Three-mode four-stage nine-topping centerless dynamic self-healing space-based satellite anti-jamming networking system

Abstract

This invention discloses a three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-jamming networking system, belonging to the field of software-defined satellite and space-based anti-jamming communication technology. It uses four peer satellites as the smallest autonomous networking unit, with each satellite configured with three x three nine independent communication beams. It features three switchable operating modes: static civilian fixed network, dynamic free networking, and wartime millisecond self-healing. No modifications are made to the satellite hardware, and no additional routing satellites are required. A networking program with all built-in mathematical models is remotely injected via the uplink, enabling each on-orbit satellite to possess distributed routing and forwarding capabilities and local autonomous computation. It can be used in conjunction with already deployed space-based routing satellite networks, providing mutual backup and ensuring continuous and stable communication. All topology rules, calculation formulas, and constraints can be directly used in subsequent software source code development, improving the space-based network's anti-jamming and anti-destruction capabilities, reducing constellation construction costs, and is suitable for both military and civilian space-based communication scenarios.

Description

Technical Field

[0001] This invention belongs to the technical fields of software-defined satellites, space-based networking, space-based electromagnetic interference-resistant communication, and distributed self-healing networks. Specifically, it relates to a three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system. This invention is a space-based satellite networking algorithm system with a complete mathematical model that can be directly programmed and implemented. It is suitable for applications such as low-Earth orbit communication constellations, medium- and high-Earth orbit relay networks, and dual-use (military and civilian) space-based emergency communication. Background Technology

[0002] Currently, global commercial and military communication satellite networks generally adopt a static constellation system with fixed orbits, fixed topology, and fixed grouping. Satellite orbital parameters are fixed in the long term, and the members of the cluster network are pre-planned and remain unchanged over the long term; inter-satellite links are pre-locked, and routing and forwarding rely on dedicated space-based routing satellites for centralized scheduling; communication links are stable with low latency and jitter, which can meet the basic requirements of long-term stable communication scenarios in civilian applications.

[0003] However, as space warfare enters the stage of invisible electronic warfare, adversaries can use soft-kill methods such as electromagnetic suppression and targeted radio frequency jamming to sever communication links between satellites and ground stations, and between satellites themselves, without directly destroying spacecraft. In extreme cases, they can even carry out hard-kill attacks on key communication nodes, completely severing space-based information links. It is precisely because of the practical need to address these space communication security risks that my country has proposed a plan to launch two dedicated space router satellites. These router satellites will be responsible for inter-satellite data exchange and cross-constellation forwarding functions, thereby changing the previous model of satellite communication overly reliant on ground gateway stations.

[0004] A search revealed the following existing solutions: First, a master-slave architecture-based autonomous satellite cluster approach: This scheme sets up a master satellite and slave satellites in the cluster, with the master satellite responsible for topology management and data forwarding, while the slave satellites only establish communication links with the master satellite. This scheme has a significant centralized dependency defect; the failure of the master satellite will paralyze the cluster.

[0005] Second, a distributed topology planning method using a three-satellite rigid formation is adopted: This scheme uses three satellites to form a rigid formation, with each pair of satellites interconnected to form a triangular topology, which has a certain degree of topology self-stabilization capability. However, the topology structure of the three-satellite rigid formation is fixed, lacking the ability to flexibly expand and dynamically splice across units, and the anti-interference redundancy of the three beams is limited.

[0006] Third, a self-healing method relying on centralized ground computing is adopted: after a satellite failure occurs, the failure information is reported to the ground station, where the ground computing center uniformly calculates a new routing scheme and then injects rerouting commands into the satellite through the telemetry, tracking, and command network. This scheme has a recovery latency of seconds to minutes, which cannot meet the timing requirements of millisecond-level warfare in wartime.

[0007] Fourth, an anti-interference communication method based on frequency hopping spread spectrum is adopted: This method avoids narrowband interference and improves the anti-interference capability of the link by switching between the time and frequency domains. However, this method only improves the anti-interference capability at the single link level and does not comprehensively solve the anti-interference problem at the system level, such as satellite beam redundancy and topology redundancy.

[0008] Based on system analysis and engineering assessment, existing space-based communication networking solutions have at least the following five significant shortcomings: First, the network architecture adopts a star-shaped centralized topology. Data forwarding, topology management, and path scheduling rely entirely on dedicated space-based routing core nodes, with all operational satellites communicating only with these routing satellites. When both the primary and backup routing satellites are destroyed, the entire subnet collapses, posing a fatal single point of failure risk. This fragile architecture, where a single failure can have catastrophic consequences, constitutes a fundamental security vulnerability in space-based communication systems.

[0009] Second, the fault recovery process lags significantly behind the wartime engagement sequence. Traditional fault recovery relies entirely on ground station involvement: fault reporting to the ground station, ground computing center resolving routing replanning schemes, transmitting replanning commands to satellites via the telemetry, tracking, and command (TT&C) network, and network-wide negotiation and reconstruction. The recovery latency of this entire process is in the range of seconds to minutes, while the engagement window for modern space warfare, space-based guidance, and instantaneous strikes is in the millisecond range. The recovery speed of traditional solutions is far from meeting the timing requirements of transient wartime engagements.

[0010] Third, traditional satellites are equipped with only three communication beams. The probability of a single satellite failing when all three beams are suppressed by electromagnetic interference is... When the enemy implements high-power electromagnetic suppression, it is extremely vulnerable to being completely suppressed and rendered communication silent, exhibiting severely inadequate resistance to radio frequency interference. In contrast, if any one of the nine beams is jammed, the satellite still retains eight usable links, while the failure of any beam in a three-beam architecture results in the loss of more than one-third of its communication capacity.

[0011] Fourth, existing solutions lack a complete and coded mathematical model system. Publicly available technical documents only implement single functions such as self-healing and redundant communication in a scattered manner, without a complete mathematical evaluation model for the dedicated topology. They also lack a unified mathematical description of the two sets of operating logics for normal steady state and wartime transient state, making it impossible to enable existing satellites to acquire distributed self-healing networking capabilities through pure software upgrades. Furthermore, it is impossible to solidify all mathematical formulas, matrix models, and timing constraints and directly write them into source code for on-orbit operation.

[0012] Fifth, the construction cost is high and existing resources cannot be utilized effectively. Existing solutions rely on continuously launching new satellites to improve network performance and on stacking hardware to achieve network stability, resulting in a large number of satellite deployments and extremely high redundancy costs. Launching two dedicated space router satellites requires high research and development and launch costs, and existing satellites already in orbit cannot be upgraded synchronously through software to gain self-healing networking capabilities, resulting in a relatively low overall return on investment.

[0013] In summary, the industry currently lacks a complete set of mathematical formulas, matrix models, and time-series constraints that are all fixed and can be directly coded into a decentralized self-healing networking solution. It also lacks a unified networking theoretical system that is stable for civilian use, transiently resistant to damage, decentralized, has no routing core, requires the fewest satellites, has extremely high anti-interference capabilities, and millisecond-level self-healing. This invention addresses these gaps and deficiencies in existing technologies. Summary of the Invention

[0014] To address the combined shortcomings of existing space-based communication systems—relying on dedicated space-based routing satellites for centralized forwarding, the ease with which a single point of failure can cause a complete network outage, the inability of existing satellites in orbit to gain self-healing networking capabilities through software upgrades, excessively long fault recovery delays that cannot adapt to wartime millisecond-level countermeasures windows, insufficient anti-radio frequency interference performance, high construction costs, and difficulty in revitalizing existing resources—this invention provides a three-mode, four-level, nine-topology, decentralized, dynamic self-healing space-based satellite anti-interference networking system to achieve the following technical objectives: First, completely eliminating centralized routing nodes; four peer satellites can constitute the smallest autonomous networking unit, and any single point of failure will not cause a complete network outage; Second, through remote source code injection, existing satellites in orbit can acquire distributed routing forwarding and local mathematical calculation capabilities without hardware modifications; Third, compressing the communication self-healing time from the seconds to minutes of existing schemes to less than one hundred milliseconds, adapting to wartime millisecond-level countermeasures windows; Fourth, using a single-satellite nine-beam redundant configuration and Probabilistic model to achieve anti-interference gain The exponential improvement; fifth, replacing the traditional six-star routing scheme with a four-star scheme, thereby improving availability while reducing construction costs.

[0015] This invention aims to establish a complete mathematical model system with directly readable source code. It uses four peer satellites as the smallest autonomous networking unit, with each satellite configured with nine independent communication beams (three by three). The system remotely injects a networking program containing the complete mathematical model into the on-orbit satellites without modifying satellite hardware or launching additional routing satellites. This enables each on-orbit satellite to possess distributed routing and local mathematical computation capabilities. In the event of satellite loss of connection, link interference, or satellite damage, the system can automatically eliminate faulty nodes, update the topology matrix, and regenerate closed-loop subnets, keeping the communication self-healing time within one hundred milliseconds. It is also compatible with existing space-based routing satellite networks, providing mutual backup and ensuring continuous and stable communication.

[0016] This invention overcomes the shortcomings of existing technologies and provides a three-mode, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system, including: a remote source code injection module, a four-level autonomous topology operation module, a 3×3 nine-beam probability calculation module, a three-mode switching control module, a topology matrix operation module, a reliability mathematical solution module, a millisecond-level self-healing reconstruction module, and a space-based routing compatibility backup module.

[0017] The core functions, such as network program injection, topology autonomous operation, anti-interference probability calculation, mode switching control, topology matrix maintenance, reliability mathematical calculation, self-healing reconfiguration execution, and external system compatibility backup, are undertaken by independent modules. Each module performs its own function while cooperating with each other to ensure that the entire system can operate autonomously in orbit without relying on real-time ground intervention. This ensures that the space-based communication network can remain continuously available even under extreme conditions such as wartime or when ground stations are cut off.

[0018] The remote source code injection module is used to remotely inject the networking program, which encapsulates the complete mathematical model, into existing in-orbit satellites without modifying satellite hardware or launching dedicated space routing satellites. This enables each satellite to possess distributed routing and forwarding capabilities as well as local mathematical calculation capabilities. By upgrading the functionality of existing satellites already in orbit through pure software, self-healing networking capabilities can be achieved without waiting for the launch of new satellites, significantly shortening the capability generation cycle and reducing construction costs.

[0019] The remote source code injection module also supports injecting networking logic roles into heterogeneous satellites in orbit. When a topological pole vacancy occurs in the network, it wakes up compliant heterogeneous satellites in low-power standby mode, injects networking role parameters and topological constraints into them, and makes them new poles to fill the fourth-level unit and participate in the nine-level topology expansion. This maximizes the utilization of available resources on existing heterogeneous platforms in orbit, enabling different types of satellites to network collaboratively under a unified topology framework, further raising the threshold for the enemy to completely paralyze the space-based network.

[0020] The four-level autonomous topology operation module uses four satellites as the smallest autonomous unit. The four satellite nodes are completely peer-to-peer and there is no fixed routing center. The entire system has no fixed master control satellite, and the network management logic can be dynamically migrated among surviving satellites. The loss of a single satellite will not cause the entire network to be paralyzed. The centralized nodes are completely eliminated at the underlying topology level, so that there are no vulnerable nodes in the entire space-based network that can be accurately located and eliminated by the enemy. The loss of any satellite will not affect the overall network connectivity.

[0021] The four-level autonomous topology operation module sets a virtual center constraint point at the geometric center of the four satellites. The four satellites form a quadrilateral topological unit with this virtual center constraint point as the geometric anchor point. Within the unit, each satellite node performs relative position drift and coordinate adaptive adjustment around the virtual center constraint point to maintain the overall stability of the unit's geometric configuration. This provides a common geometric reference for the four high-speed moving satellites, allowing each node to perform small-range position drifts to adapt to relative position changes caused by orbital motion while maintaining overall configuration stability. This achieves a flexible topology effect of large-structure self-stabilization and small-range adaptation.

[0022] The quadrilateral topology unit is a software-defined flexible logical topology, not a physically fixed structure. The logical topology roles carried by each physical satellite node within the unit are decoupled from the physical satellite entities. If any physical satellite fails, its logical topology role can be migrated to another on-orbit satellite to take over, maintaining the logical integrity of the quadrilateral topology unit. In traditional fixed topology schemes, satellite numbers are bound one-to-one with topology locations; once the satellite corresponding to a number is damaged, the topology location is permanently lost. The decoupling scheme allows logical roles to migrate freely within the network, and vacant topology locations can be filled by any on-orbit satellite, significantly improving the network's structural fault tolerance.

[0023] The topological coordinates of each satellite node within the unit drift and dynamically adapt in real time with the relative motion between satellites, making it compatible with satellite node networking across all orbital altitudes and speed ranges (high, medium, and low Earth orbits). This allows the system to incorporate heterogeneous satellites with different orbital altitudes and operating speeds into the same network architecture, eliminating the need to design different networking schemes for different orbital types, and demonstrating excellent versatility and scalability.

[0024] The four satellite nodes of the quadrilateral topology unit are fully connected, naturally containing four sets of independent three-point rigid triangular subnets. When any one node fails, the remaining three nodes automatically form one of the three-point rigid triangular subnets to maintain communication within the unit. Utilizing the geometric property that the four-point fully connected unit naturally contains four sets of triangular subnets, and considering that a triangle is the most stable geometric structure in space, any three points in the four-point unit naturally form a rigid closed triangle. The four sets of triangles overlap and are redundant. Therefore, after any point fails, there is no need to generate a new topology; simply switching to an existing triangular subnet is sufficient for self-healing. This mechanism ensures millisecond-level latency for self-healing reconstruction.

[0025] The four-level autonomous topology operation module incorporates a real-time link quality assessment subunit. It uses the signal-to-noise ratio (SNR) and bit error rate (BER) of real-time inter-satellite electromagnetic wave transmission as priority criteria for link establishment and partner selection, replacing the pre-configured static link table on the ground. The space electromagnetic environment is complex and variable, and the pre-calculated link table on the ground cannot reflect the true channel status in real time. Using SNR and BER as real-time criteria ensures that the inter-satellite links established at any given time are the highest quality available communication paths, avoiding data routing to links that have been interfered with or degraded in quality.

[0026] Each satellite node autonomously and in a distributed manner completes the dynamic selection of networking partners and the construction of topology units. This decentralizes the decision-making power for networking from centralized ground control to the onboard processor of each satellite, completely eliminating dependence on ground stations and telemetry and control networks. Even if ground facilities are cut off or attacked, the space-based network can still autonomously complete networking and reconstruction.

[0027] The four-level autonomous topology operation module has built-in unit availability determination rules: the unit availability condition is that the number of intact satellites is ≥3, and the unit failure condition is that the number of intact satellites is ≤2. It clearly defines the minimum hardware conditions for maintaining communication connectivity among the four units, meaning that at least three of the four satellites are available to maintain communication. This is equivalent to allowing any single point of failure without affecting network functionality, reducing the system's reliability requirement for a single satellite from 100% to 75%, significantly improving the system's fault tolerance.

[0028] The 3×3 nine-beam probability calculation module corresponds to nine independent communication beams, three groups each along the X, Y, and Z axes of a single satellite configuration. These nine independent communication beams are software-defined logical beams, dynamically allocated and reconfigured at the software level by an onboard beam scheduling algorithm, without requiring nine additional physical hardware connections to the satellite. The nine beams are independent of each other; the failure of any one beam does not affect the operation of the remaining beams. This configuration increases the number of communication links per satellite from three in the traditional scheme to nine, tripling the link redundancy. With three groups of nine beams deployed along each of the three axes in three-dimensional space, communication can be maintained even if any axis is blocked or interfered with.

[0029] Built-in anti-interference mathematical model: The failure probability when all nine beams of a single satellite are suppressed is... The overall usability probability of a single star is ,in The probability of failure due to single-beam electromagnetic interference. This method calculates the probability of physical destruction of a single satellite. It transforms the anti-jamming performance of a single satellite from intuitive judgment into a quantifiable mathematical expression, providing a precise probabilistic basis for subsequent beam scheduling, link switching, and self-healing triggering. It also provides a unified mathematical benchmark for quantitative comparison with existing solutions.

[0030] The three-mode switching control module is configured with three operating modes: static civilian fixed network mode, dynamic instantaneous free networking mode, and wartime centerless millisecond self-healing mode. Civilian communication seeks long-term stable and low-jitter service quality, dynamic networking seeks adaptation to orbital motion, and wartime confrontation seeks survival and rapid recovery under extreme conditions. These three requirements are difficult to meet simultaneously with a single mode in the same architecture. Through mode switching, the same hardware payload and the same software algorithm can operate in the optimal way in the three scenarios, achieving military-civilian compatibility and peacetime-wartime integration.

[0031] The topology matrix operation module has three built-in adjacency matrices that can be implemented programmatically: a static four-star fully connected matrix and a dynamic time-varying topology matrix. A three-point closed-loop reconstruction matrix for fault dimensionality reduction is implemented; matrix elements are automatically updated and failed nodes are removed when a fault occurs. The adjacency matrix is ​​a standard mathematical tool in graph theory for describing network topology. After converting the network topology into a programmable two-dimensional array, all topology maintenance, node addition / deletion, and path search operations can be completed through standard matrix operations, facilitating efficient implementation in spaceborne embedded systems. A static matrix is ​​used for civilian steady-state locking scenarios, a dynamic matrix for orbital adaptation scenarios, and a fault matrix for self-healing reconstruction scenarios; all three cover all operational conditions.

[0032] The four-level units can be dynamically spliced ​​together. Any two or more four-level units can be interconnected through nine-level fission branches between poles to form 6-star, 8-star, 9-star, and larger-scale cluster networks. A 6-star cluster is formed by two four-level units sharing a pair of interconnection branches; an 8-star cluster is formed by two four-level units spliced ​​through two sets of parallel branches; and a 9-star cluster is formed by three four-level units connected end-to-end in a ring interconnection. The splicing scale is unlimited and can be continuously expanded according to task requirements. Using four-star units as gene blocks, the network expands outward through standardized pole interconnection interfaces to form cluster networks of any size. This gene fission-like growth mechanism ensures that the network size is not limited by the initial design and can be expanded at any time according to actual task needs, exhibiting strong flexibility and scalability.

[0033] In this invention, "nine-topology extension" means: using the four satellite poles of a level-four autonomous unit as topology extension anchor points, each pole extends outward along three different spatial directions with three sets of interconnected branches, for a total of twelve sets of branches. After deducting three redundant branches that overlap with the existing four-point fully connected topology within the unit, nine effective new topology extension directions are formed, hence the name "nine-topology fission." This "nine-topology extension" mechanism ensures that each level-four unit has at least nine independent topology extension paths when expanding outward, providing sufficient redundant interfaces for constructing a full-network closed topology. It should be noted that "nine-topology extension" specifically refers to the topology extension direction and has no direct correspondence with the number of "nine beams" configured on a single satellite. Nine beams represent redundant design for single-satellite communication links, while nine-topology extension represents redundant design for the entire network topology extension.

[0034] The reliability mathematical calculation module incorporates a closed-form formula for the availability of four-star units. Meanwhile, the average availability of the built-in system Cost-effectiveness model Anti-interference gain ,in The availability calculation for the four-point unit is simplified from a complex calculation involving enumerating four failure scenarios to a single mathematical expression, which can be directly calculated in real time on the onboard processor. This provides a real-time availability assessment basis for mode switching, self-healing triggering, and beam scheduling. Considering that the instantaneous availability fluctuates over time due to the high-speed motion of satellites in orbit, the availability value at a single moment is insufficient to assess the long-term stability of the network. By integrating and averaging the instantaneous availability within a time window, the average availability of the system over a longer time scale can be obtained, which can be used for network status assessment and scheduling decisions in dynamic modes. Combining availability with the number of satellites evaluates the resource utilization efficiency of the scheme, providing a unified quantitative indicator of cost-effectiveness for horizontal comparison of different networking schemes. The improvement factor in anti-interference capability of the nine-beam architecture compared to the traditional three-beam architecture is directly quantified through exponential calculation, providing a concise and convincing quantitative conclusion for system design review, scheme demonstration, and performance indicator reporting.

[0035] The millisecond-level self-healing reconfiguration module incorporates segmented triggering logic: when the number of faulty nodes is ≤1, a three-point closed-loop reconfiguration is performed only within the current Level 4 autonomous unit, without triggering external link adjustments; when the number of faulty nodes is ≥2, a cross-unit, network-wide secondary reconfiguration is initiated. Most satellite faults are single-point faults. Limiting the recovery of single-point faults to within their respective units avoids widespread signaling flooding during recovery, significantly reducing network-wide signaling overhead. Minimizing the self-healing scope also means maximizing self-healing speed. When the fault scale exceeds the fault tolerance limit of a single unit, cooperation from multiple surrounding units is required to complete topology repair. In this case, sacrificing some recovery speed for overall network stability is necessary.

[0036] The millisecond-level self-healing reconstruction module has built-in timing constraints: self-healing convergence delay. With a recovery time of ≤100ms, all fault identification, matrix recalculation, and link reconstruction operations are completed locally on-board, requiring no ground command intervention. 100ms is on the order of one-hundredth of the characteristic time of satellite orbital motion; within this timescale, the satellite's relative position changes minimally, and the topology after self-healing remains physically valid and stable. This on-board completion, eliminating ground intervention, completely removes the propagation delays of fault reporting, ground calculation, and command uploading inherent in traditional methods, accelerating recovery speed from seconds to milliseconds and meeting the timing requirements of wartime operations.

[0037] The space-based routing compatible backup module is deployed in parallel with the hardware space routing network. When the routing satellite fails, this system independently undertakes the communication task, forming a dual communication guarantee. This system does not replace or compete with the existing national space-based routing satellite construction plan, but coexists as a parallel backup system. When the hardware routing satellite is working normally, the two systems operate collaboratively and complement each other; when the hardware routing satellite is interfered with or damaged, this system can independently undertake all communication tasks, ensuring that space-based communication is not interrupted due to a single path being cut off, forming a dual security defense line for my country's space-based communication.

[0038] The system constructs a spherical closed topology through a four-level, nine-stage topological splitting process, eliminating weak points at the network ends, as is common in linear or strip topologies. Every node is located on a closed loop, enabling multi-point redundancy and multi-path routing. Traditional linear or strip topologies have terminal nodes; if a terminal node is damaged, the connection between all downstream nodes and the network core is severed. In a spherical closed topology, there are no terminal nodes. Even if any node is damaged, information can be relayed indirectly through the other direction of the closed loop. This topologically guarantees network connectivity under any single or multiple point damage conditions, fundamentally eliminating the structural defect where a broken node leads to the breakdown of a chain of nodes.

[0039] Furthermore, the reliability mathematical solution module also includes a traditional hardware networking comparison model: Dual-route satellite failure probability ; Traditional three-beam failure probability structure ; Traditional 6-star network availability ; Availability metrics used for real-time comparison of two networking schemes.

[0040] in, The probability of a single satellite being physically destroyed is given by the given value. The probability of both routing satellites being completely destroyed is the product of their individual destruction probabilities. .in, Let be the probability of single-beam electromagnetic interference failure. The probability of all three beams failing is the product of their individual failure probabilities, i.e. Availability of traditional 6-star networking middle, The probability that at least one of the routing system's components is intact. The probability of a single operational satellite being available is given by its total availability. The fourth power of this probability is the probability that all four operational satellites are available.

[0041] During system operation, the availability difference between this solution and the traditional solution is calculated in real time, and the performance gain of this solution is presented to ground monitoring terminals and decision-makers in a visual manner, while providing a performance benchmark for the system itself. When the availability of this solution is lower than that of the traditional solution, it may only occur under extreme parameter conditions, and the system can automatically issue an alarm to avoid blindly pursuing advantages while ignoring boundary conditions.

[0042] Furthermore, in the topology matrix operation module, the static four-star fully connected adjacency matrix... for: ,

[0043] The topological relationships between the four satellites are fully connected using an array of zeros and ones. Zeros on the diagonal indicate that a node does not establish a link with itself, while one off-diagonal indicates that there is a direct communication link between any two nodes. This fully connected matrix is ​​a mathematical expression for a fixed topology in a civilian static model, where the matrix elements do not change over time, maintaining the absolute stability of the topology.

[0044] When any node fails, a three-point closed-loop reconstruction matrix is ​​automatically generated. for: ,

[0045] A 3x3 array of zeros and one-values ​​describes the closed-loop connection relationship of the three remaining nodes after the failure node is removed. The three nodes are fully connected in pairs, forming a rigid triangle, which is the most stable and simplest communication topology in space. The operation of reducing the dimension from a 4x4 matrix to a 3x3 matrix corresponds to the matrix operation of removing the columns and rows of the failure node. All operations can be completed by the onboard processor in microseconds.

[0046] Matrix elements in dynamic mode Updated in real time according to the satellite's orbital position; 1 represents a connected link, and 0 represents a disconnected link. ,in , The time-varying topology is precisely described by mathematical functions, and the topological state at any given time can be given a deterministic mathematical expression by this matrix, avoiding the fuzzy description of unpredictable dynamic changes in the topology.

[0047] Furthermore, the 3×3 nine-beam probability calculation module supports real-time calculation of the single-satellite communication silence probability under different interference intensities, compared to the traditional 3-beam architecture. The anti-interference effect is improved by multiples through exponential difference quantification. The output evaluation results are used for beam scheduling strategy adjustment. When taking any value between zero and one All values ​​are greater than one, proving that the anti-interference capability of this scheme is superior to the traditional 3-beam scheme under any interference intensity. The stronger the interference, the better. The closer to one, the smaller the gain factor, but it is always greater than one; the weaker the interference, the smaller the gain factor. The closer it gets to zero, the more the gain increases dramatically.

[0048] Furthermore, the reliability mathematical solution module incorporates a binomial distribution failure probability derivation formula: Four-star unit independent damage failure probability ; in This represents the probability of physical destruction of a single satellite. A four-satellite unit fails if and only if the number of intact satellites is ≤2, i.e., exactly two, exactly three, or exactly four satellites are destroyed. The probability of exactly two satellites being destroyed is the number of combinations. When unfolded, it becomes The probability of exactly three stones being damaged is the number of combinations. When unfolded, it becomes The probability of exactly four stones being damaged is the number of combinations. ,equal The sum of the three terms, combined with like terms, yields the result. .

[0049] Overlay cluster damage corrections , , Obtain the corrected failure probability In a real battlefield environment, the enemy's weapons of mass destruction may cause multiple satellites to be destroyed simultaneously. In this case, the failure of each satellite is no longer an independent event, and a simple binomial distribution will underestimate the probability of multiple satellites failing at the same time. , , As an adjustable correction parameter filled in after engineering measurement, it is used to correct the correlation deviation caused by cluster damage, so that the mathematical expectation of failure probability is closer to the statistical value in real combat environment.

[0050] Cluster damage correction item , , The calibration method can adopt the following exemplary process: By constructing a Monte Carlo simulation model that includes the spatial distribution of multiple satellites, weapon damage radius, and explosion timing, a large number of damage scenarios are randomly sampled and generated. The frequency of simultaneous damage to exactly two, three, and four satellites is counted, and each is divided by the total number of scenarios to obtain the empirical probability. Then, the probability corresponding to the independent binomial distribution model is subtracted to obtain the final probability. , , The calibration value can be determined based on the statistical frequency of simultaneous multi-satellite destruction events in historical battles. When measured data is lacking, , , The default value can be zero, in which case the correction failure probability degenerates into an independent binomial distribution model. The system can still operate normally.

[0051] Unit availability The unit failure probability is converted in reverse to the availability probability, as mentioned above. This forms a comparative expression that describes the same physical quantity from different perspectives. Describing from the available positive perspectives, From the perspective of the unusable opposite, the two are numerically complementary and mutually verify the consistency of the mathematical derivation.

[0052] All formulas are encapsulated as callable functions and written into the source code, so that the engineering implementation of the mathematical model does not require manual conversion between theoretical documents and on-orbit code, ensuring the consistency between theoretical derivation and engineering implementation.

[0053] Furthermore, in the static civilian fixed network mode, the topology matrix remains fixed, and the system continuously outputs steady-state availability. To maintain stable communication with low jitter, dynamic integration calculations are not performed. Civilian communication services such as broadcast television relay, broadband internet access, and navigation enhancement signal broadcasting have strict requirements for long-term stability and low latency jitter in communication links; the additional jitter introduced by frequent topology changes is unacceptable. In this mode, the system sacrifices the flexibility of dynamic adaptation in exchange for the same level of stability as existing traditional fixed networks, ensuring that civilian services are not affected by dynamic topology adjustments.

[0054] Furthermore, in the dynamic instantaneous free networking mode, the system periodically integrates to calculate the average availability within a certain time window. The system uses real-time scheduling of nine beams to perform link splitting and reconfiguration, adapting to topology changes caused by satellite orbital motion. In a normal operating environment where high-speed satellite motion leads to continuous topology changes, the availability value at any single moment cannot reflect the overall network performance over a longer period. By calculating the time-averaged availability using a sliding window integration method, the interference of instantaneous fluctuations on scheduling decisions is eliminated, ensuring that beam scheduling and link reconfiguration decisions are based on an accurate assessment of the network's long-term trends.

[0055] Furthermore, in the wartime decentralized millisecond self-healing mode, the system enforces timing constraints. ,in The characteristic time constant of satellite orbital topology deformation is 100ms. Ground-based centralized scheduling is disabled, and fault diagnosis, matrix refresh, and subnet reconstruction are performed independently on-board. The characteristic time constant of satellite orbital motion is on the order of tens of seconds, while 100ms is only one-thousandth to one-hundredth of this time constant. Within the 100ms time window, the relative position change of the satellite is minimal and negligible relative to the inter-satellite distance. Therefore, the geometric conditions of the rigid triangular subnet formed by the remaining three points after self-healing remain stable before and after self-healing, preventing invalid self-healing where the topology is completely distorted due to orbital motion by the time reconstruction is completed. Disabling ground-based centralized scheduling aims to completely sever the dependence on ground facilities, enabling the space-based network to maintain independent operation even when the land-based telemetry and control network is cut off or suppressed.

[0056] Furthermore, the remote source code injection module encapsulates all matrix operation functions, probability calculation formulas, integral solving functions, and modality judgment branch code. After the source code is remotely distributed, the satellite embedded system directly calls the entire mathematical model to complete real-time networking operations without requiring additional hardware upgrades. All mathematical derivations, topology rules, and modality switching logic are pre-encapsulated into a standardized function library. On-orbit satellites only need to execute function calls to complete complex matrix updates, probability calculations, and integral solving, without needing to rewrite or compile the code in the onboard embedded environment. Since all mathematical operations are performed in the existing function library, the satellite's processor load and storage resource overhead are kept to a minimum, allowing even existing satellites with limited computing power to run the system completely.

[0057] Furthermore, the system calculates the cost-effectiveness index of the two solutions in real time: Cost-effectiveness of this solution ; Cost-effectiveness of traditional routing solutions ; Automatic output gain difference ; ; As output parameters for network status assessment, they support the software's visual monitoring functions. Traditional solutions use six satellites to achieve availability. This solution uses 4 satellites to achieve availability. The cost-effectiveness metric divides availability by the number of satellites, reflecting the resource utilization efficiency of how much availability each satellite contributes. and This directly quantifies the gains of this scheme compared to traditional schemes in terms of availability and resource efficiency, providing an intuitive quantitative basis for the status assessment of space-based networks in ground command and control centers.

[0058] Furthermore, it also includes a millisecond-level predictive seamless switching submodule: Based on the relative motion speed between stars , orbital deviation rate Real-time inter-satellite distance and signal propagation delay Dynamically calculate the remaining visible time window of the current link. ; Before the current link leaves line of sight, initiate a full-domain scan and pre-matching of neighboring satellites in advance, complete the pre-handshake of the new link, and achieve a seamless transition with zero interruption between the disconnection of the old link and the establishment of the new link.

[0059] The function The explicit form of expression is: ,

[0060] in The maximum inter-satellite communication distance is determined by the satellite's transmit power, antenna gain, receiver sensitivity, and space propagation loss. To preset the maximum visible time window, it can be half of the orbital period or a larger fixed value; This term is introduced to account for signal propagation delay, reserving margin for signal transmission and processing. In engineering implementation, this function can also be used for refined modeling based on actual orbital parameters and link budget, including considering the effects of factors such as Earth's occlusion and atmospheric attenuation.

[0061] Traditional satellite communication schemes employ a delayed model of searching for a new link after a link is lost. There is a gap between the link loss and the establishment of a new link, involving a search, matching, and handshaking process, during which communication is interrupted. This scheme, by predicting the remaining lifespan of the old link while it is still operational and establishing a backup link in advance, transforms the serial process of link loss, search, and reconstruction into a parallel process of pre-search, pre-handshaking, and seamless handover, eliminating the communication interruption window during the handover process.

[0062] Furthermore, the specific method by which the remote source code injection module performs network logic role injection on in-orbit heterogeneous satellites is as follows: when the number of intact satellites in a Level 4 autonomous unit drops to 2 or less, or when a topology pole vacancy is detected in the entire network, the system automatically selects in-orbit heterogeneous satellites that have unified public networking frequency bands and have radio frequency and laser full-duplex transceiver capabilities, and injects them with a network logic package containing topology role parameters, adjacency constraints, and availability calculation functions, so that the heterogeneous satellite is added to the damaged unit as a new pole, and the topology matrix and nine-topology fission branch relationship are updated simultaneously.

[0063] There are numerous satellite platforms of different types, such as communication satellites, remote sensing satellites, and navigation satellites, in orbit. However, most of the time they only perform their primary tasks, with their communication modules idle or under low load. Through logical role injection, the idle communication capabilities of these heterogeneous platforms can be borrowed to the network architecture, expanding the network resource pool from dedicated communication satellites to all in-orbit platforms with basic communication capabilities, maximizing the utilization of existing in-orbit assets.

[0064] Furthermore, the remote source code injection module, the four-level autonomous topology calculation module, the 3×3 nine-beam probability calculation module, the three-mode switching control module, the topology matrix calculation module, the reliability mathematical solution module, the millisecond-level self-healing reconfiguration module, and the space-based routing compatible backup module are connected by a decoupled software interface. Each module has an independent version identifier and supports independent on-orbit upgrades and replacements without affecting the operation of other modules. The modules transmit topology matrices, availability indicators, mode switching commands, and link status parameters through a standardized data exchange format.

[0065] Once a spacecraft system is operational in orbit, major software upgrades across the entire satellite are difficult to perform. In traditional monolithic software architectures, fixing defects in any module requires recompiling and re-implementing all code, resulting in extremely high risks and costs. Decoupled architectures allow for the replacement of only the problematic module in orbit, while other modules remain unchanged, significantly reducing the risks and costs of on-orbit software maintenance. Standardized data exchange formats between modules ensure that even if the interface format changes after an upgrade of one module, as long as its output conforms to the standardized format, the remaining modules can continue to function normally without any modifications.

[0066] Furthermore, the four-level nine-topology fission expansion is based on the four poles of each four-level autonomous unit, and then expands outward to form three new branches, nested layer by layer to form a spherical closed network topology; the spherical closed topology has no strip-shaped end nodes, and there are at least two non-overlapping communication paths between any two nodes in the network.

[0067] A spherical closed topology provides at least two physically distinct information transmission paths between any two communication nodes. When an intermediate node on one path fails, the information can automatically reroute to the destination node via another path without recalculating the entire route. There are no topologically significant terminal nodes in a spherical closed topology; each node is on a closed loop and has the same topological status and resilience.

[0068] Furthermore, in the dynamic instantaneous free networking mode, all inter-satellite links are transient temporary links. Each satellite node dynamically establishes communication connections through instantaneous scanning, instantaneous rendezvous identification, and instantaneous handshake link establishment. The link is instantly disconnected after a satellite node leaves line of sight. The entire network relies on the continuous dynamic link establishment, dynamic link disconnection, and dynamic iteration of each node to maintain communication connectivity, without relying on a fixed link table pre-configured on the ground.

[0069] In traditional fixed link tables, links are pre-planned and locked for a long time. Once a link is interrupted due to orbital motion or satellite failure, a new link table needs to be planned and added on the ground, resulting in a long response cycle and poor flexibility. In contrast, in transient link mode, the lifecycle of each link is entirely determined autonomously by the satellite. A link is established when an available neighboring satellite is detected, and the link is disconnected when the satellite moves out of line of sight. The entire link establishment and disconnection process can be completed within milliseconds. Combined with a predictive switching mechanism, continuous dynamic refreshing of the entire network of links can be achieved without creating communication gaps.

[0070] Furthermore, the dynamic splicing method between the four-level units is as follows: the four poles of each four-level unit extend outward along the nine-topology splitting direction to form three sets of interconnected branches, and adjacent units are spliced ​​through the interconnection of branches between poles; in the cluster network formed after splicing, there are at least two non-overlapping interconnected paths between any two four-level units.

[0071] When each fourth-level unit expands outward, it is not simply attached to the outside of the parent unit. Instead, it extends three sets of branches from each of the four poles of the parent unit, which cross-interconnect with the corresponding branches of adjacent units. This multi-point interconnection method ensures that the interface between units also has redundancy characteristics. If a single interconnection branch fails, the other branches can still maintain communication between units.

[0072] Furthermore, under normal circumstances, the networking communication modules of heterogeneous satellites in orbit are in a low-power sleep standby state and do not participate in network topology maintenance. When the Level 4 autonomous topology operation module detects a topology pole gap or the number of intact satellites in the unit drops to 2 or less, the system sends a wake-up command to the heterogeneous satellites in orbit that have a unified public networking frequency band and are capable of radio frequency and laser full-duplex transceiver. After the heterogeneous satellites are woken up, the remote source code injection module injects networking logic roles into them and incorporates them into the entire network topology.

[0073] The energy budget for satellites in orbit is extremely limited, with strict constraints on solar panel area, battery capacity, and heat dissipation capabilities. Keeping the communication modules of all heterogeneous satellites constantly active would consume a large amount of precious on-orbit energy, shortening the satellites' service life. A normal hibernation design ensures that the networking communication modules of heterogeneous satellites do not consume energy most of the time, only being activated and deployed for networking when necessary, achieving the engineering goals of on-demand scheduling and energy-efficient operation.

[0074] Furthermore, the mathematical criterion for the three-point rigid triangular subnet is: for any three nodes in a quadrilateral topological unit... , , If the link states between any two of them satisfy Then the three points form a rigid closed triangular subnet.

[0075] The geometric concept of three points forming a rigid closed triangular subnet is transformed into a programmable mathematical judgment expression. If every pair of the three nodes is connected, a closed triangle is formed; if any edge is missing, a zero appears, and the triangle is not closed and does not form a rigid subnet.

[0076] In a quadrilateral fully connected topology, the following four sets of independent three-point rigid triangular subnets naturally exist: , , , It clearly demonstrates that any three nodes out of the four nodes form a rigid closed triangle, and the four triangles are independent of each other. A failure in any one triangle does not affect the availability of the other three triangles.

[0077] When any pole fails, the system automatically switches to the corresponding rigid closed triangular subnet among the remaining three poles, maintaining uninterrupted communication within the cell. The purpose of this switching logic is to: In the case of failure, the system automatically switches to This set of triangular subnets. This switching requires no computational generation because... , , Links between any two nodes naturally exist in a fully connected topology. The system only needs to execute "disable failed nodes". The matrix marking operation automatically forms a closed triangle with the remaining three points connected. The switching speed is only limited by the matrix update time, which is much less than the timing constraint of 100ms.

[0078] The beneficial effects of this invention compared to the prior art are: 1. This invention employs a complete eight-module system architecture comprised of a remote source code injection module, a four-level autonomous topology operation module, a 3×3 nine-beam probability calculation module, a three-mode switching control module, a topology matrix operation module, a reliability mathematical solution module, a millisecond-level self-healing reconfiguration module, and a space-based routing compatible backup module. Combined with a four-pole quadrilateral flexible embedded topology unit, virtual center constraint point geometric anchoring, complete decoupling of logical topology roles and physical satellite entities, and a real-time coordinate drift adaptive mechanism, this invention completely eliminates centralized nodes at the underlying architecture level. The decentralized nature of the entire network prevents the enemy from identifying and locking onto critical targets that could lead to network paralysis. When any single satellite is damaged, its logical role can be migrated to other satellites for replacement. If any point in a four-point unit fails, the remaining three points automatically form a rigid closed-loop triangular subnet, ensuring that the unit does not collapse and the link is not interrupted. This topology design provides a level of resilience not found in traditional three-star rigid formations and random mesh networking schemes; it represents a generational upgrade at the architectural level rather than a gradual optimization at the parameter level.

[0079] 2. This invention replaces the fixed mode of the traditional scheme with a static link table pre-configured on the ground by using the signal-to-noise ratio and bit error rate of real-time inter-satellite electromagnetic wave transmission as the priority criteria for link establishment and partner selection. Each satellite node autonomously and in a distributed manner completes the dynamic selection of networking partners and the construction of topology units. This dynamic link establishment mechanism decentralizes the network decision-making power from centralized ground control to the onboard processor of each satellite. The space-based network can still maintain independent operation even under extreme conditions where the ground-based telemetry and control network is cut off or suppressed, completely eliminating dependence on ground stations. At the same time, using channel quality as the priority criterion ensures that the inter-satellite links established at each moment are the highest quality communication paths currently available, avoiding routing data to links that have been interfered with or degraded in quality. This represents a fundamental shift from preset links to real-time optimal selection in terms of anti-interference capability and communication quality assurance.

[0080] 3. This invention establishes nine independent communication beams (three groups each along the X, Y, and Z axes) on a single satellite and sets up... Nine-beam total failure probability model and Based on the single-satellite integrated availability probability model, this invention increases the number of communication links per satellite from three in the traditional scheme to nine. It is important to emphasize that these nine backbone links are all software-defined logical links, dynamically allocated and reconfigured at the software level by the onboard beam scheduling algorithm; they are not new physical hardware connections, and existing RF front-ends and antenna arrays can still be reused at the hardware level. Compared to the traditional three-beam scheme… The failure probability of the nine-beam scheme is reduced, and the anti-interference gain reaches [a certain value]. Times. With Taking a value of 0.3 as an example, the failure probability of a traditional three-beam scheme is... =0.027, the failure probability of the nine-beam scheme in this proposal. = =0.000019683, Anti-interference gain =0.027 / 0.000019683≈1371 times. This gain is in The value is always greater than one in the interval (0,1), and it increases with each other. Decrease and then increase rapidly. When When K = 0.1, K = 10 6 =1,000,000 times. This improvement in anti-interference capability is exponential, not a gradual improvement on a percentage level, and its mathematical proof is in It holds strictly within the range of greater than zero and less than one, possessing irrefutable quantitative persuasiveness.

[0081] 4. This invention utilizes a built-in... The closed-form formula for the availability of four-point units simplifies the calculation of four-point unit availability from a complex operation involving enumerating four failure scenarios to a single mathematical expression that can be directly calculated in real time on the onboard processor. This closed-form solution is strictly equivalent to the unit availability determination rule that the number of intact satellites is greater than or equal to three, mathematically proving that the reliability of four-point units is far superior to the physical intuition of three-point units. In a three-point unit, the network is interrupted when any point fails, while in a four-point unit, the failure of any point does not interrupt communication, significantly improving availability. Under equal adversarial conditions... =0.02、 In the numerical example with a value of 0.3, the availability of the traditional six-star routing scheme is... ≈0.883, while the availability of the four-point unit in this scheme is... The efficiency is approximately 0.9986, representing an improvement in availability of about 11.56%. The derivation of this formula is complete and without skipping steps, and can be directly compiled into source code that runs on orbit, achieving a seamless transformation from mathematical theory to engineering code.

[0082] 5. This invention employs a segmented triggering logic: when the number of faulty nodes is ≤1, a three-point closed-loop reconstruction is performed only within the unit without triggering network-wide signaling; when the number of faulty nodes is ≥2, a cross-unit network-wide secondary reconstruction is initiated. This achieves a layered self-healing mechanism that prioritizes self-healing within the unit and provides backup reconstruction for the entire network. In traditional solutions, any fault triggers a unified network-wide traversal and rerouting, resulting in slow recovery, signaling congestion, and a large impact. In this solution, most single-point faults are confined to their respective fourth-level units for self-healing, leaving other units unaffected. This significantly reduces signaling overhead and topology disturbance during the self-healing process. Furthermore, the three-point closed-loop reconstruction within the unit, due to its small topology scale and lack of negotiation requirement, allows for stable control of the recovery speed. With a recovery latency of ≤100ms, compared to the traditional solution's recovery latency of seconds to minutes, this represents an improvement of one to three orders of magnitude, fundamentally meeting the timing requirements of millisecond-level combat in wartime.

[0083] 6. This invention achieves the design goal of military-civilian compatibility and peacetime-wartime integration by configuring three switchable working modes—static civilian fixed network, dynamic instantaneous free networking, and wartime decentralized millisecond self-healing—within the same underlying architecture. In static mode... Communication stability is comparable to traditional fixed networks, meeting the stringent requirements of low jitter and high stability for long-term civilian communication; in dynamic mode, through The sliding window integral average availability assessment and nine-beam real-time scheduling and reconfiguration adapt to the time-varying topology environment caused by the high-speed motion of satellites in high, medium, and low orbits, achieving adaptive networking under normal maneuvering conditions; and mandatory execution under wartime mode. ≤100ms and 100ms is much less than By eliminating timing constraints and disabling centralized ground scheduling, the satellite independently performs all self-healing calculations, maintaining network survival under extreme adversarial conditions. All three modes share the same hardware payload and software algorithm, switching the optimal operating logic according to different scenarios. This eliminates the need to deploy multiple heterogeneous systems for different scenarios, significantly reducing system complexity and construction costs.

[0084] 7. This invention encapsulates all matrix operation functions, probability calculation formulas, integral solving functions, and modal judgment branch code into a standard function library and distributes it remotely via a remote source code injection module. This allows existing on-orbit satellites to acquire complete distributed routing and forwarding, local mathematical calculation, and millisecond-level self-healing networking capabilities through software upgrades without any hardware modifications. Compared to traditional solutions that rely on continuously launching new satellites to improve networking performance, this solution significantly reduces the cost and time required for system upgrades, maximizing the value of on-orbit satellites and demonstrating significant engineering economics.

[0085] 8. This invention uses a millisecond-level predictive seamless switching submodule to calculate the relative motion velocity between satellites in real time. , orbital deviation rate Real-time inter-satellite distance and signal propagation delay And dynamically calculate the remaining visible time window of the current link. Its explicit expression is ,

[0086] In engineering implementation, refined modeling can be performed based on actual orbital parameters and link budgets. Before the current link moves out of line of sight, a full-domain scan of neighboring satellites, pre-search and pre-matching, and a pre-handshake for the new link can be initiated in advance. This invention achieves seamless, zero-interruption transition between the disconnection of an old link and the establishment of a new link. Traditional solutions only begin searching for a new link after the old one is disconnected, resulting in a gap between link disconnection, search, matching, and handshake. This invention, by transforming the serial process into a parallel process, completely eliminates the communication interruption window during the handover process, solving the common problems of network-wide lag, disconnection, and paralysis after satellites in traditional constellations move out of line of sight.

[0087] 9. This invention achieves the goal of maintaining the long-term sustainability of a space-based network without relying on frequent launches of new satellites, solely through the dynamic scheduling of existing heterogeneous platforms in orbit. This is achieved by maintaining low-power sleep / standby states for heterogeneous satellites under normal circumstances, and automatically waking up compliant heterogeneous satellites and injecting them with networking logic roles only when a topological pole gap is detected or the number of intact satellites in a unit drops to two or less. The communication modules of heterogeneous satellites do not consume valuable on-orbit energy most of the time, and are only activated for networking when necessary, achieving a balance between energy efficiency and networking capability. This mechanism extends the network's schedulable resources from dedicated communication satellites to all on-orbit heterogeneous platforms with basic communication capabilities, maximizing the resource pool size and redundancy of the space-based network without increasing hardware requirements.

[0088] 10. This invention achieves modular and iterative engineering capabilities for the system architecture by connecting all eight modules through decoupled software interfaces. Each module has an independent version identifier and supports independent on-orbit upgrades and replacements without affecting the operation of other modules. After aerospace systems are operational in orbit, it is difficult to perform major version upgrades of the entire satellite's software. In traditional monolithic software architectures, fixing defects in any module requires recompiling and re-uploading all code, which is extremely risky and costly. The decoupled architecture allows for the replacement of only the problematic module, while other modules remain unchanged. This innovative engineering architecture enables the invention to be continuously iterated and optimized, constantly generating new technical solutions, throughout the satellite's on-orbit service life of several to several decades, possessing long-term vitality and scalability.

[0089] 11. This invention constructs a spherical closed topology through a four-level, nine-stage topological splitting process, eliminating weak points at the network ends (as in strip topologies). Every node is located on a closed loop, and there are at least two non-overlapping communication paths between any two nodes. This topological geometry ensures network connectivity even under single or multiple point damage conditions. Traditional linear or strip topologies have terminal nodes; if a terminal node is damaged, the connection between all downstream nodes and the network core is severed, creating a structural vulnerability. The spherical closed topology fundamentally eliminates this structural defect. After any node is damaged, information can be relayed indirectly through the other direction of the closed loop, automatically bypassing the original path without requiring recalculation. This elevates network resilience from mere recovery to a higher level of structurally guaranteed uninterrupted operation.

[0090] 12. This invention utilizes a cost-performance model. and traditional solutions Real-time comparison, and and The automatic output of gain difference provides an intuitive quantitative basis for assessing the state of space-based networks. Under equal adversarial conditions... =0.02、 In the numerical example with a value of 0.3, the availability of the traditional six-star routing scheme is... ≈0.883, cost-effectiveness ≈0.147; Availability of the four-point unit in this scheme ≈0.9986, cost-effectiveness The efficiency is approximately 0.2496, representing a 69.8% improvement in cost-effectiveness. These figures have been fully theoretically derived and mathematically proven, and can be verified through field measurements after engineering implementation. This solution achieves higher availability, stronger anti-interference capabilities, and faster self-healing speed with fewer satellites (four compared to six). Its overall performance is significantly superior to the traditional approach of adding new hardware routing, and it can serve as an optimized supplement to existing space-based routing satellite systems, raising the survivability ceiling of the entire national space-based communication system. Attached Figure Description

[0091] The present invention will now be further described with reference to the accompanying drawings.

[0092] Figure 1 This is a schematic diagram of the four-level four-star fully connected topology of the present invention.

[0093] Figure 2 This is a schematic diagram of the three-point closed-loop self-healing topology structure after a single-point failure according to the present invention.

[0094] Figure 3 This is a flowchart of the three-mode switching process of the present invention.

[0095] Figure 4This is a schematic diagram of the overall system framework of the present invention. Detailed Implementation

[0096] The following is in conjunction with the appendix Figures 1 to 4 Specific embodiments of the present invention will be described in detail below. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0097] As a prerequisite for the implementation of this invention, satellites participating in the network must possess basic inter-satellite communication transceiver hardware, including RF / laser transceiver front-ends, antenna systems, and signal processing units, and have duplex transmission, reception, and signal identification capabilities in both RF and laser modes. This invention protects the software algorithms and networking protocols running on the aforementioned hardware platform, not the satellite hardware itself. This hardware capability is possessed by most existing on-orbit communication satellites and some next-generation remote sensing and navigation satellites. For satellites lacking this hardware capability, they can work collaboratively with hardware-routing satellites through the space-based routing compatibility backup module of this invention; full-duplex capability is not mandatory for all satellites. Furthermore, various heterogeneous satellites require a ground-based aerospace control system to routinely pre-configure a unified public networking frequency band, uniformly adjusting the operating frequencies of each satellite to a common communication channel that allows for mutual recognition and handshaking. Before participating in networking, heterogeneous satellites must have their working frequency, channel parameters, networking key, and identity identifier uploaded by the ground station in advance. All satellites must be uniformly adjusted to a common networking frequency band that allows them to recognize and handshake with each other in order to meet the transient networking conditions. Satellites that have not been uploaded by the ground station in advance must not be awakened or given a networking role.

[0098] The working boundaries of the ground station are limited to: before the network deployment mission starts, parameters such as frequency, channel parameters, network key, and identity identifier are uploaded, as well as status monitoring and mission-level command issuance during operation; real-time network deployment operations such as topology reconstruction, matrix operation, beam scheduling, and self-healing decision-making are all completed locally on the satellite, and the ground station does not participate in real-time network deployment operations.

[0099] This system is applicable to a variety of typical operating conditions: Under normal on-orbit operation, the system provides low jitter and high stability communication in a static civilian fixed network mode; under conditions of local electromagnetic interference, the system uses a nine-beam probability calculation module to assess the single-satellite silence probability in real time and schedule beam avoidance; under conditions of single-point or a few-node damage, the system automatically completes three-point closed-loop self-healing reconstruction within the unit; under conditions of multi-point large-area damage or strong confrontation emergency, the system initiates cross-unit full-network secondary reconstruction and switches to wartime centerless millisecond self-healing mode; under extreme emergency conditions such as ground station disconnection or telemetry and control link failure, the system still relies on on-board local computing to independently maintain network operation.

[0100] Example 1: Implementation of Static Civil Networking Mode like Figure 1As shown, in static civilian fixed network mode, the system injects the networking program, which encapsulates all mathematical models and topology matrices, into four in-orbit satellites via a remote source code injection module. The remote source code injection module first authenticates the target satellites and assesses channel quality. After confirming that the satellites have both RF and laser full-duplex transceiver capabilities and have been unified to a common networking frequency band, it fragments the networking program package, containing a complete matrix operation function library, probability calculation formulas, integral solving functions, and modal judgment branch code, and distributes it to the embedded storage area of ​​each satellite. After receiving and verifying the complete program package, the satellites load it into their running memory and start the four-level autonomous topology operation module.

[0101] The Level 4 Autonomous Topology Operation Module uses four satellites , , , A quadrilateral topology is constructed for the smallest autonomous unit. This module establishes a virtual center constraint point at the geometric center of the four satellites. Each satellite acquires distance data from the others through two-way inter-satellite ranging. The arithmetic mean of the four coordinates is used as the spatial position of the virtual center point, which is continuously updated during subsequent operation. The four satellite nodes use this virtual center constraint point as their geometric anchoring reference, forming a rigid closed triangular subnet of three points to anchor the overall configuration. Stable geometric constraints are formed between any three nodes and the virtual center point, thus maintaining the overall stability of the geometric configuration of the entire four-point unit under orbital perturbations and relative motion conditions. The logical topological roles of each physical satellite node are locked and bound to the physical satellite entity in this mode. The satellite's number, topological position, and functional role are fixed after initial configuration and do not migrate or drift. The topology matrix operation module loads a static four-satellite fully connected adjacency matrix. In this matrix, all four nodes are marked as connected, and the overall network topology remains fixed. In this mode, the system does not perform dynamic integration, dynamic beam reconfiguration, or instantaneous link switching; it only continuously outputs the steady-state availability. To maintain a low-jitter, highly stable communication link, adapting to service scenarios with strict requirements for jitter and stability, such as civilian long-term broadband access, broadcast television relay, and navigation signal enhancement broadcasting. During continuous operation, the reliability mathematical calculation module performs sampling calculations at fixed time intervals, such as once per minute. The current value, when When the value falls below a preset threshold, the system sends a maintenance alarm to the ground monitoring terminal, prompting the ground station to assess whether it is necessary to switch to dynamic mode or arrange network replenishment operations.

[0102] Example 2: Implementation of Dynamic Instantaneous Free Networking Mode like Figure 1 and Figure 3As shown, under normal space maneuvering conditions, such as low-Earth orbit satellite constellations operating for a long time in an orbital altitude range of 500 to 1200 kilometers, the inter-satellite relative speed can reach more than 7 kilometers per second, and the topology structure changes continuously over time. The system automatically switches to a dynamic instantaneous free networking mode through a three-mode switching control module or by ground command control.

[0103] In this mode, the fourth-level autonomous topology operation module switches the quadrilateral topology units to a software-defined flexible logical topology. The topological coordinates of each satellite node are no longer locked to their initial positions, but drift and dynamically adapt in real time with the relative motion between satellites, while maintaining the stability of the overall configuration around the virtual central constraint point. The topology matrix operation module operates from static matrices... Switch to dynamic time-varying matrix The matrix elements are updated in real time according to the satellite's orbital position. Each satellite continuously exchanges position information and link quality data via inter-satellite links. When the distance between two satellites exceeds the communication range or the link quality falls below a preset threshold, the corresponding matrix element is set to zero, indicating a broken link. When the two satellites re-enter the line of sight and the link quality meets the requirements, the corresponding matrix element is set to one, indicating a connected link. This matrix is ​​continuously refreshed at millisecond intervals to ensure that the topology matrix always reflects the current true inter-satellite connection status.

[0104] At the same time, the 3×3 nine-beam probability calculation module calculates the current interference intensity in real time. The probability of silent single-satellite communication Combined with single-star availability probability And compared with the traditional 3-beam scheme Calculate the anti-interference gain Output The value is passed to the beam scheduling logic when When the value is lower than the preset threshold, i.e. the interference intensity exceeds the system's tolerance limit, the system triggers a beam reconfiguration command to switch the currently most severely interfered axial beam to other idle frequencies or adjust the beam direction to avoid the interference source.

[0105] The reliability mathematical solution module within the specified time window The system uses discrete sampling, for example, sampling once every ten milliseconds, to obtain instantaneous availability. The average availability is approximated by numerical integration based on the time-series data. The integral result is used to drive the fission and reconfiguration decisions of the nine-beam array, when... When the integral value of the next window is lower than that of the previous window, it indicates that the network performance is deteriorating. The system should increase beam redundancy or start pre-scanning and pre-matching of candidate neighboring stars in advance to maintain the average availability above the preset threshold.

[0106] In this mode, all links between satellite nodes are transient and temporary. For example... Figure 3 As shown in the process control diagram, during continuous operation in dynamic mode, the system monitors the link status in real time and triggers beam scheduling when link quality fluctuates, ensuring that communication is always maintained on the link with the highest available quality. This process relies on each node to continuously and dynamically establish, disconnect, and update links to maintain network connectivity, without depending on a fixed link table pre-configured on the ground.

[0107] Example 3: Implementation of a wartime decentralized millisecond self-healing mode like Figure 2 and Figure 3 As shown, in sudden war or intense confrontation scenarios, such as when a satellite is subjected to enemy directional radio frequency interference, electromagnetic pulse attack, kinetic energy impact, or laser hard destruction, after the system detects a strong electromagnetic interference alarm or a satellite damage alarm, the three-mode switching control module automatically switches to the wartime centerless millisecond self-healing mode within milliseconds.

[0108] by Figure 2 Taking the scenario shown as an example, let's assume a satellite in a fourth-level unit. Permanent failure due to hard destruction or strong interference. The millisecond-level self-healing reconfiguration module first continuously monitors the health status of each node through the real-time link quality assessment subunit. When... If no response signal is received within N consecutive decision periods or the signal-to-noise ratio remains below the minimum usable threshold, Marked as a failed node. The topology matrix operation module immediately and automatically performs a matrix update: converting the original 4x4 adjacency matrix... middle The corresponding fourth row and fourth column are completely removed, and the dimension is reduced to generate a 3x3 three-point closed-loop reconstruction matrix. This matrix describes , , The pairwise fully connected relationships between the three remaining nodes.

[0109] During this process, the unit availability determination rules of the Level 4 Autonomous Topology Operation Module determine the availability status of the current unit in real time: if the number of intact satellites is three, it meets the availability condition of three or more, so the unit is not determined to be faulty. At the same time, since the number of faulty nodes is one, it meets the self-healing trigger condition within the unit of less than or equal to one, and the millisecond-level self-healing reconstruction module only performs three-point closed-loop reconstruction within the current Level 4 Autonomous Unit. , , The two nodes are naturally connected in the original fully connected topology. The system only needs to switch the routing table from a four-node full mesh to a three-node triangular closed loop route without triggering any external link adjustments or network-wide signaling broadcasts. Other units in the network are completely unaffected, and the signaling overhead generated during the self-healing process is kept to a minimum.

[0110] like Figure 2 As shown, The nodes marked with dashed lines are "faulty or interference failure nodes". , , A rigid triangular closed-loop subnet is formed by bidirectional solid lines. The entire reconfiguration process is completed locally on-board. Fault identification is handled by autonomous on-board monitoring, matrix recalculation by the onboard processor, and link reconfiguration by onboard beam control, all without ground-based command intervention. The entire self-healing reconfiguration process, from fault occurrence to the new topology taking effect, is strictly controlled within [timeframe missing]. Within 100ms. Timing constraints. ≤100ms and 100ms is much less than The physical meaning lies in: due to the characteristic time constant of satellite orbital motion The time constant is on the order of tens of seconds, and 100ms is only one-thousandth to one-hundredth of that time constant. At the instant that the self-healing reconstruction is completed, , , The relative positions of the three satellites have not changed perceptibly. Therefore, the geometric conditions of the three-point rigid closed triangular subnet on which self-healing depends remain stable before and after self-healing. The self-healing result is effective and there will be no situation where the topology has failed due to orbital motion after reconstruction.

[0111] In more severe cases, such as the simultaneous failure of two satellites in a Level 4 unit, reducing the number of intact satellites to two, the unit can be deemed to have failed if the number is less than or equal to two, making it impossible to maintain a three-point triangular closed loop within a single unit. In this situation, the millisecond-level self-healing reconfiguration module initiates a cross-unit, full-network secondary reconfiguration: the system retrieves idle or low-load satellite nodes from neighboring Level 4 units to replenish the damaged unit. These replenished nodes are connected via a heterogeneous satellite intelligent network replenishment mechanism or redundant poles from adjacent units, restoring the four-point topology. This full-network secondary reconfiguration process takes slightly longer than intra-unit self-healing, but it is still completed through onboard autonomous computation, without relying on ground scheduling, ensuring continuous communication connectivity across the entire network.

[0112] In this mode, the system also performs real-time calculations based on the traditional hardware networking comparison model. When the ground monitoring terminal issues a comparison query command, the system collects the actual availability of this solution at the current moment. Theoretical availability compared to traditional solutions This solution offers excellent cost-effectiveness. Cost-effectiveness of traditional solutions and gain difference , The formula calculates the gain difference and transmits the calculation results back to the ground monitoring terminal in the form of numerical values ​​and trend charts to assist ground command personnel in deciding whether to switch operating modes or start the network replenishment process.

[0113] Example 4: Implementation of Millisecond-Level Predictive Seamless Switching In dynamic or wartime modes, all satellite nodes are constantly in high-speed relative motion. The system monitors the current link's survivability in real time through a millisecond-level predictive seamless switching submodule. Specifically, each satellite obtains the real-time inter-satellite distance through two-way inter-satellite ranging and signal arrival time difference measurement. Inter-satellite relative motion velocity is obtained by Doppler frequency shift measurement. and orbit away from the rate Propagation delay is obtained by measuring the round-trip time of the signal. Substitute these real-time parameters into the remaining visible time window function. Its explicit expression is: ,

[0114] in The maximum inter-satellite communication distance is determined by the satellite's transmit power, antenna gain, receiver sensitivity, and space propagation loss. To preset the maximum visible time window, it can be half of the orbital period or a larger fixed value; (Introduction) This allows for margins in signal transmission and processing. During engineering implementation, the function can also be used for refined modeling based on actual orbital parameters and link budgets, including consideration of factors such as Earth's shielding and atmospheric attenuation.

[0115] In the current link When the time drops below a preset threshold, such as ten milliseconds, the system determines that the current link is about to lose line-of-sight. At this point, instead of waiting for the link to actually disconnect, the system initiates a global scan of neighboring satellites while the current link is still operational. The beam control module on the satellite, while maintaining the current communication link, rapidly scans other directions with an idle beam to detect available candidate satellites in the vicinity. Upon detecting a candidate satellite, the system immediately performs rendezvous identification, verifying the other party's legitimacy as a network node by exchanging identity identifiers and orbital parameters, and performs pre-matching to assess the compatibility between the candidate satellite and the current serving satellite in terms of orbital position, communication angle, and load status. After successful pre-matching, the system completes a pre-handshake with the new satellite within the window before the current link disconnects, exchanging encryption keys, synchronizing communication parameters, and reserving bandwidth resources. When the old link momentarily disconnects due to the satellite losing line-of-sight, the new link has already completed all preparations for establishment during the pre-handshake, achieving a synchronous and instantaneous handshake, seamlessly taking over communication services without any gaps or network interruptions.

[0116] Example 5: Implementation of Intelligent Network Supplementation for Heterogeneous Satellites like Figure 3 As shown, during normal system operation, the networking communication modules of heterogeneous satellites in orbit, such as remote sensing satellites and navigation satellites in the same orbit as communication satellites, are in a low-power sleep standby state, only retaining the basic wake-up signal receiving circuit to work, and do not participate in network topology maintenance, so as to save on-orbit energy.

[0117] When the Level 4 autonomous topology operation module detects a topology pole gap, for example Figure 2 In In the event of a failure, if no available homogeneous communication satellites can be added to the four-point unit to restore full connectivity, or if the number of intact satellites in the unit drops to two or less, the system triggers a heterogeneous satellite wake-up procedure. The wake-up procedure first selects candidate platforms with both radio frequency and laser full-duplex transceiver capabilities from a list of heterogeneous satellites operating on a unified public networking frequency band. This selection is based on an onboard list of heterogeneous satellite capabilities, which is pre-maintained and periodically updated by the ground-based aerospace control system. During the selection process, multiple indicators are comprehensively evaluated, including the relative positional deviation of the candidate satellite from the current four-point unit, orbital plane consistency, remaining on-orbit lifetime, current load margin, and communication link quality. These are ranked by weighted scores, and the satellite with the highest score is selected as the replacement satellite. After a compliant heterogeneous satellite is selected, the system sends a wake-up command to the wake-up receiving circuit of that heterogeneous satellite via an inter-satellite link.

[0118] Upon receiving the wake-up command, the heterogeneous satellite's networking communication module transitions from a low-power sleep state to an active state, activating the RF and laser transceiver front-ends, calibrating the frequency reference, synchronizing time slots, and establishing a preliminary communication link with the node issuing the activation command. Subsequently, the remote source code injection module injects a networking logic role package into the heterogeneous satellite. This logic package contains the topology role parameters required by the currently damaged unit, such as supplementary parameters. The location becomes the fourth pole, and the adjacency constraint condition is the same as... , , The requirements for connectivity between them and the complete set of availability calculation functions. After receiving and loading the logical packet, the heterogeneous satellite updates its own adjacency matrix configuration and moves towards... , , After initiating a link establishment request and completing a three-way handshake, the damaged unit is officially added as a new pole, simultaneously updating the entire network topology matrix and the nine-topology fission branch relationships. The entire process operates autonomously in a closed loop on-board, requiring no ground-based human intervention. After the network replenishment process is completed, the system can reassess whether to switch to static or dynamic mode based on the new topology structure.

[0119] Example 6: Implementation of Dual-System Collaborative Backup This system operates in parallel with the deployed hardware space-based routing satellites. Under normal circumstances, the space-based routing satellites undertake the main tasks of inter-satellite data exchange and cross-constellation forwarding. This system is in a hot backup standby state, with each module loaded with complete mathematical models and topology matrices. All satellites in orbit have distributed routing and forwarding capabilities, but routing decisions are still dominated by the hardware routing satellites.

[0120] When a space-based routing satellite suffers electromagnetic interference or physical damage, such as when the primary routing satellite in a dual-satellite system is suppressed by interference or the backup routing satellite loses contact due to a malfunction, the space-based routing compatible backup module detects a timeout in the routing satellite's response or a sudden drop in the signal-to-noise ratio, automatically triggering a system takeover command. During the takeover process, the system first confirms that all modules are in normal condition and the topology matrix is ​​complete. Then, it switches the routing decision mode of each satellite from "following the routing satellite's scheduling" to "independent distributed routing forwarding." Each satellite node rescans its neighboring satellites and establishes transient links through a transient link-building dynamic iteration mechanism, completing the self-organized reconstruction of the entire network topology in a very short time and independently undertaking all space-based communication tasks. Since this system shares the same inter-satellite link physical layer and frequency band resources with the hardware routing satellites, and the system's mathematical model and topology matrix have already been pre-injected into each satellite, the takeover process does not require re-injecting the program or modifying the hardware configuration; it only requires switching the mode flag, and the switching time is much shorter than the communication service interruption tolerance threshold. The dual-system collaborative operation provides dual communication assurance. When the hardware routing satellite is working normally, the two systems complement each other. When the hardware routing satellite fails, this system seamlessly takes over, ensuring that space-based communication is not interrupted due to a single system failure.

[0121] Example 7: Heterogeneous Multi-orbit Satellite Networking Scenario In a certain networking scenario, the network consists of multiple existing heterogeneous satellites in low Earth orbit (LEO) and medium Earth orbit (MEO). The ground station pre-loads frequency and channel parameters, and each satellite enters a network-ready state. During operation, one LEO satellite experiences radio frequency interference and its link fails. The system immediately triggers the fault detection threshold, switches the compensation iteration matrix, recalculates and generates a detour communication link, and completes the self-healing reconfiguration of the entire network topology within 100 milliseconds. The service transmission of the remaining satellites is unaffected. After the interference is eliminated, the system automatically switches back to the normal main topology matrix and restores the original logical backbone link working mode.

[0122] Example 8: Satellite Node Damage Case Multiple satellites in orbit have completed a four-level-nine-level topology network, and the network is in normal service transmission. When one of the participating satellites is damaged by an impact and goes offline, all existing related links are broken. The networking software quickly identifies the node's disconnection status, migrates the management logic to the remaining surviving satellites, and initiates compensation iterative matrix operations. Within 100 milliseconds, the link replanning and topology reconstruction are completed, and the network continues to maintain complete connectivity, ensuring normal transmission and reception of ground target service data. If the damaged satellite cannot be recovered, the network continues to maintain stable operation of the compensation topology. The core of this invention includes two software operation models. Under normal conditions, a positive nine-order main topology matrix is ​​used to complete the routing and scheduling of nine logical backbone links. When a node is damaged or a link is interrupted due to radio frequency interference, an offset iterative compensation matrix is ​​initiated to perform iterative operations, generating detour compensation links. The iteration convergence time is controlled within 100 milliseconds, completing transient self-healing topology reconstruction. After the fault is cleared, it automatically recovers to the positive nine-order main topology matrix. The matrix operations are executed by on-board software and do not rely on real-time ground calculations. The matrix is ​​a logical model, not a physical connection matrix of the satellite hardware.

[0123] In practical engineering implementation, the 3×3 nine-beam probability calculation module will... The calculation is encapsulated in the function Calc_Beam9(p), which includes the anti-interference gain. The computation is encapsulated as the function Calc_Gain_K(p); the four-level autonomous topology operation module encapsulates node availability judgment as the function NodeAvailability(n); the three-mode switching control module encapsulates the switching logic of the three modes as the functions Mode_Static(), Mode_Dynamic(), and Mode_SelfHeal(), respectively. Each function is implemented in standard C language in the spaceborne embedded system and is called by other modules through a unified calling interface.

[0124] Cluster damage correction item , , The specific calibration process is as follows: First, a Monte Carlo simulation model is constructed, including the spatial coordinates of four satellites in our four-level unit, and parameters such as the damage radius, explosion timing, and attack sector of enemy area-of-effect weapons (e.g., electromagnetic pulse warheads, kinetic energy interceptor groups, laser blinding arrays, etc.). At least 100,000 damage scenarios are randomly generated, and random variables such as weapon impact point, explosion time, and damage radius in each scenario are sampled according to a preset probability distribution. The frequency of exactly two satellites being damaged simultaneously in each scenario is counted. The frequency of exactly three satellites being destroyed simultaneously The frequency of exactly four satellites being destroyed simultaneously Divide the above frequencies by the total number of scenarios. Gain experience while increasing the probability of damage , , Simultaneously, the corresponding probabilities are calculated using an independent binomial distribution model: , , Then the correction term = , , If simulation conditions are lacking, historical case studies can be used: extract the number of simultaneous multi-satellite destruction events from publicly available records of space combat tests and anti-satellite weapon tests, and calculate using the same method described above. When no actual measurement or simulation data is available, [the following can be used]... That is, using an independent damage model, the corrected failure probability degenerates into... .

[0125] In actual engineering implementation, the software code for each of the above modules must be written according to aerospace-grade C language specifications, conforming to military standards such as GJB 5369-2005 "Design Guidelines for Reliability and Security of Spacecraft Software" and international aerospace standards such as ECSS-Q-ST-80C "Space Product Assurance". All floating-point operations must be limited to prevent numerical overflow. Integral operations in the embedded environment should use discrete numerical integration, such as the trapezoidal rule or Simpson's rule, instead of analytical integration to reduce computational resource consumption. Remote source code injection must employ encrypted transmission and digital signature authentication mechanisms. The uplink must adopt a spread spectrum anti-interference system to prevent the enemy from injecting malicious code or intercepting the network logic. The embedded systems of each satellite must have watchdog timers and program rollback capabilities. When a newly injected network program malfunctions, it can automatically revert to the previous stable version within a specified time window to ensure that the satellites in orbit do not completely lose control due to software defects.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A three-mode four-level nine-dimensional centerless dynamic self-healing space-based satellite anti-jamming networking system, characterized in that, include: Remote source code injection module, four-level autonomous topology operation module, 3×3 nine-beam probability calculation module, three-mode switching control module, topology matrix operation module, reliability mathematical solution module, millisecond-level self-healing reconstruction module, and space-based routing compatible backup module; The remote source code injection module is used to remotely inject the networking program, which encapsulates the complete set of mathematical models, into existing on-orbit satellites without modifying the satellite hardware or launching dedicated space routing satellites, so that each satellite has distributed routing and forwarding capabilities and local mathematical calculation capabilities. The remote source code injection module also supports injecting networking logic roles into heterogeneous satellites in orbit. When a topological pole is missing in the network, it wakes up a compliant heterogeneous satellite in a low-power standby state and injects networking role parameters and topological constraints into it, so that it can be added as a new pole to the fourth-level unit and participate in the nine-level fission expansion. The four-level autonomous topology operation module uses four satellites as the smallest autonomous unit. The four satellite nodes are completely equal and there is no fixed routing center. The entire system has no fixed master control satellite, and network management logic can dynamically migrate between surviving satellites. The failure of a single satellite will not cause the entire network to collapse. The four-level autonomous topology operation module sets a virtual center constraint point at the geometric center of the four satellites. The four satellites form a quadrilateral topology unit with this virtual center constraint point as the geometric anchor point. Each satellite node within the unit performs relative position drift and coordinate adaptive adjustment around the virtual center constraint point to maintain the overall stability of the unit's geometric configuration. The quadrilateral topology unit is a software-defined flexible logical topology, not a physically fixed structure. The logical topology role carried by each physical satellite node within the unit is decoupled from the physical satellite entity. When any physical satellite fails, its logical topology role can be migrated to other in-orbit satellites. Satellites take over the transmission, maintaining the logical integrity of the quadrilateral topology unit. The topological coordinates of each satellite node within the unit drift and dynamically adapt in real time with the relative motion between satellites, compatible with satellite node networking across all orbital altitudes and speed ranges, including high, medium, and low orbits. The four satellite nodes of the quadrilateral topology unit are fully connected, naturally containing four sets of independent three-point rigid triangular subnets. When any node fails, the remaining three nodes automatically form one of the three-point rigid triangular subnets to maintain communication within the unit. The four-level autonomous topology operation module has a built-in real-time link quality evaluation subunit, using the signal-to-noise ratio and bit error rate of real-time inter-satellite electromagnetic wave transmission as the priority criteria for link establishment and partner selection, replacing the pre-configured static link table on the ground. Each satellite node autonomously and in a distributed manner completes the dynamic selection of networking partners and the construction of topology units; the four-level autonomous topology operation module has built-in unit availability determination rules: the unit availability condition is that the number of intact satellites is ≥3, and the unit failure condition is that the number of intact satellites is ≤2; The 3×3 nine-beam probability calculation module corresponds to nine independent communication beams (three groups each along the X, Y, and Z axes) configured for a single satellite. These nine independent communication beams are software-defined logical beams, dynamically allocated and reconfigured at the software level by an onboard beam scheduling algorithm, without requiring nine additional physical hardware connections on the satellite. The nine beams are independent of each other; the failure of any one beam does not affect the operation of the remaining beams. A built-in anti-interference mathematical model states that the failure probability of all nine beams on a single satellite being suppressed is... The overall usability probability of a single star is ,in The probability of failure due to single-beam electromagnetic interference. The probability of physical destruction of a single satellite; The three-mode switching control module is configured with three working modes: static civilian fixed network mode, dynamic instantaneous free networking mode, and wartime centerless millisecond self-healing mode. The topology matrix operation module has three built-in adjacency matrices that can be implemented programmatically: a static four-star fully connected matrix and a dynamic time-varying topology matrix. A three-point closed-loop reconstruction matrix for fault dimensionality reduction; automatic updating of matrix elements and removal of failed nodes when a fault occurs; dynamic splicing between the four-level units; any two or more four-level units can be interconnected through nine-level fission branches between poles to form 6-star, 8-star, 9-star and larger-scale cluster networks. The reliability mathematical calculation module incorporates a closed-form formula for the availability of four-star units: Meanwhile, the average availability of the built-in system Cost-effectiveness model Anti-interference gain ,in The number of satellites in the network; The millisecond-level self-healing reconfiguration module incorporates segmented triggering logic: when the number of faulty nodes is ≤1, a three-point closed-loop reconfiguration is performed only within the current fourth-level autonomous unit, without triggering external link adjustments; when the number of faulty nodes is ≥2, a cross-unit full-network secondary reconfiguration is initiated; the millisecond-level self-healing reconfiguration module incorporates timing constraints: self-healing convergence delay ≤100ms, all fault identification, matrix recalculation, and link reconstruction operations are completed locally on the satellite without the need for ground commands; The space-based routing compatibility backup module and the hardware space routing are deployed in parallel in a network. When the routing satellite fails, this system independently undertakes the communication task, forming a dual communication guarantee. The system constructs a net-like closed topology structure through a four-level unit progressive nine-level fission. The entire network has no weak points at the ends of the strip topology, and any node is located on a closed loop, realizing multi-point cross redundancy and multi-path detour routing.

2. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, In the topology matrix operation module, a static four-star fully connected adjacency matrix is ​​used. for: ， When any node fails, a three-point closed-loop reconstruction matrix is ​​automatically generated. for: ， Matrix elements in dynamic mode Updated in real time according to the satellite's orbital position; 1 represents a connected link, and 0 represents a disconnected link. ,in , .

3. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The 3×3 nine-beam probability calculation module supports real-time calculation of the single-satellite communication silence probability under different interference intensities, compared with the traditional 3-beam architecture. The anti-interference effect is improved by multiples through exponential difference quantification. The output evaluation results are used to adjust the beam scheduling strategy.

4. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The reliability mathematical calculation module incorporates a binomial distribution failure probability derivation formula: Four-star unit independent damage failure probability ; Overlay cluster damage corrections , , Obtain the corrected failure probability ; Unit availability ; All formulas are encapsulated as callable functions and written into the source code.

5. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, In wartime, decentralized, millisecond-level self-healing mode, the system enforces timing constraints. ,in The orbital topology deformation characteristic time; ground-based centralized scheduling is disabled, and fault diagnosis, matrix refresh, and subnet reconstruction are completed independently on the satellite.

6. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The system calculates the cost-effectiveness index of the two solutions in real time: Cost-effectiveness of this solution ; Cost-effectiveness of traditional routing solutions ; Automatic output gain difference ; ； As output parameters for network status assessment, they support the software's visual monitoring function.

7. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, It also includes a millisecond-level predictive seamless switching submodule: Based on the relative motion speed between stars , orbital deviation rate Real-time inter-satellite distance and signal propagation delay Dynamically calculate the remaining visible time window of the current link. The function The explicit form of expression is: ， in This represents the maximum inter-satellite communication distance. Set the maximum visible time window; Before the current link leaves line of sight, initiate a full-domain scan and pre-matching of neighboring satellites in advance, complete the pre-handshake of the new link, and achieve a seamless transition with zero interruption between the disconnection of the old link and the establishment of the new link.

8. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The specific method by which the remote source code injection module performs networking logic role injection on heterogeneous satellites in orbit is as follows: When the number of intact satellites in a Level 4 autonomous unit drops to two or less, or when a topological pole vacancy is detected across the entire network, the system automatically selects on-orbit heterogeneous satellites that have unified public networking frequency bands and have RF and laser duplex transceiver capabilities. It then injects a networking logic package containing topological role parameters, adjacency constraints, and availability calculation functions into the satellite, allowing it to fill the damaged unit as a new pole, and simultaneously updates the topology matrix and the nine-topology fission branch relationship.

9. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The four-level nine-topology fission expansion is based on the four poles of each four-level autonomous unit, and then expands outward to form three new branches, which are nested layer by layer to form a spherical closed network topology; the spherical closed topology has no strip-shaped end nodes, and there are at least two non-overlapping communication paths between any two nodes in the network.

10. The three-modal, four-level, nine-topology, centerless, dynamic self-healing space-based satellite anti-interference networking system according to claim 1, characterized in that, The mathematical criteria for determining the three-point rigid triangular subnet are as follows: For any three nodes in a quadrilateral topological element , , If the link states between any two of them satisfy Then the three points form a rigid closed triangular subnet; In a quadrilateral fully connected topology, the following four sets of independent three-point rigid triangular subnets naturally exist: 、 、 、 ； When any pole fails, the system automatically switches to the corresponding rigid closed triangular subnet at the remaining three points to maintain uninterrupted communication within the cell.

Images