Multi-layer progressive hybrid interstellar network adaptive planning method, device and medium

Abstract

The application provides a kind of multilayer progressive hybrid inter-satellite network adaptive planning method, equipment and medium, it is related to network planning field.The application constructs hybrid inter-satellite network optimization model, and global chain building problem is decomposed into laser fixed topology optimization, microwave static topology optimization and dynamic topology optimization three sub-problems.The method first optimizes laser backbone link to determine fixed laser topology chain building strategy;Second, on the basis of laser fixed topology, microwave static link is optimized to determine the static microwave topology chain building strategy of isomorphism and heterostructure, maintains basic signaling transmission and full connectivity;Finally, in combination with current network state and inter-satellite visibility, with network average geometric distribution factor and maximum hop number of extraterritorial to inland satellite transmission as optimization goal, real-time microwave dynamic link adjustment solution is solved by using dynamic programming algorithm based on simulated annealing.The application significantly reduces the on-board networking calculation complexity, and greatly improves the robustness of self-organizing network.

Description

Technical Field

[0001] This invention relates to the field of inter-satellite network planning technology, and more specifically, to a multi-layered progressive hybrid inter-satellite network adaptive planning method. Background Technology

[0002] Global Navigation Satellite Systems (GNSS) provide inter-satellite observation data and communication capabilities through inter-satellite links, which greatly enhances the overall performance of navigation constellations in clock bias determination, orbit determination, and autonomous navigation. To maximize the effectiveness of the inter-satellite network, inter-satellite link network planning is a critical issue that must be addressed first during constellation operation. In existing technical solutions, navigation constellations typically employ a method where ground stations pre-plan time slot tables and centrally upload them to establish the entire network of satellites. However, this traditional ground-based centralized planning method exhibits significant limitations when facing scenarios with highly dynamic space environments and frequent topology changes. On the one hand, the uplink and downlink transmission of satellite-to-ground links are subject to visible time window limitations and communication delays. If a sudden node failure or link interruption occurs in the inter-satellite network, ground stations struggle to respond in real time and re-upload the time slot tables, resulting in extremely poor application flexibility. On the other hand, as the scale of the inter-satellite network continues to expand, the massive computational load and communication overhead brought by centralized computing become an unbearable burden on the system.

[0003] Furthermore, the future development trend of inter-satellite links will gradually evolve from single microwave links to a heterogeneous link form that combines microwave and laser technologies. In this hybrid inter-satellite network, laser links typically serve as the constellation backbone, providing high-speed data transmission and high-precision inter-satellite measurements. However, due to the difficulty and time-consuming process of laser terminal acquisition, tracking, and aiming, the connection relationships of laser links usually need to remain relatively fixed, making frequent dynamic switching difficult. In contrast, microwave links have the advantages of wide beam coverage and rapid link establishment, making them more flexible than laser links. They can be dynamically adjusted to adapt to instantaneous changes in network topology and are typically used as supplementary and access networks, providing link supplementation during autonomous orbit determination and rapid transmission of inter-satellite signaling. While this microwave-laser hybrid system combines the advantages of both, it leads to an exponential increase in the combination space and constraints of network links.

[0004] Faced with such highly complex heterogeneous networks containing both dynamic and static attributes, traditional ground-based pre-planning methods not only struggle to solve the massive and highly coupled global network link optimization problem, but also fail to meet the real-time requirements of space network self-healing. With the significant improvement in onboard processing capabilities of modern satellite platforms, onboard adaptive planning has become an inevitable way to overcome these technical bottlenecks. Onboard adaptive planning can effectively enhance the autonomous networking capabilities of navigation constellations, significantly reducing the system's dependence on ground control. Thus, even when some network nodes fail or links are abnormally interrupted, the inter-satellite network can still maintain a good orbit determination configuration and transmission performance, effectively improving the reliability and robustness of the entire inter-satellite network. Therefore, considering the heterogeneous and multi-layered physical characteristics of future laser-microwave hybrid inter-satellite networks, the industry urgently needs an efficient, robust, and low-complexity onboard adaptive planning networking algorithm to reasonably reduce and decompose the complex global optimization problem, effectively supporting the comprehensive realization of constellation adaptive planning and self-organizing capabilities, thereby optimizing the satellite geometry configuration during autonomous orbit determination and reducing the transmission latency of data returned from overseas. Summary of the Invention

[0005] The present invention aims to solve at least one of the aforementioned technical problems existing in the prior art.

[0006] To this end, the first aspect of the present invention provides an adaptive planning method for multi-layer progressive hybrid inter-satellite networks.

[0007] A second aspect of the present invention provides an electronic device.

[0008] A third aspect of the present invention provides a computer-readable storage medium.

[0009] This invention provides an adaptive planning method for multi-layer progressive hybrid inter-satellite networks, comprising: The entire navigation constellation operation cycle is divided into several time segments of equal duration. The inter-satellite visibility matrix between satellites is obtained in each time segment. A hybrid inter-satellite network optimization model is constructed, and the multi-layer optimization problem is decomposed into three sub-problems: laser fixed topology optimization, microwave static topology optimization, and dynamic topology optimization. The laser backbone link is optimized, and the transmission hop count and transmission delay are evaluated based on the differences in satellite connection methods. A fixed laser topology link establishment strategy is determined to construct a fixed laser link network. Based on the laser fixed link network, the microwave static link is optimized to determine the link establishment strategy for co-track and heterogeneous static microwave topologies, thereby constructing a microwave static link network to transmit signaling and maintain full network connectivity when the laser link is abnormally interrupted. Based on the current network status and satellite visibility, and taking the network average geometric distribution factor and the maximum number of hops in satellite transmission from overseas to domestic locations as optimization objectives, a dynamic programming algorithm is used to dynamically adjust and optimize microwave links in real time, generating a dynamic link network. The hybrid inter-satellite network link establishment planning results are then output by combining the laser fixed link network and the microwave static link network.

[0010] The multi-layer progressive hybrid inter-satellite network adaptive planning method according to the above-described technical solution of the present invention may also have the following additional technical features: In the above technical solution, in the hybrid inter-satellite network optimization model, the visibility matrix between each satellite is defined as follows: The topology matrix of the hybrid inter-satellite network is defined as follows: ,in, This is a fixed link topology matrix, containing microwave fixed link topologies. and laser fixed link topology ; This is a dynamic link topology matrix, containing microwave dynamic link topology. and laser dynamic link topology The expression for decomposing the original global hybrid inter-satellite network link establishment optimization problem using the idea of ​​hierarchical optimization is as follows:

[0011] in, The objective function is composed of the network's average geometric distribution factor and the maximum number of hops in satellite transmissions from overseas to domestic locations; T represents the transpose matrix operation; i and j are both indices of matrix elements.

[0012] In the above technical solution, when determining the fixed laser topology link establishment strategy, based on the distribution characteristics of the satellite orbital plane, inter-satellite continuous visibility, and ground station coverage elevation angle constraints, the maximum number of hops and transmission delay of satellite data transmission from overseas to domestic are evaluated, and satellites are continuously tracked and linked by laser to construct a multi-ring backbone network.

[0013] In the above technical solution, when determining the static microwave topology link establishment strategy, the set of continuously visible satellites of the target satellite is obtained based on inter-satellite continuous visibility analysis. A first number of co-orbit microwave links are allocated in the set of continuously visible satellites to maintain co-orbit connectivity. Combined with the evaluation results of the network's average geometric distribution factor performance, a second number of disorbit microwave links are allocated to construct a heterogeneous microwave network.

[0014] In the above technical solution, the expression for the model that optimizes the dynamic link topology in the real-time dynamic adjustment and optimization solution of the microwave link using a dynamic programming algorithm is as follows:

[0015] The dynamic optimization results are obtained by solving the model.

[0016] In the above technical solution, a dynamic link optimization algorithm based on simulated annealing is used to solve the model in real time, including: Set the annealing initialization temperature, and based on the visibility matrix between satellites and the position coordinates between satellites obtained from the current topology planning time segment, use an integer programming algorithm to establish links according to the specified number of links required for each satellite to obtain the initial dynamic link establishment matrix. A dual-link switching strategy is used to search the neighborhood under each temperature condition. Two existing dynamic links (a,b) and (c,d) are randomly selected based on the current link establishment matrix, i.e., satellite a is currently connected to satellite b, and satellite c is connected to satellite d. If satellite a and satellite c are mutually visible and not currently linked, and satellite b and satellite d are mutually visible and not currently linked, the topology is changed to switch the two links, so that satellite a is linked to satellite c and satellite b is linked to satellite d, thus obtaining the new link establishment matrix under the current temperature.

[0017] In the above technical solution, after generating the new state chain matrix, the average geometric distribution factor and the minimum number of hops from outside the territory to inside the territory for each satellite are calculated, and the new generation value is obtained through objective function calculation. and with current generation values Compare; according to the Metropolis criterion, when When choosing to accept the new state, At that time, the decision on whether to accept the new state is based on the calculated acceptance probability. The method for calculating the acceptance probability is as follows:

[0018] Where p is the acceptance probability; T is the simulated annealing temperature in the current state.

[0019] In the above technical solution, after deciding whether to accept the new state, according to Update the simulated annealing temperature, where, This represents the simulated annealing temperature at the k-th iteration; This represents the simulated annealing temperature at the (k+1)th iteration; It is the temperature decay factor, and satisfies The algorithm continues until the upper limit of iteration is reached, at which point it terminates and the current chain-building matrix is ​​output as the final optimization result.

[0020] The present invention also provides an electronic device, including a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to enable the electronic device to perform a multi-layer progressive hybrid inter-satellite network adaptive planning method as described in any of the above technical solutions.

[0021] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements a multi-layer progressive hybrid inter-satellite network adaptive planning method as described in any of the above technical solutions.

[0022] In summary, due to the adoption of the above-mentioned technical features, the beneficial effects of the present invention are: This invention provides a multi-layered, progressive, hybrid inter-satellite network adaptive planning method. By constructing a hybrid inter-satellite network optimization model, it cleverly decomposes the originally computationally complex and difficult-to-solve global link building planning problem into three dimensionality-reduced sub-problems: laser fixed topology optimization, microwave static topology optimization, and dynamic topology optimization. This multi-layered, progressive, and dynamic-static combined solution architecture not only significantly reduces the computational complexity and resource consumption on-board, but also completely eliminates the strong dependence of traditional centralized ground planning on the communication window and timeliness of satellite-to-ground links, truly realizing efficient, low-complexity autonomous networking and adaptive planning of hybrid inter-satellite networks on satellite.

[0023] This invention fully leverages the heterogeneous physical characteristics of laser and microwave links through a layered deployment. The laser inter-satellite link serves as the backbone network, maintaining a relatively fixed topology to ensure the stability of high-speed data transmission and high-precision measurements, effectively avoiding the frequent switching and reconnection costs caused by the difficulty of laser terminal acquisition and tracking. Simultaneously, the agile microwave link serves as a supplementary and access network. While retaining some co-orbital and heterogeneous static microwave links to maintain basic signaling transmission and full connectivity, a dynamic programming algorithm is used to dynamically adjust the microwave links in real time based on the current network status. This design leverages both the backbone advantages of the laser link and the flexible access characteristics of the microwave link, significantly optimizing satellite geometry during autonomous orbit determination and reducing transmission latency for data return from overseas.

[0024] Based on the planning method of this invention, under normal operating conditions, the average geometric distribution factor of the hybrid inter-satellite network can reach 1.001, with the maximum and minimum geometric distribution factors controlled within an excellent range of 1.132 and 0.941, respectively. Furthermore, all data from overseas satellites can be returned to China via laser inter-satellite links in a single hop, significantly reducing transmission latency. A more prominent benefit lies in the network's high robustness and fault self-healing capability. Even when facing a network link outage probability as high as 20%, thanks to the effective and real-time replenishment of critical links in the network by dynamic links, the average geometric distribution factor can still be improved to 1.136, representing a performance improvement of approximately 47.1% compared to traditional link establishment methods under the same fault conditions. It still strictly meets the requirement of one-hop transmission of business data or signaling data from overseas to China. In addition, this invention introduces a dual-link switching search strategy based on simulated annealing algorithm in the dynamic link optimization solution, effectively escaping the trap of local optima and further ensuring the quality and reliability of the global topology solution under large-scale constellation dynamic networking.

[0025] Additional aspects and advantages of the invention will become apparent in the following description or may be learned by practice of the invention. Attached Figure Description

[0026] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a schematic diagram of an adaptive planning process for a multi-layer progressive hybrid inter-satellite network provided in an embodiment of the present invention; Figure 2 This is a schematic diagram comparing different connection methods of laser ring-shaped fixing topologies provided in embodiments of the present invention; Figure 3 A schematic diagram of a continuously visible satellite for a microwave link provided in an embodiment of the present invention; Figure 4 A flowchart of dynamic link topology optimization solution based on simulated annealing algorithm provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of the overall planning of hybrid inter-satellite network link establishment provided in an embodiment of the present invention; Figure 6 A schematic diagram of the PDOP performance of a multilayer progressive adaptive planning hybrid network provided in an embodiment of the present invention; Figure 7 A schematic diagram illustrating the minimum number of hops between the outside and inside of a multi-layered progressive adaptive planning hybrid network provided in an embodiment of the present invention; Figure 8 This is a PDOP performance analysis diagram after link interruption provided in an embodiment of the present invention; Figure 9This is a performance analysis diagram of PDOP after adaptive planning provided in an embodiment of the present invention; Figure 10 A performance analysis diagram of transmission hop count after link interruption provided in an embodiment of the present invention; Figure 11 The transmission hop count performance analysis diagram provided for an embodiment of the present invention is shown. Detailed Implementation

[0027] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0028] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0029] The following reference Figures 1 to 11 This describes a multi-layer progressive hybrid inter-satellite network adaptive planning method, apparatus, and medium provided according to some embodiments of the present invention.

[0030] Some embodiments of this application provide a multi-layered progressive hybrid inter-satellite network adaptive planning method.

[0031] Before detailing the specific implementation methods, the core technical problem solved by this invention will be explained in general. In future hybrid inter-satellite networks, laser inter-satellite links will serve as the constellation backbone, providing high-speed data transmission and high-precision inter-satellite measurements. However, due to the difficulty of laser acquisition, tracking, and aiming, the connection relationships of laser links are usually relatively fixed. Microwave links, as supplementary and access networks, provide link supplementation and inter-satellite signaling transmission during autonomous orbit determination. Microwave links are agile and more flexible than laser links, and can dynamically adjust to adapt to network changes. Combining the aforementioned heterogeneous characteristics of laser and microwave links, this invention provides a multi-layered progressive hybrid inter-satellite network adaptive planning method. It rationally decomposes the global network link establishment optimization problem from both laser and microwave perspectives, as well as dynamic and static perspectives, and solves each optimization sub-problem step by step to address the problem of insufficient timeliness in traditional ground planning methods.

[0032] like Figure 1 As shown, the first embodiment of the present invention proposes a multi-layer progressive hybrid inter-satellite network adaptive planning method, including the following steps S1 to S4.

[0033] S1. Divide the entire navigation constellation operation cycle into several time segments of equal duration. Obtain the inter-satellite visibility matrix between each satellite in each time segment, construct a hybrid inter-satellite network optimization model, and decompose the multi-layer optimization problem into three sub-problems: laser fixed topology optimization, microwave static topology optimization, and dynamic topology optimization.

[0034] Specifically, step S1 constructs a hybrid inter-satellite network optimization model and decomposes the multi-layer optimization problem into three independent sub-problems: laser fixed topology optimization, microwave static topology optimization, and dynamic topology optimization. In practice, the entire navigation constellation's operational cycle is divided into several equally long planning time segments. Within each time segment, it is assumed that changes in the relative positions of the satellites are insufficient to cause substantial changes in visibility, thus maintaining a fixed link structure and simplifying the design space for link topology design. The visibility matrix between each satellite is defined as follows: This matrix is ​​represented as an n-dimensional 0-1 square matrix. In this matrix, if satellite i and satellite j are spatially visible in the current time segment and satisfy the physical conditions for establishing a link, then the elements are defined as follows: ,otherwise The topology matrix of the hybrid inter-satellite network is defined as follows: ,in, This is a fixed link topology matrix, containing microwave fixed link topologies. and laser fixed link topology , represented as ; This is a dynamic link topology matrix, containing microwave dynamic link topology. and laser dynamic link topology , represented as .

[0035] Based on the above matrix definition, the original problem is decomposed into dimensions using the idea of ​​hierarchical optimization. The expression of the constructed hybrid inter-satellite network optimization model is shown below:

[0036] in, The objective function is composed of the network's average geometric distribution factor and the maximum number of hops in satellite transmissions from overseas to domestic locations; T represents the transpose matrix operation; i and j are both indices of matrix elements.

[0037] In some embodiments, since the objective function is a multi-objective optimization, a normalized weighted summation model can be used in actual calculations, that is, the objective function expression can be constructed as:

[0038] in, Represents topological variables; This represents the normalized result of the network's average geometric distribution factor; This represents the normalized result of the maximum number of hops in satellite transmissions from overseas to within China. and These are the weighting coefficients of the corresponding indicators, and their sum is 1.

[0039] Understandably, the calculation of the geometric distribution factor for a single satellite can be achieved by constructing a direction cosine observation matrix based on the relative position coordinates of other satellites connected to the satellite in the current topology. The covariance matrix of this observation matrix is ​​then calculated, and the square root of the summation of the main diagonal elements is taken. The network average geometric distribution factor is calculated by averaging the geometric distribution factors of all satellites participating in orbit determination within the constellation. The calculation of the minimum hop count from outside the country to inside the country involves constructing a network adjacency graph based on the current topology matrix. All satellites directly connected to domestic ground stations are used as the target node set, and outside satellites are used as source nodes. A graph theory shortest path algorithm is used to calculate the number of edges on the shortest path from each source node to the target node set, and the maximum value is extracted as a penalty term.

[0040] It should be noted that the above-mentioned average geometric distribution factor of the network, the maximum number of hops for satellite transmission from overseas to domestic, and the optimization objective function formed by the two are only illustrative representations of the embodiments of this disclosure. Those skilled in the art can adjust the calculation methods of the above-mentioned average geometric distribution factor of the network and the maximum number of hops for satellite transmission from overseas to domestic according to actual needs. Similarly, the specific form of the optimization objective function can also be adaptively adjusted, which will not be elaborated here.

[0041] The constraints of the above hybrid inter-satellite network optimization model have clear physical meaning: and The visibility and link establishment status of inter-star links must be undirected and bidirectionally symmetrical; The visibility state is limited to a Boolean value; It is the most fundamental physical constraint, which mandates that if there is no spatial visibility between two satellites (i.e., ... If they cannot establish any form of communication link, then they absolutely cannot establish any communication link between them. It must be 0. Considering that the various optimization sub-objectives are independent of each other in engineering implementation, this invention adopts a three-level progressive optimization mode for step-by-step solution.

[0042] S2. Optimize the laser backbone link, evaluate the number of transmission hops and transmission delay based on the differences in satellite connection methods, determine the fixed laser topology link establishment strategy, and thus construct a fixed laser link network.

[0043] Specifically, when determining the fixed laser topology link establishment strategy, based on the distribution characteristics of the satellite orbital plane, inter-satellite continuous visibility, and the coverage elevation angle constraints of the ground station, the maximum number of hops and transmission delay of satellite data transmission from overseas to domestic are evaluated, and satellites are continuously tracked and linked via laser to construct a multi-ring backbone network.

[0044] In some embodiments, reference Figure 2 The diagram illustrates different laser ring-shaped fixed topology connection methods. Medium Earth orbit (MEO) satellites can currently be categorized into three laser link establishment schemes based on their connection methods: one loop formed by a handshake between satellites, four loops formed by linking every two satellites, and two loops formed by linking every one satellite. To ensure the universality of this strategy, a comprehensive assessment is needed when determining the fixed laser topology link establishment strategy, considering the distribution characteristics of the satellite orbital planes, inter-satellite visibility, and the elevation angle constraints of the ground stations. For example, ground stations within the country might be located in specific geographical locations (such as cities A and B), with a ground station visibility angle threshold of 30°. By evaluating the maximum number of hops and transmission delay for data transmission from overseas to domestic satellites under each scheme, it is found that if a link establishment scheme with every one satellite is used, two loops need to be formed in each orbital plane, which severely limits the laser terminal resources onboard the satellite. If a handshake between satellites is used, overseas data often needs to be relayed over long distances through multiple satellites before entering the country, resulting in excessively high end-to-end transmission delays. After comprehensive evaluation and comparison, this embodiment prefers the second scheme shown in the figure, which, for medium Earth orbit satellites, uses a method of establishing links with two satellites spaced apart. The satellites are continuously tracked by laser to establish links, thus constructing four stable ring-shaped backbone networks and completing the laser fixed link network. The construction of the network is as follows. For geostationary orbit or inclined geosynchronous orbit satellites, in order to maintain the robustness of the network across layers, a fixed link establishment method is also adopted to continuously track and connect to the above-mentioned ring network via laser.

[0045] S3. Based on the laser fixed link network, the microwave static link is optimized to determine the link establishment strategy for co-track and heterogeneous static microwave topologies, thereby constructing a microwave static link network to transmit signaling and maintain full network connectivity when the laser link is abnormally interrupted.

[0046] Specifically, when determining the static microwave topology link establishment strategy, the set of continuously visible satellites of the target satellite is obtained based on inter-satellite continuous visibility analysis. A first number of co-orbit microwave links are allocated in the set of continuously visible satellites to maintain co-orbit connectivity. Combined with the evaluation results of the network's average geometric distribution factor performance, a second number of disorbit microwave links are allocated to construct a heterogeneous microwave network.

[0047] In some embodiments, reference Figure 3The diagram showing a continuously visible satellite for microwave links illustrates that, based on complex celestial orbit dynamics and inter-satellite continuous visibility analysis, in a typical BeiDou navigation constellation configuration, any medium Earth orbit (MEO) satellite typically has eight continuously visible neighboring satellites. Of these eight continuously visible satellites, four are in the same orbital plane as the target satellite, while the remaining four are evenly distributed across two other MEO planes. When determining the static microwave topology link establishment strategy, it is necessary to rationally allocate a first number of co-orbit microwave links and a second number of inter-orbit microwave links within this set. Considering that the aforementioned laser ring backbone link has already been deployed within the same orbit, to prevent single-point failures and improve link robustness, this embodiment selects to allocate two co-orbit microwave links within the same orbit to maintain basic connectivity. These links are mainly used for transmitting low-rate control signaling and, in the event of abnormal interruption due to space environment interference, act as an emergency backup to maintain minimum full network connectivity. Furthermore, based on simulation evaluation results of the network's average geometric distribution factor performance, the allocation of the number of inter-orbit microwave links is compared and verified on the basis of the fixed laser links. Analysis shows that when only two microwave fixed heterogeneous links are allocated, the overall geometric distribution factor performance of the constellation deteriorates significantly, failing to meet the requirements of high-precision navigation and positioning. Therefore, considering both connectivity requirements and geometric configuration optimization indicators, this embodiment ultimately determines to construct the constellation by retaining two co-orbit microwave links and four heterogeneous microwave links, thereby outputting a complete microwave static link network. .

[0048] S4. Combining the current network status and satellite visibility, with the network average geometric distribution factor and the maximum number of hops in satellite transmission from overseas to domestic as optimization objectives, a dynamic programming algorithm is used to dynamically adjust and optimize microwave links in real time, generating a dynamic link network. The hybrid inter-satellite network link establishment planning results are then output by combining the laser fixed link network and the microwave static link network.

[0049] In step S4, the dynamic link optimization module, combined with the current network status and satellite visibility, dynamically adjusts and optimizes microwave links in real time to further improve network robustness and fill blind spots in the fixed topology. This is because the dynamic link topology matrix... The network topology optimization requires frequent dynamic changes due to factors such as the opening and closing of visibility windows and the reliability of network nodes. Traditional static solutions are inapplicable; therefore, dynamic programming algorithms must be used for real-time solutions. The local mathematical model expression for this dynamic link topology optimization is as follows:

[0050] The dynamic optimization result is obtained by solving the objective function.

[0051] In some embodiments, a dynamic link optimization algorithm based on simulated annealing is used to solve the local mathematical model of the dynamic link topology optimization in real time.

[0052] Considering that the simulated annealing algorithm has the excellent characteristic of accepting inferior solutions with a certain probability in combinatorial optimization problems, thus effectively escaping the trap of local optima, this embodiment specifically adopts a dynamic link optimization algorithm based on simulated annealing for real-time solution. (Reference) Figure 4 The flowchart shown is for solving dynamic link topology optimization based on simulated annealing algorithm. At the beginning of the algorithm, the satellite set, inter-satellite visibility matrix, and precise position coordinates between satellites are required as input, and the annealing initialization temperature is set. Subsequently, based on the inter-satellite visibility constraints obtained from the current topology planning time segment, and according to the maximum concurrent link establishment requirement specified by the radio frequency terminal of each satellite, an integer programming algorithm is used for preliminary link establishment, thereby obtaining an initial dynamic link establishment matrix that satisfies the physical constraints, denoted as the current state. And calculate its initial cost. .

[0053] Under each fixed temperature condition, the algorithm needs to perform an extensive search of the neighborhood of the current topological solution to find the optimal solution with lower system energy. In a specific implementation, a specially designed dual-link exchange strategy is used to perturb the neighborhood and generate a new solution. The specific implementation process is as follows: Based on the current link establishment matrix, two existing dynamic links (a,b) and (c,d) are randomly selected, meaning satellite a is currently connected to satellite b, and satellite c is connected to satellite d. Then, the system performs visibility and connectivity checks. If it is detected that satellite a and satellite c are mutually visible in the current time segment and have not yet established a link, and simultaneously, it is detected that satellite b and satellite d are also mutually visible and have not yet established a link, then a change in the topology is triggered. The system will disconnect the two existing links and cross-swap the link endpoints, changing the connection to indicate that satellite a and satellite c are connected, and satellite b and satellite d are connected. Through this local perturbation operation, known in graph theory as edge swapping, a new state link establishment matrix is ​​obtained at the current temperature, denoted as [equation missing]. .

[0054] Establishing a chain matrix for generating new states Then, the algorithm engine needs to recalculate the average geometric distribution factor and minimum hop count from outside the territory for each satellite under this topology, and substitute them into the aforementioned optimization objective function to calculate the new generation value. Subsequently, the new generation of value will be... Current generation value before mutation A rigorous comparison is performed to determine the direction of the algorithm's evolution. This decision-making process is executed based on the Metropolis criterion, the core of the simulated annealing algorithm: when When the new network topology generated after the dual-link switching is superior to the previous state in terms of navigation accuracy or transmission latency, the algorithm unconditionally accepts the new state and updates the current state matrix and cost value; when This indicates that the newly generated topology has degraded network performance. To prevent the algorithm from prematurely converging and getting trapped in a local minimum, the algorithm calculates an acceptance probability, which is expressed as follows:

[0055] Where p is the acceptance probability; T is the simulated annealing temperature in the current state.

[0056] After calculating the theoretical probability value, the computing unit generates a uniformly distributed random number within the continuous interval (0,1) using a pseudo-random number generator. The system compares this random number with the calculated acceptance probability p. If the generated random number is strictly less than the acceptance probability p, the system "tolerates" this performance degradation and accepts the new state; otherwise, if the random number is greater than or equal to the probability p, the system decisively rejects the new state and rolls back the chain topology to before the mutation.

[0057] After deciding whether to accept the new status, according to Update the simulated annealing temperature, where, This represents the simulated annealing temperature at the k-th iteration; This represents the simulated annealing temperature at the (k+1)th iteration; It is the temperature decay factor, and satisfies The algorithm continues until the upper limit of iteration is reached, at which point it terminates and the current chain-building matrix is ​​output as the final optimization result.

[0058] The aforementioned neighborhood search, cost calculation, Metropolis probability assessment, and temperature decay process will continuously iterate, with the number of iterations denoted as `loop`, until the system temperature drops to a set threshold or the number of iterations in the outer loop reaches a set iteration limit. At this point, the simulated annealing algorithm ends, and the system outputs the current link establishment matrix as the final dynamic topology optimization result. Finally, the optimized dynamic link network is superimposed and fused with the laser fixed link network and microwave static link network determined in the previous steps to output the complete link establishment planning result of the hybrid inter-satellite network within the current time segment.

[0059] To verify the technical effectiveness of the method proposed in this invention, performance evaluation was conducted using specific simulation scenarios. (Reference) Figure 5The diagram shows the overall planning of the hybrid inter-satellite network link establishment. It illustrates the three-dimensional spatial topology generated using the aforementioned hierarchical progressive optimization method under simulation conditions with a ground station elevation angle constraint of 30°. In the diagram, links marked with red solid lines represent laser backbone links fixed by the scheme; links marked with green solid lines represent microwave fixed links used to maintain basic connectivity; and the interleaved network marked with blue solid lines represents microwave dynamic links obtained through real-time optimization using the simulated annealing algorithm. This multi-layered network structure exhibits a well-defined and complementary structure.

[0060] Further analysis of network performance under this topology is needed, refer to... Figure 6 and Figure 7 These two figures respectively illustrate the geometric distribution factor performance of the hybrid network after adaptive planning and the statistics of the minimum hop count from overseas to domestic locations. The simulation data clearly shows that when the network is in a fully functional and healthy state, the average geometric distribution factor of the entire constellation remains at an excellent level of 1.001. The maximum geometric distribution factor for a single satellite is only 1.132, and the minimum is as low as 0.941, indicating that the constellation possesses extremely high spatial orbit determination geometric accuracy at this point. More importantly, under the scheduling of this planning scheme, all satellite data located in non-visible areas overseas can be accurately matched to routes and return to domestic ground stations in just one hop via high-speed laser inter-satellite links, effectively ensuring extremely low latency for service data backhaul.

[0061] The most significant technical advantage of this solution lies in its self-healing and reconstruction capabilities when the network encounters localized damage or sudden link interruptions. To simulate the harsh real-world space environment, the probability of random abnormal interruptions in various communication links within the hybrid inter-satellite network is set to as high as 20%. Under these stringent fault boundary conditions, the network performance of the traditional pre-established link method and the adaptive link establishment scheme of this invention are compared. (Reference) Figure 8 and Figure 10 This demonstrates the disastrous performance resulting from a large-scale link outage when using a traditional rigid link establishment method. Due to the sudden failure of numerous pre-set links and the inability to reconstruct them in real time, the network topology deteriorated rapidly, causing the average geometric distribution factor (GDP) of the entire network to plummet to 2.149. Some nodes were even in a semi-isolated state, the maximum satellite GDP deteriorated to an unacceptable 4.410, and the minimum satellite GDP also dropped to 1.293. Regarding transmission performance, a significant abnormal peak appeared in the minimum hop count statistics from overseas to domestic locations, indicating that data packets from at least two overseas satellites could no longer return to domestic locations via a single hop. This inevitably required multi-level store-and-forward, severely hindering and fatally impacting the transmission of high-real-time business data and inter-satellite signaling.

[0062] In stark contrast, see reference Figure 9 and Figure 11 This demonstrates the recovery effect of the system after initiating the multi-layer progressive optimization scheme of this invention under the adverse condition of an equal 20% link interruption probability. Because the dynamic link optimization module is rapidly activated upon sensing topology changes, the algorithm can effectively and quickly replenish the blocked critical links in the network by recombination using idle microwave radio frequency resources based on the remaining visibility matrix, resulting in adaptive orbit changes and reorganization of the network topology. After adaptive planning, the network's average geometric distribution factor was forcibly increased and restored to a usable level of 1.136, an improvement of approximately 47.1% compared to the traditional method's 2.149. The maximum satellite geometric distribution factor within the network was suppressed to approximately 1.386, and the minimum satellite geometric distribution factor was restored to approximately 0.986, ensuring that the degraded orbit determination accuracy under damaged conditions still meets the requirements of most navigation tasks. Crucially, Figure 11 Hop count analysis shows that the redesigned topology once again fully meets the stringent requirements for one-hop transmission of all business data or signaling data from overseas to domestic locations, completely eliminating communication silos and demonstrating extremely strong resilience and network robustness.

[0063] At the hardware implementation level, this invention also provides an electronic device for executing the aforementioned multi-layered progressive hybrid inter-satellite network adaptive planning method. This electronic device can be a high-performance onboard computer deployed on a satellite platform or a server cluster located in a ground-based telemetry and control center. The core hardware architecture of this electronic device includes interconnected processors and memory, as well as a system bus for data interaction between internal components. The memory is typically a non-volatile storage medium, such as a high-capacity onboard solid-state drive or flash memory, internally used to persistently store the computer programs and mathematical libraries implementing the various optimization algorithms described above. The processor, as the computing engine, can employ a field-programmable gate array (FPGA) with high-speed floating-point operation capabilities or a dedicated digital signal processing chip. By reading and executing the computer program code stored in the memory, it sequentially completes a series of complex logical operations, including time segment partitioning, matrix construction, laser and microwave static topology strategy allocation, and the most computationally intensive simulated annealing dynamic link optimization, thereby ultimately enabling the electronic device to fully execute the method flow claimed in this invention.

[0064] Furthermore, the technical solution of this invention can also be embodied in the form of a computer-readable storage medium. This computer-readable storage medium internally stores a compiled set of computer program instructions. When these program instructions are loaded and executed by an execution unit such as a central processing unit, they can accurately reproduce the various physical steps and mathematical judgment logic defined in the claims, thereby realizing the multi-layered progressive hybrid inter-satellite network adaptive planning method described above. This storage medium encompasses a wide range of forms, including but not limited to portable hard drives, read-only optical discs, random access memory, magnetic tape, and all other physical media capable of carrying data and instructions in a computer-readable form.

[0065] In this specification, the illustrative expressions of the terms used do not necessarily refer to the same embodiments or examples. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0066] Any modifications, equivalent substitutions, or improvements made within the spirit and principles of this invention shall be included within the scope of protection of this invention.

Claims

1. A multi-layered progressive hybrid inter-satellite network adaptive planning method, characterized in that, include: The entire navigation constellation operation cycle is divided into several time segments of equal duration. The inter-satellite visibility matrix between satellites is obtained in each time segment. A hybrid inter-satellite network optimization model is constructed, and the multi-layer optimization problem is decomposed into three sub-problems: laser fixed topology optimization, microwave static topology optimization, and dynamic topology optimization. The laser backbone link is optimized, and the transmission hop count and transmission delay are evaluated based on the differences in satellite connection methods. A fixed laser topology link establishment strategy is determined to construct a fixed laser link network. Based on the laser fixed link network, the microwave static link is optimized to determine the link establishment strategy for co-track and heterogeneous static microwave topologies, thereby constructing a microwave static link network to transmit signaling and maintain full network connectivity when the laser link is abnormally interrupted. Based on the current network status and satellite visibility, and taking the network average geometric distribution factor and the maximum number of hops in satellite transmission from overseas to domestic locations as optimization objectives, a dynamic programming algorithm is used to dynamically adjust and optimize microwave links in real time, generating a dynamic link network. The hybrid inter-satellite network link establishment planning results are then output by combining the laser fixed link network and the microwave static link network.

2. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 1, characterized in that, In the hybrid inter-satellite network optimization model, the visibility matrix between satellites is defined as follows: The topology matrix of the hybrid inter-satellite network is defined as follows: ,in, This is a fixed link topology matrix, containing microwave fixed link topologies. and laser fixed link topology ; This is a dynamic link topology matrix, containing microwave dynamic link topology. and laser dynamic link topology The expression for decomposing the original global hybrid inter-satellite network link establishment optimization problem using the idea of ​​hierarchical optimization is as follows: in, The objective function is composed of the network's average geometric distribution factor and the maximum number of hops in satellite transmissions from overseas to domestic locations; T represents the transpose matrix operation; i and j are both indices of matrix elements.

3. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 1, characterized in that, When determining the fixed laser topology link establishment strategy, based on the distribution characteristics of the satellite orbital plane, inter-satellite continuous visibility, and ground station coverage elevation angle constraints, the maximum number of hops and transmission delay of satellite data transmission from overseas to domestic are evaluated, and satellites are continuously tracked and linked via laser to construct a multi-ring backbone network.

4. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 1, characterized in that, When determining the static microwave topology link establishment strategy, the set of continuously visible satellites of the target satellite is obtained based on inter-satellite continuous visibility analysis. A first number of co-orbit microwave links are allocated in the set of continuously visible satellites to maintain co-orbit connectivity. Combined with the evaluation results of the network's average geometric distribution factor performance, a second number of disorbit microwave links are allocated to construct a heterogeneous microwave network.

5. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 2, characterized in that, The expression for the model that optimizes the dynamic link topology in the real-time dynamic adjustment and optimization solution of the microwave link using the dynamic programming algorithm is as follows: The dynamic optimization results are obtained by solving the model.

6. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 5, characterized in that, A dynamic link optimization algorithm based on simulated annealing is used to solve the model in real time, including: Set the annealing initialization temperature, and based on the visibility matrix between satellites and the position coordinates between satellites obtained from the current topology planning time segment, use an integer programming algorithm to establish links according to the specified number of links required for each satellite to obtain the initial dynamic link establishment matrix. A dual-link switching strategy is used to search the neighborhood under each temperature condition. Two existing dynamic links (a,b) and (c,d) are randomly selected based on the current link establishment matrix, i.e., satellite a is currently connected to satellite b, and satellite c is connected to satellite d. If satellite a and satellite c are mutually visible and not currently linked, and satellite b and satellite d are mutually visible and not currently linked, the topology is changed to switch the two links, so that satellite a is linked to satellite c and satellite b is linked to satellite d, thus obtaining the new link establishment matrix under the current temperature.

7. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 6, characterized in that, After generating the new state chain matrix, the average geometric distribution factor and minimum hop count from outside the territory for each satellite are calculated, and the new generation value is obtained through objective function calculation. and with current generation values Compare; according to the Metropolis criterion, when When choosing to accept the new state, At that time, the decision on whether to accept the new state is based on the calculated acceptance probability. The method for calculating the acceptance probability is as follows: Where p is the acceptance probability; T is the simulated annealing temperature in the current state.

8. The multi-layer progressive hybrid inter-satellite network adaptive planning method according to claim 7, characterized in that, After deciding whether to accept the new status, according to Update the simulated annealing temperature, where, This represents the simulated annealing temperature at the k-th iteration; This represents the simulated annealing temperature at the (k+1)th iteration; It is the temperature decay factor, and satisfies The algorithm continues until the upper limit of iteration is reached, at which point it terminates and the current chain-building matrix is ​​output as the final optimization result.

9. An electronic device, characterized in that, It includes a processor and a memory, the memory being used to store a computer program, and the processor being used to execute the computer program stored in the memory to cause the electronic device to perform a multi-layer progressive hybrid inter-satellite network adaptive planning method as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements a multi-layer progressive hybrid inter-satellite network adaptive planning method as described in any one of claims 1 to 8.

Images