. The software-defined intelligent forwarding jamming system according to claim 4, characterized in that: The replay interference module includes a filter component, a frequency conversion component, a power amplifier circuit and a directional antenna; The filtering component is used to filter and purify the IQ signal to obtain a noise reduction interference source; The frequency conversion component is used to perform up-conversion processing on the noise reduction interference source; The power amplifier circuit is used to set the playback interference intensity to amplify the power of the noise reduction interference source after up-conversion; The directional antenna is used to transmit the noise reduction interference source after power amplification to the outside.

6. The software-defined intelligent forwarding jamming system according to claim 1, characterized in that: The interference effectiveness evaluation unit includes an evaluation module and a feedback module; The evaluation module evaluates the interference signal based on the communication failure rate, storage resource consumption space and system power consumption; The feedback module sends the communication failure rate, storage resource consumption space and system power consumption to the multi-layer neural network unit.

7. The software-defined intelligent forwarding jamming system according to claim 6, characterized in that: The communication failure rate is a function of the bit error rate and the packet loss rate: P f =w1·P e +w2·P l Among them, P f is the communication failure rate, P e is the bit error rate, P l is the packet loss rate, w1 and w2 are weight coefficients.

8. The software-defined intelligent forwarding jamming system according to claim 1, characterized in that: The multi-layer neural network unit includes an input module, a sample parameter set production module, a training module and a calculation module; The input module is used to receive the evaluation result, time-frequency characteristics, sampling window value and replay interference intensity; The sample parameter set preparation module is used to rate the interference signals according to the evaluation results, select the best-level interference signal as a sample, and record its sample parameters, wherein the sample parameters include time-frequency characteristics, sampling window value, and replay interference intensity, and prepare a sample parameter set according to the sample parameters; The training module is used to iteratively train the multi-layer neural network according to the sample parameter set to obtain the relationship function between the time-frequency characteristics, the sampling window and the replay interference intensity; The calculation module is used to calculate the current sampling window and replay interference intensity according to the relationship function and the current time-frequency characteristics in an actual environment, and send them to the forwarding interference unit. The forwarding interference unit sends the actual interference signal according to the current sampling window and replay interference intensity.

9. The software-defined intelligent forwarding jamming system according to claim 8, characterized in that: The calculation module provides feedback on the cost function related to communication failure rate, storage resource consumption space, and system power consumption based on the current time-frequency characteristics in the actual environment, adjusts the weight coefficients of each layer of the neural network unit, and calculates the current sampling window and replay interference intensity.

10. The software-defined intelligent forwarding jamming system according to any one of claims 1 to 9, characterized in that: The forwarding interference unit obtains the current time-frequency characteristics of the communication unit through short-time Fourier transform.

Images