CN-121239346-B - Unmanned aerial vehicle group dynamic interception method and system based on distributed interference array cooperation

CN121239346BCN 121239346 BCN121239346 BCN 121239346BCN-121239346-B

Abstract

The invention relates to the technical field of electronic countermeasure, in particular to a method and a system for dynamically intercepting unmanned aerial vehicle groups based on distributed interference array cooperation. The method comprises the steps of dynamically deploying an interference node cluster with autonomous logic association in a defending domain, forming an elastically reconfigurable distributed interference domain through a software defined radio architecture, extracting time-varying spectrum features through an anti-neural network, constructing a dynamic threat fingerprint library, generating an interference vector with frequency band following and power self-adaption based on the dynamic threat fingerprint library, updating feature weights of the dynamic threat fingerprint library in real time according to interference efficiency feedback, generating a layered interference wave order sequence based on a multi-agent consensus protocol according to situation evolution trend of the dynamic threat fingerprint library, and forming a space-time coupled interception chain. The invention realizes the whole-process dynamic countermeasure capability of the unmanned aerial vehicle group, maintains the high-efficiency interception performance in a complex electromagnetic environment, and reduces the interference risk to legal communication.

Inventors

HUANG JIEWU
CHEN QISHEN
CHEN JIERUI
HUANG JUNHAO
HUANG RUOXIN
KE ZHIQIANG

Assignees

深圳市雅诺讯科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251010

Claims (7)

1. The unmanned aerial vehicle group dynamic interception method based on the cooperation of the distributed interference array is characterized by comprising the following steps: s102, dynamically deploying an interference node cluster with autonomous logic association in a defending domain, and forming a distributed interference field capable of being elastically reconstructed through a software defined radio architecture, wherein each node shares a destruct-resistant communication channel; S104, based on the real-time captured unmanned aerial vehicle group communication signal flow, a dynamic threat fingerprint library comprising a protocol behavior mode and a frequency hopping evolution path is constructed by extracting time-varying frequency spectrum features through an antagonistic neural network; s106, generating an interference vector with frequency band following and power self-adaption based on the dynamic threat fingerprint library, and updating the characteristic weight of the dynamic threat fingerprint library in real time according to interference efficiency feedback; S108, generating a layered interference wave sequence based on a multi-agent consensus protocol according to the situation evolution trend of the dynamic threat fingerprint library to form a space-time coupled interception chain; the S102 includes: synchronously acquiring a signal arrival angle, an instantaneous bandwidth and a polarization scattering matrix through a software-defined radio node array, and constructing a dynamically updated space-time-frequency three-dimensional tensor; Performing bispectrum analysis on the space-time-frequency three-dimensional tensor to extract nonlinear phase coupling characteristics, and generating a bispectrum fingerprint template with polarization sensitivity by adopting tensor decomposition; calculating the polarization isolation between nodes based on the bispectrum fingerprint template, dividing the channels into orthogonal polarization codebook clusters through dynamic polarization coding, and generating a polarized channel topological graph with conflict marks; Establishing a node polarization constraint relation according to conflict marks of a polarization channel topological graph, calculating node position adjustment vectors in real time by adopting a gradient descent method with polarization mismatch as a target, and driving a cluster to form a polarization isolation array configuration; Inputting the bispectrum fingerprint template into a pre-trained variable self-encoder to generate an interference waveform codebook set containing polarization deception parameters, wherein the codeword index of the interference waveform codebook set is bound with a polarization codebook cluster; the waveform codebook index is broadcast to each node through a polarization codebook cluster, and corresponding digital predistortion coefficients and polyphase filter bank parameters are loaded based on a software defined radio architecture to form a polarization self-adaptive coherent interference beam network.
2. The unmanned aerial vehicle group dynamic interception method based on distributed interference array cooperation according to claim 1, wherein the method is characterized in that the inter-node polarization isolation is calculated based on the bispectrum fingerprint template, channels are divided into orthogonal polarization codebook clusters through dynamic polarization coding, and a polarized channel topological graph with conflict marks is generated, specifically: Carrying out main polarization component decomposition on the bispectrum fingerprint template, and obtaining complex correlation coefficients of polarization state vectors among nodes to generate an N multiplied by N polarization correlation matrix, wherein N is the number of nodes; Comparing each off-diagonal element of the polarization correlation matrix with a preset polarization isolation threshold, and marking the off-diagonal element as a reusable node pair when a certain diagonal element is smaller than the preset polarization isolation threshold, or marking the off-diagonal element as a polarization conflict node pair; For the reusable node pairs, adopting the Galame-Schmitt orthogonalization to process the polarization state vector thereof to generate a group of orthogonalization polarization bases; and (3) using the nodes as vertexes, constructing an edge relation according to conflict marks in the polarized codebook set, adding red conflict edges into the topological graph when the two nodes share the same polarization coding index and are conflict node pairs, otherwise, adding green multiplexing edges, and finally outputting the bipolarized topological graph with the color marks.
3. The method for dynamically intercepting unmanned aerial vehicle groups based on distributed interference array cooperation according to claim 1, wherein the step S104 is specifically: performing time-frequency atomic decomposition on the received communication signal flow, extracting transient pulse components and steady carrier components, and generating a time-frequency atomic sparse representation matrix; Inputting the time-frequency atomic sparse representation matrix into a generator for generating an antagonism network, synthesizing a time-frequency mask with antagonism through multi-layer time sequence convolution and an attention mechanism, simultaneously calculating the spectral correlation loss of the time-frequency mask by a discriminator based on the signal cyclostationary characteristic, and outputting a time-frequency characteristic tensor with enhanced antagonism; carrying out protocol semantic analysis on the time-frequency characteristic tensor with the countermeasure enhancement, matching a preset protocol state machine template through a sliding window, extracting a frequency hopping period, a frequency point offset and a modulation constellation diagram variation parameter in a protocol field, and constructing a protocol behavior mode vector; carrying out Viterbi grid search on the time-frequency characteristic tensor, tracking phase continuity characteristics and power spectrum entropy change in the frequency point switching process, and generating a state transition probability matrix of a frequency hopping evolution path; Tensor splicing is carried out on the protocol behavior pattern vector and the state transition probability matrix, and a dynamic threat fingerprint library with space-time correlation is generated through self-organizing mapping neural network clustering, wherein each fingerprint unit comprises protocol behavior pattern cluster heart coordinates and frequency hopping path Markov chain transition probability.
4. The method for dynamically intercepting unmanned aerial vehicle groups based on distributed interference array cooperation according to claim 1, wherein the step S106 is specifically: Extracting a protocol behavior pattern cluster center coordinate and a frequency hopping path Markov chain transition probability at the current moment from a dynamic threat fingerprint library, predicting the frequency band center frequency and the bandwidth of the next frequency hopping point through a Viterbi decoder, and generating an initial interference frequency band vector followed by the frequency band; Performing complex domain point multiplication operation on the initial interference frequency band vector and the power spectrum density of the unmanned aerial vehicle group signal acquired in real time, calculating the signal-to-interference ratio gain of each frequency band, and adjusting the interference power distribution weight based on the power spectrum entropy when the signal-to-interference ratio gain is lower than a preset gain threshold value to generate a power self-adaptive interference vector; After the interference vector is applied, acquiring a frequency offset error vector and a protocol retransmission rate of the interfered unmanned aerial vehicle through a software defined radio node array, calculating an interference efficiency evaluation index, and when the interference efficiency evaluation index exceeds a dynamic threshold, judging that the current interference strategy is effective and triggering fingerprint library update; correcting the cluster center coordinates of the protocol behavior mode according to the Euclidean distance of the frequency offset error vector, and adjusting the Markov chain weight coefficient of the state transition probability matrix according to the correlation coefficient of the protocol retransmission rate and the frequency hopping path; And re-inputting the updated cluster center coordinates and state transition probability matrix into the self-organizing mapping neural network for topology mapping, generating a dynamic threat fingerprint library with space-time correlation, and completing iterative optimization of the feature weights.
5. The method for dynamically intercepting unmanned aerial vehicle groups based on distributed interference array cooperation according to claim 1, wherein the step S108 is specifically: Extracting a protocol behavior mode cluster heart coordinate and a frequency hopping path Markov chain state transition probability matrix at the current moment from a dynamic threat fingerprint library, predicting a communication frequency band migration track of the unmanned aerial vehicle group in a future time window through a space-time convolution network, and generating a threat situation evolution tensor containing a time stamp-frequency band mapping relation; Inputting threat situation evolution tensor into a multi-agent consensus protocol framework, and calculating space-time coverage of each unit to threat tracks according to interference beam coverage of a soft interference unit and physical interception radius of a hard interference unit to generate an initial interception efficiency evaluation matrix; Based on space-time coverage difference in the interception efficiency evaluation matrix, coordinating decision weights of all the intelligent agents, performing joint optimization on the frequency band suppression priority of the soft interference unit and the interference time sequence of the hard interference unit, and outputting a layered interference wave order sequence with a time synchronization mark; Triggering a guidance radar of a hard interference unit to start up while applying an interference vector according to a time synchronous mark in the layered interference wave sequence, so that a frequency spectrum suppression blind area of soft interference and an interception coverage window of the hard interference are complementary; and monitoring threat situation evolution tensor residual errors after the interference wave sequence is executed in real time, when the threat situation evolution tensor residual errors exceed a preset residual error threshold, reallocating interception time sequences of the inactive standby units through a multi-agent consensus protocol, and adjusting frequency band interception coupling parameters of subsequent waves to maintain space-time continuity of an interception chain.
6. The method for dynamically intercepting a group of unmanned aerial vehicles based on cooperation of a distributed interference array according to claim 1, wherein the method for dynamically intercepting a group of unmanned aerial vehicles further comprises: detecting the polarization coding state of each node interference beam in real time, and judging as a failure node when the deviation degree of the polarization coding and the preset orthogonal constraint exceeds a threshold value or the time-frequency characteristic tensor output by the node cannot be synchronously updated with the system reference; If the failure node exists, removing the corresponding vertex and the associated edge from the polarized channel topological graph, generating a bipartite polarized topological subgraph of the rest nodes, and extracting the number of polarized conflict edges of each node in the subgraph as a logic association index; calculating a dynamic survivability coefficient according to the logic association degree and the historical destruction-resistant duration of the rest nodes, preferentially selecting nodes with survivability coefficients higher than a preset coefficient threshold as communication relay nodes, and generating a polarization isolation constraint set of a relay cluster based on the polarization codebook index; reconstructing a coherent interference beam network by using a polarization isolation constraint set of a relay cluster, loading a digital predistortion coefficient bound with a residual node polarization codebook through a software-defined radio architecture to form a polarization self-adaptive beam coverage network taking a relay node as a core, and marking a failure interception layer corresponding to a beam coverage blind area; according to the space-time coordinates of the beam coverage blind area, screening standby nodes with polarization coding indexes matched with the blind area from a standby unit pool, calculating position adjustment vectors of the standby nodes by a gradient descent method to fill the blank of an interception layer, and synchronously updating the cluster center coordinates of the protocol behavior mode in the dynamic threat fingerprint library to adapt to new topology; Based on the reconstructed polarized self-adaptive wave beam coverage network, the space-time coverage in the multi-agent consensus protocol is recalculated, and the time synchronization mark of the layered interference wave order sequence is adjusted, so that the interception time sequence of the standby unit and the frequency band interception coupling parameter of the original wave order form a space-time continuous interception chain.
7. A distributed interference array collaboration-based unmanned cluster dynamic intercept system for performing the method of any of claims 1 to 6, comprising: The full-frequency band programmable interference node module adopts a software defined radio architecture, supports 300MHz-6GHz full-frequency band dynamic programmable interference, has polarization self-adaptive beam forming capability, can synchronously suppress GPS/GLONASS navigation signals, 2.4GHz remote control links and 5.8GHz image transmission bands, and realizes the elastic reconstruction of interference beams through space-time-frequency three-dimensional tensor analysis; The intelligent spectrum countermeasure and threat fingerprint library module is used for extracting the frequency hopping pattern, the protocol semantic features and the cyclostationary features of the unmanned aerial vehicle group in real time based on the countermeasure neural network, constructing a dynamic threat fingerprint library, and optimizing the frequency hopping path tracking by fusing Viterbi grid search; The self-adaptive interference decision module dynamically generates interference vectors of frequency band following and power self-adaptation according to the Markov chain transition probability of the threat fingerprint library, and optimizes the interference strategy in real time through complex domain signal-to-interference ratio feedback; and the multi-agent collaborative interception scheduling module is used for calculating the space-time coverage of the soft interference unit and the hard interference unit based on a distributed consensus protocol, generating a layered interference wave sequence and realizing space-time coupling of navigation suppression, remote control chain breaking and physical interception.

Description

Unmanned aerial vehicle group dynamic interception method and system based on distributed interference array cooperation Technical Field The invention relates to the technical field of electronic countermeasure, in particular to a method and a system for dynamically intercepting unmanned aerial vehicle groups based on distributed interference array cooperation. Background With the rapid development of unmanned aerial vehicle technology, the wide application of unmanned aerial vehicle technology also brings new challenges of airspace security management, and especially under the scenes of airport periphery, major activity security and the like, unauthorized unmanned aerial vehicle activity can cause serious threat to public security. The current mainstream unmanned aerial vehicle countering technology mainly relies on GPS interference or full-band suppression, but the method has the problems of low interference efficiency, easy accidental injury of legal communication and the like. Interference schemes based on spectrum monitoring in the prior art often lack dynamic adaptability to frequency hopping communication protocols, and are difficult to cope with complex scenes of collaborative invasion of multiple unmanned planes. In addition, the traditional fixed type interference equipment has the defects of limited coverage range, inflexible deployment and the like, and cannot meet the dynamic defense requirement. In the technical implementation level, the existing scheme mostly adopts a centralized control architecture, has poor survivability and lacks a real-time optimization mechanism for an interference strategy. More importantly, the problem of reasonable allocation of spectrum resources in the interference process is not solved effectively in the prior art, and excessive interference is easy to cause and violate radio management regulations. Therefore, development of an unmanned aerial vehicle group management and control technology with intelligent spectrum sensing and dynamic interference coordination capability is needed to be developed, and interference risks to surrounding legal radio services are reduced to the greatest extent while effective interception is ensured. Disclosure of Invention The invention overcomes the defects of the prior art and provides a method and a system for dynamically intercepting unmanned aerial vehicle groups based on cooperation of a distributed interference array. The technical scheme adopted by the invention for achieving the purpose is as follows: the invention discloses a unmanned aerial vehicle group dynamic interception method based on distributed interference array cooperation, which comprises the following steps: s102, dynamically deploying an interference node cluster with autonomous logic association in a defending domain, and forming a distributed interference field capable of being elastically reconstructed through a software defined radio architecture, wherein each node shares a destruct-resistant communication channel; S104, based on the real-time captured unmanned aerial vehicle group communication signal flow, a dynamic threat fingerprint library comprising a protocol behavior mode and a frequency hopping evolution path is constructed by extracting time-varying frequency spectrum features through an antagonistic neural network; s106, generating an interference vector with frequency band following and power self-adaption based on the dynamic threat fingerprint library, and updating the characteristic weight of the dynamic threat fingerprint library in real time according to interference efficiency feedback; S108, generating a layered interference wave order sequence based on a multi-agent consensus protocol according to the situation evolution trend of the dynamic threat fingerprint library to form a space-time coupled interception chain. Preferably, the S102 specifically is: synchronously acquiring a signal arrival angle, an instantaneous bandwidth and a polarization scattering matrix through a software-defined radio node array, and constructing a dynamically updated space-time-frequency three-dimensional tensor; Performing bispectrum analysis on the space-time-frequency three-dimensional tensor to extract nonlinear phase coupling characteristics, and generating a bispectrum fingerprint template with polarization sensitivity by adopting tensor decomposition; calculating the polarization isolation between nodes based on the bispectrum fingerprint template, dividing the channels into orthogonal polarization codebook clusters through dynamic polarization coding, and generating a polarized channel topological graph with conflict marks; Establishing a node polarization constraint relation according to conflict marks of a polarization channel topological graph, calculating node position adjustment vectors in real time by adopting a gradient descent method with polarization mismatch as a target, and driving a cluster to form a polarization isolation array configur