A distributed intelligent anti-jamming communication system based on space-air-ground integrated self-organizing network

Abstract

The application discloses a kind of distributed intelligent anti-interference communication systems based on space-air-ground integration self-organizing network, belong to space-air communication and electromagnetic anti-interference field.It includes multi-layer multi-orbit dynamic grid topology, dual cluster head hot backup anti-interference architecture and other core designs, and fuses the cross-level mixed anti-interference algorithm system of orbital dynamics, with full-dimensional dynamic performance evaluation system and special scene adaptation mechanism, to realize the elastic adaptation of small / medium / large 3 types of scales, 3 types of special scenes.The application solves the problems of poor topology adaptation and fragmented anti-interference system in the prior art, improves the communication reliability and survivability in harsh environments, and is suitable for national defense communication, emergency rescue and other fields with high communication requirements.

Description

Technical Field

[0001] This invention belongs to the field of aerospace communication and electromagnetic interference suppression technology, specifically involving a distributed intelligent interference suppression communication system based on an integrated aerospace-ground self-organizing network. It features an integrated distributed self-organizing network architecture design combining low-Earth orbit satellites, unmanned aerial vehicles (UAVs), and ground stations; a cross-level hybrid interference suppression algorithm integrating orbital dynamics, scale characteristics, and scenario characteristics; and a comprehensive dynamic performance evaluation system oriented towards space environment disturbances. It is applicable to low-Earth orbit satellite communication, UAV swarm navigation communication, and seamless aerospace-ground communication scenarios under harsh conditions such as complex electromagnetic interference, ionospheric disturbances, and solar activity, covering three scale categories (small / medium / large) and three special scenarios (maritime / polar / high altitude). Its key applications include defense communication, emergency rescue, low-altitude economic activities, ocean communication, and polar scientific research—fields with extremely high requirements for communication reliability and resilience. Background Technology

[0002] Current Status of Industry Development With the evolution of 6G communication technology and the large-scale deployment of low-Earth orbit satellite constellations (Starlink, OneWeb, etc.), integrated air-space-ground networks have become the core architecture for global communication. It is estimated that by 2030, the number of low-Earth orbit satellites worldwide will exceed 100,000, and the scale of drone swarms will reach thousands. At the same time, the demand for communication in extreme environments (at sea, in the polar regions, and at high altitudes) will continue to grow, but existing technologies cannot meet the dual requirements of "flexible scaling and compatibility with special scenarios".

[0003] Poor topology dynamic adaptability: Traditional fixed grid topology increases the frequency of polar link reconfiguration by 42.7%, with reconfiguration latency > 2s, communication interruption rate ≥ 15%, and does not consider the impact of terrain / environmental occlusion in special scenarios; Fragmented anti-interference system: Satellite, UAV and ground station anti-interference systems are independent, without cross-node / cross-scenario interference information sharing mechanism, and cannot cope with distributed interference and strong interference in special scenarios; Lack of scale adaptability: With fixed network parameters and anti-interference algorithms, the outage rate increases by 40% after scaling up small-scale models, and the resource redundancy rate is ≥30% after scaling down large-scale models. Insufficient compatibility with special scenarios: The hardware lacks extreme environment protection design, and the algorithm does not incorporate scenario characteristics. In environments such as sea salt spray, polar low temperature, and high altitude low pressure, the system reliability drops to below 60%. Performance evaluation distortion: The evaluation results deviate from the actual scenario by more than 20% because the scale adaptability and scenario adaptability indicators are not integrated. Summary of the Invention

[0004] Multi-layer, multi-orbit dynamic mesh topology (scalable): Integrating Walker-Delta configuration with dedicated redundant links for polar regions, it supports flexible networking of satellite constellations of small, medium, and large scales. The frequency of link reconfiguration in polar regions is reduced by 53.2%, the topology reconfiguration latency is ≤500ms, and the link stability attenuation is ≤5% after scaling up or down. Dual cluster head hot backup anti-interference architecture (cluster elastic adaptation): HEED algorithm dynamically elects primary / backup cluster heads, the number of cluster heads is automatically adjusted according to the cluster size, combined with GCN anti-interference path planning + DSAC dynamic channel selection, the full-scale cluster communication interruption rate is ≤0.08%, and the scale expansion and contraction adaptation time is ≤10s. High-altitude ground station relay subsystem (coverage elastic matching): 20km high-altitude HAPS deployment, supports 1-4 HAPS elastic networking, 2 HAPS are equivalent to 11 traditional ground stations, satellite visibility time is extended by 87.5%, and coverage matching can be achieved for clusters of different sizes / scenarios; A cross-level hybrid anti-interference algorithm system integrating orbital dynamics (full-scale adaptation): algorithm parameters dynamically iterate with node size, interference identification accuracy ≥99.2%, can resist 22.3dB interference-to-noise ratio, and anti-interference performance degradation ≤3% in small / medium / large scale scenarios; A comprehensive dynamic performance evaluation system (scale-based quantitative evaluation): The system adds a scale adaptability evaluation index, which, combined with spatial environmental disturbance factors, achieves an evaluation deviation of ≤4.2%, providing a precise quantitative basis for optimizing the anti-disturbance strategies of clusters of different sizes. Dynamic adaptation mechanism for special scenarios: Optimize hardware protection and algorithm parameters for the environmental characteristics of marine (salt spray / turbulence), polar (low temperature / polar day / polar night), and high-altitude (low air pressure / strong radiation) environments, and achieve triple dynamic adaptation of "environment-scale-anti-interference strategy", with system reliability ≥98% under special scenarios; Cross-scenario spectrum coordination technology: Based on the electromagnetic propagation characteristics of special scenarios, a dynamic switching mechanism for Ku / Ka / V bands is constructed, which improves spectrum utilization by more than 40% and reduces link interruption rate to below 0.05% in marine / polar / high-altitude scenarios.

[0005] Mathematical Derivation of Core Algorithm

[0006] GCN UAV Anti-interference Path Planning Algorithm (Scale-Scenario Collaborative Version) Graphical model definition: V represents the set of drone nodes (size) E represents the communication link set and the node feature matrix. (d is the basic feature dimension, k is the scene feature dimension, including temperature, air pressure, interference intensity, and terrain complexity).

[0007] Layer propagation formula:

[0008] in, As a scale-scenario synergy factor, , ( (Base size number of nodes) For scene feature vectors, For the scene weight matrix, This is a bias term.

[0009] Objective function:

[0010] New Scene interference penalty item, adapted to the strong interference characteristics of special scenes.

[0011] DSAC Dynamic Channel Selection Algorithm State space: (Integration Scale - Scenario Characteristics) Action space: (The number of channels is dynamically adjusted according to the scale and scenario)

[0012] Core optimization: Dynamic iteration of dual network parameters is achieved through scale-scenario coordination factors, with channel switching latency ≤30ms.

[0013] Disturbance identification model integrating orbital dynamics Orbital dynamics constraints:

[0014] New (Number of satellite constellation sizes) (Scale perturbation coefficient) (Scene pressure perturbation compensation item) Adapts to satellite orbit characteristics of different scales / scenes.

[0015] Model structure: ResNet18 (signal features) + LSTM (orbit features + scale features + scene features) + attention fusion layer, supporting cross-node joint recognition of 4 types of interference (narrowband continuous wave / pulse / wideband noise / spoofing) and 3 types of scenes, with recognition accuracy ≥98.5% and response time ≤80ms. Attached Figure Description

[0016] Figure 1 System overall architecture diagram Figure 2 Schematic diagram of multi-layer, multi-orbit dynamic grid topology and polar redundancy links for low-Earth orbit satellites. Figure 3 Schematic diagram of UAV swarm dual-cluster head cluster topology and GCN anti-interference path planning Figure 4Schematic diagram of the working principle of the relay subsystem of the high-altitude ground station Figure 5 Flowchart of Distributed Intelligent Anti-interference Communication Implementation Method Figure 6 Linear graph comparing interference suppression ratios under different interference types Figure 7 : Scale-Scenario Adaptation Parameter Comparison Table.

Claims

1. A distributed intelligent anti-interference communication system based on an integrated air-space-ground self-organizing network, characterized in that, The system comprises a low-Earth orbit (LEO) satellite constellation subsystem, a drone swarm self-organizing subsystem, a high-altitude ground station relay subsystem, an air-space-ground collaborative anti-interference control center, and a dynamic performance evaluation subsystem. The LEO satellite constellation subsystem employs a multi-layer, multi-orbit dynamic grid topology, integrating a laser communication module, a multi-band adaptive radio frequency module, an orbital dynamics sensing module, and a scale-scenario adaptation module. The drone swarm self-organizing subsystem uses a dual-cluster head topology, configured with a cluster head dynamic election module, dynamically electing primary / backup cluster head nodes using the HEED algorithm, with the number of nodes automatically adjusting according to system scale. The high-altitude ground station relay subsystem deploys 1-4 high-altitude platform stations, integrating a DTN architecture data caching module, a phased array antenna module, a space environment sensing module, and a scene protection module. The air-space-ground collaborative anti-interference control center is equipped with an AI intelligent anti-interference model that integrates orbital dynamics, scale, and scene characteristics, with algorithm parameters dynamically iterating according to scale and scene. The dynamic performance evaluation subsystem constructs a comprehensive evaluation index system including scale adaptability and scene adaptability, employing a greedy algorithm + genetic algorithm + ... Tabu search algorithm completes quantitative evaluation of system anti-interference performance; each subsystem achieves interconnection and interoperability by building a unified control plane through software-defined networking, supporting flexible networking and collaborative anti-interference in three scale categories (small / medium / large) and three special scenarios (marine / polar / high altitude).

2. The system according to claim 1, characterized in that, The three types of low-Earth orbit satellite constellations are configured as follows: 12 orbital planes × 22 satellites, 24 orbital planes × 44 satellites, and 36 orbital planes × 66 satellites, with 1 / 2 / 3 polar redundancy links and a topology reconfiguration delay of ≤500ms; the laser communication module transmission rate is configured to 50Gbps, 100Gbps, and 100Gbps depending on the scale, and the radio frequency module supports dual-band or full-band adaptation.

3. The system according to claim 1, characterized in that, The three types of drone swarm configurations are 32, 96, and 256 drones, with cluster heads of 1 main and 1 backup, 2 main and 2 backup, and 4 main and 4 backup, respectively. The cluster head switching time is ≤10ms, and the swarm size scaling adaptation time is ≤10s.

4. The system according to claim 1, characterized in that, The number of high-altitude platform stations deployed is configured as 1 / 2 / 4 depending on the scale. Two high-altitude platform stations have the coverage performance equivalent to 11 traditional ground stations, and the satellite visibility time is extended by more than 87.5%. In addition, for marine / polar / high-altitude scenarios, scenario protection modules with resistance to salt spray corrosion, low temperature start-up, and low air pressure heat dissipation are configured respectively, and the hardware has a continuous working time of ≥800 hours.

5. The system according to claim 1, characterized in that, The scale-scene adaptation module obtains γ'=γ・σ(W) by adjusting the scale factor γ=n0 / n (n0 is the baseline scale number of nodes, n is the actual number of nodes) in conjunction with the scene feature vector E=(e1,e2,e3,e4). e •E+b e This enables collaborative adaptation, and the link stability attenuation is ≤5% after large-scale adaptation.

6. The system according to claim 1, characterized in that, The AI ​​intelligent anti-interference model integrates ResNet18 and LSTM networks, and can identify four types of interference, including narrowband continuous wave and pulse interference, with an average recognition accuracy of ≥98.8%. The drone swarm anti-interference path planning adopts the GCN algorithm and incorporates a scene interference penalty term to adapt to strong interference in special scenes.

7. The system according to claim 1, characterized in that, The dynamic channel selection adopts the DSAC algorithm, and the channel switching delay is ≤30ms. In the system's hierarchical anti-interference strategy, the physical layer uses the MVDR algorithm + improved ANF filter, the network layer uses hybrid game theory + PSO algorithm, and the application layer uses the PR-VM-EASI blind source separation algorithm, which can resist interference-to-noise ratios of more than 21.5dB.

8. The system according to claim 1, characterized in that, The comprehensive evaluation index system includes six dimensions such as communication quality and anti-interference capability. It is configured with 15 / 18 / 23 indicators under three scale categories, and the evaluation results deviate from the actual scenario by ≤4.5%.

9. The system according to claim 1, characterized in that, The key performance indicators of the system are: communication interruption rate ≤ 0.09%, transmission latency ≤ 60ms, and bit error rate ≤ 9.2 × 10⁻⁻⁻⁶. 7 Interference suppression ratio ≥21.5dB, system reliability ≥98% in special scenarios; data transmission reliability of high-altitude ground station relay subsystem ≥93% in marine / polar / high-altitude scenarios, and ≥95% in general scenarios.

10. A distributed intelligent anti-interference communication implementation method based on an integrated air-space-ground self-organizing network, applied to the system described in any one of claims 1-9, characterized in that, The process includes the following steps: Step 1: Distributed networking across air, space, and ground: After each subsystem completes its own networking, it synchronizes hardware status, link information, interference, and scale scenario data through the unified control plane of SDN; Step 2: Joint interference perception and identification: Each node uploads data to the control center, and the AI ​​intelligent anti-interference model completes cross-node joint identification of interference type, intensity, and scale scenario adaptability; Step 3: Cooperative anti-interference strategy generation: Based on the identification results and the scale-scenario coordination factor γ', the control center generates hierarchical anti-interference strategies and distributes them to each node; Step 4: Dynamic strategy execution and adaptation: Each node adjusts hardware and algorithm parameters, and the cluster head module automatically adjusts according to the scale scenario, achieving millisecond-level dynamic adaptation of the strategy; Step 5: Performance evaluation and closed-loop optimization: The evaluation subsystem completes quantitative evaluation of indicators, and when the threshold is not reached, iteratively optimizes the anti-interference strategy based on the adaptation results to form a closed-loop control.

Images