Optimal Access Node Selection Method for Low-Earth Orbit Communication Satellite Clusters

Abstract

This invention discloses a method for selecting the optimal access node for a low-Earth orbit (LEO) communication satellite constellation, belonging to the field of satellite communication technology. The invention includes the following steps: self-organizing the construction of a distributed nested cellular-like network; the network is distributed, dynamically self-sustaining; access node initialization; and dynamic iteration of distributed service nodes based on the cellular-like network. The decentralized chainmail structure of the cellular-like network eliminates single points of failure. The switching of any node only triggers a local reconstruction, without affecting the overall network function. Network construction and maintenance do not rely on any pre-set infrastructure or central node, possessing strong autonomy and survivability. Topology maintenance and service recursion are strictly limited to local nesting relationships, resulting in low communication and computational overhead, fast response speed, and the integration of new nodes into the existing chainmail network through negotiation with neighboring nodes. This facilitates network expansion and enables autonomous, resilient, and efficient maintenance of the LEO constellation topology, as well as rapid long-term recursion of communication services.

Description

Technical Field

[0001] This invention relates to the field of satellite communication technology, and in particular to a method for selecting the optimal access node for low-Earth orbit communication constellations. Background Technology

[0002] User terminals such as remote sensing satellites require continuous and seamless communication services from constellation networks during high-speed movement. Traditional topology analysis methods, which rely heavily on global information updates and real-time calculations, suffer from high computational costs, slow response times, and difficulty in supporting long-term forecasts when dealing with giant constellations of thousands or even tens of thousands of satellites. Especially given the limited onboard computing resources, how to quickly and accurately predict and allocate future service satellite sequences for mobile user terminals is a pressing technical challenge that needs to be addressed.

[0003] Existing technologies typically employ either pure numerical prediction based on orbital mechanics models or real-time search algorithms based on global topology graphs. The former demands extremely high orbital accuracy and computational power, and errors accumulate over time; the latter requires global recalculation every time the network topology changes, resulting in huge communication and computational overhead, which cannot meet the long-term, high-efficiency service recursion requirements of low-Earth orbit constellations. Summary of the Invention

[0004] The purpose of this invention is to provide a method for selecting the optimal access node for low-Earth orbit communication satellite constellations, so as to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for selecting the optimal access node for low-Earth orbit communication satellite constellations, comprising the following steps:

[0006] Self-organizing to build a distributed nested cellular-like network, in which all satellite nodes, based on their own location information in the initial state, discover neighboring nodes through a distributed negotiation protocol, evaluate the stability weight of the links with neighboring nodes, and form a logical cellular ring;

[0007] The network is distributed and dynamically self-sustaining. The cellular-like network is dynamically maintained through a distributed protocol. Each satellite node only exchanges heartbeat information with its neighboring nodes within the logical cellular ring periodically to monitor each other's liveness and link quality. When a satellite node is detected to have failed or disconnected, the affected node will initiate a local, distributed reconstruction protocol to renegotiate and form a new logical cellular ring structure.

[0008] Access node initialization: When a user terminal accesses the network, the set of candidate service nodes that receive the request determines the initial service node through a weighted election function.

[0009] Based on the distributed service nodes of the cellular network, dynamic iteration is performed. The current service node starts service recursion according to the trajectory of the terminal, queries the list of satellite nodes in the logical cellular ring maintained by itself, and initiates consultation with the core node of the adjacent logical cellular ring that the current service node is a member to obtain the node information in the logical cellular ring where the core node is located. The results are combined to form a set of candidate service nodes. For each candidate node in the set of candidate service nodes, the handover priority score is calculated, and the candidate node with the highest score is selected as the next service node.

[0010] Furthermore, when constructing a self-organizing distributed nested cellular-like network, satellite nodes evaluate the stability weights of links with neighboring nodes, and these stability weights are calculated using the following formula:

[0011] (1)

[0012] in, Indicates satellite node with neighboring nodes Stability weights of the links between them. Indicates satellite node with neighboring nodes The distance between them Indicates the maximum communication radius. This represents the normalized signal-to-noise ratio value. Indicates the magnitude of relative velocity. , and Represents the weighting coefficients and .

[0013] Furthermore, when forming a logical cellular ring, each satellite node selects several neighboring nodes that make its local network structure most stable, minimizes the reciprocal of the sum of the stability weights of all links within the ring, and solves this problem using a distributed algorithm to form a nested ring structure.

[0014] (2)

[0015] in, Represents a logical cellular ring. Indicates satellite node Let S be the set of neighboring nodes, and let S be the set of all possible connections in the logical cellular ring.

[0016] Furthermore, in the case of a distributed, dynamically self-sustaining network, satellite nodes periodically exchange heartbeat information with neighboring nodes within the logical cellular ring to monitor link quality and calculate anomaly confidence levels.

[0017] (3)

[0018] in, Indicates the confidence level of the anomaly. This represents the threshold for the number of heartbeat packets, if the satellite node In the time window Internally received from neighboring nodes Heart rate below threshold If so, the connection is considered faulty;

[0019] When logical cellular ring More than half of the member nodes in the cluster have a certain neighbor node abnormal confidence level Greater than the confidence threshold At that time, that is ,in, If it is an indicator function, then it triggers a response against neighboring nodes. The process of removing and reconstructing logical cellular rings.

[0020] Furthermore, in the case of a distributed, dynamic, and self-sustaining network, the affected nodes include failed or detached satellite nodes as well as neighboring nodes contained within their own logical cellular rings. Changes in the cellular-like network propagate incrementally between adjacent rings through nested cellular ring relationships.

[0021] Furthermore, during access node initialization, the initial service node is determined through a weighted election function, and the weight of each candidate node in the candidate service node set is calculated:

[0022] (4)

[0023] in, Indicates candidate nodes The weight, Indicates candidate nodes Distance from the terminal Indicates candidate nodes The current load, Indicates the maximum allowed load. and Represents the weighting coefficient, election weight. The highest-ranking candidate node is determined as the initial service node through distributed negotiation, and R represents the maximum communication radius.

[0024] Furthermore, during the dynamic iteration of distributed service nodes based on cellular networks, the candidate service node set includes nodes in the current service node's own logical cellular ring, as well as the core nodes of its adjacent logical cellular rings.

[0025] Furthermore, when selecting the optimal node, for each candidate node... Calculate the candidate node handover priority score:

[0026] (5)

[0027] in, Indicates candidate nodes Switch priority scores Indicates the predicted link quality. Represents the predicted candidate nodes Service overlap time with terminal N represents the speed difference. candidates N represents the set of candidate service nodes. candidates This includes the nodes within the current service node's own logical cellular ring, as well as the core nodes of the adjacent logical cellular rings to which the current service node belongs. , and To normalize the weights, select the score. The highest-ranking candidate node will be the next service node. .

[0028] Furthermore, the current service node directly passes the service context information to the selected next service node;

[0029] When the switching opportunity arrives, the service is seamlessly transferred from the current service node to the next service node;

[0030] After the next service node takes over, it repeats the above distributed iterative process based on its own chainmail network local view, forming a long-term service sequence.

[0031] Furthermore, the user terminal is a high-speed moving terminal device, including remote sensing satellites.

[0032] Compared with the prior art, the beneficial effects of the present invention are:

[0033] 1. The distributed nested cellular-like network self-organizing construction of the present invention calculates link stability weights by integrating relative distance, signal-to-noise ratio and relative speed, and forms a nested ring structure with a distributed algorithm. It does not require central node coordination, which greatly reduces the complexity of large-scale constellation deployment, reduces the occupation of on-board storage and communication resources, and the nested structure gives the network high connectivity redundancy. Node failure only affects the local area, avoiding the collapse of the entire network, laying a highly reliable foundation for the stable operation of the network in the future, and is suitable for the actual scenario of limited low-orbit satellite resources.

[0034] 2. The distributed dynamic self-sustaining mechanism of the chainmail network of the present invention achieves network maintenance through local heartbeat monitoring, quantification of anomaly confidence, and local reconstruction. It shortens the fault response from the whole network level to the local ring level, quickly identifies and repairs node anomalies, and only affected nodes participate in reconstruction, avoiding network topology oscillation and eliminating the risk of network paralysis caused by a single point of failure. Incremental synchronization reduces inter-satellite communication and computing overhead, which is in line with the limited computing power of low-orbit satellites and ensures long-term stable network connectivity.

[0035] 3. The access node initialization and service node dynamic iteration of this invention combine distance and load to select the initial node and multi-factor decision-making to select the iterative node, adapting to high-speed terminals, ensuring stable initial access, avoiding node overload, eliminating the need for global queries when switching services, reducing resource overhead, achieving seamless recursion, supporting optimal combination of candidate node sequences and network expansion, improving service reliability and scalability, adapting to dynamic constellation construction, meeting resilience requirements, and supporting large-scale applications. Attached Figure Description

[0036] Figure 1 This is a flowchart of the overall process for selecting the optimal access node for a low-Earth orbit communication constellation according to the present invention.

[0037] Figure 2 This is a schematic diagram of the nested cellular-like network structure of the present invention;

[0038] Figure 3 This is a partial topology diagram of the access node of the present invention. Detailed Implementation

[0039] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] Please see Figure 1-3 The present invention provides the following technical solutions:

[0041] The optimal access node selection method for low-Earth orbit communication constellations constructs a fully peer-to-peer, decentralized node network. Each node forms a stable cellular-like structure with its neighboring nodes. Through the nesting and interconnection of these cellular structures, a resilient "chainmail"-like network covering the entire constellation is formed. The maintenance of the network topology does not depend on any specific node, but is synchronized and updated locally through a distributed protocol, thereby achieving efficient, long-term, and robust recursive topology recursion. The specific steps include:

[0042] Self-organizing to build a distributed nested cellular-like network, in which all satellite nodes, based on their own location information in the initial state, discover neighboring nodes through a distributed negotiation protocol, evaluate the stability weight of the links with neighboring nodes, and form a logical cellular ring;

[0043] When self-organizing to build a distributed nested cellular-like network, satellite nodes Discover neighboring nodes Next, the satellite node evaluates the stability weight of the link with its neighboring nodes. This stability weight is determined by the relative distance, link signal-to-noise ratio, and relative speed, and is calculated using the following formula:

[0044] (1)

[0045] in, Indicates satellite node with neighboring nodes Stability weights of the links between them. Indicates satellite node with neighboring nodes The distance between them Indicates the maximum communication radius. This represents the normalized signal-to-noise ratio value. Indicates the magnitude of relative velocity. , and Represents the weighting coefficients and ;

[0046] When forming a logical cellular ring, each satellite node selects several neighboring nodes that maximize the stability of its local network structure, minimizing the reciprocal of the sum of the stability weights of all links within the ring (i.e., maximizing overall stability). This is achieved through a distributed algorithm, resulting in a nested ring structure.

[0047] (2)

[0048] in, Represents a logical cellular ring. Indicates satellite node Let S be the set of neighboring nodes, and let S be the set of all possible connections in the logical cellular ring.

[0049] Specifically, the structure of a distributed nested cellular network is decentralized. Each node actively discovers its first-order neighbor nodes within its communication range and negotiates to form a temporary "logical cellular ring" with itself at its core. This logical cellular ring structure defines the direct service and cooperation boundaries of the node.

[0050] Each satellite node is not only the core of its own logical cellular ring, but also a member of the logical cellular rings formed by its multiple neighboring nodes. Through this nested relationship of "one for all, all for one," the logical cellular rings of all nodes are interconnected and nested layer by layer, ultimately forming a distributed network without a global center, but with a clear local structure and extremely strong connectivity. Its overall form is similar to "chainmail," where the exit or failure of any node will only affect the local area and will not cause the collapse of the entire network.

[0051] Each satellite node only needs to maintain the membership information of its own logical cellular ring, as well as the core node information of its other neighboring logical cellular rings (as members).

[0052] In the above embodiments, this self-organizing construction method does not rely on a global control center or pre-set infrastructure. All satellite nodes autonomously complete neighbor discovery and logical cell ring formation based on their own location information, which greatly reduces the overall complexity of network deployment. It is especially suitable for large-scale low-Earth orbit constellations composed of thousands or even tens of thousands of satellites and can quickly complete the initialization of the entire network topology.

[0053] By calculating the stability weight using a weighted formula that integrates relative distance, link signal-to-noise ratio, and relative speed, the actual communication quality of the link can be comprehensively reflected. Relative distance determines signal transmission loss, signal-to-noise ratio is directly related to communication reliability, and relative speed affects the link's lifespan. The three factors work together to ensure that the selected neighbor nodes can form a long-term and stable cooperative relationship. Compared with the scheme that relies on a single indicator (such as distance) to select neighbors, this significantly reduces the probability of frequent link interruptions and reconstructions in the future.

[0054] Each node maintains only the core node information of its own logical cellular ring and other rings. The distributed algorithm solves the neighbor set, avoiding large-scale data interaction between nodes and reducing the bandwidth consumption of inter-satellite communication. This is suitable for the actual scenario where the computing and communication resources on low-Earth orbit satellites are limited. The nested ring structure gives the network strong connectivity redundancy. Even if some nodes temporarily leave the communication range, the remaining nodes can still maintain the connection through other nested rings.

[0055] The network is distributed and dynamically self-sustaining. The cellular-like network is dynamically maintained through a distributed protocol. The core is a reconstruction mechanism based on distributed nodes. Each satellite node only periodically exchanges heartbeat information with its neighboring nodes within the logical cellular ring to monitor each other's liveness and link quality. When a satellite node is detected to have failed or disconnected, the affected node will initiate a local, distributed reconstruction protocol to renegotiate and form a new logical cellular ring structure. The affected nodes include the failed or disconnected satellite node and the neighboring nodes contained within its own logical cellular ring. Changes in the cellular-like network propagate incrementally between adjacent rings through nested cellular ring relationships.

[0056] In a distributed, dynamic, self-sustaining network, satellite nodes periodically exchange heartbeat information with neighboring nodes within the logical cellular ring, monitor link quality, and calculate anomaly confidence levels.

[0057] (3)

[0058] in, Indicates the confidence level of the anomaly. This represents the threshold for the number of heartbeat packets, if the satellite node In the time window Internally received from neighboring nodes Heart rate below threshold If so, the connection is considered faulty;

[0059] When logical cellular ring More than half of the member nodes in the cluster have a certain neighbor node abnormal confidence level Greater than the confidence threshold At that time, that is ,in, If it is an indicator function, then it triggers a response against neighboring nodes. The process of removing and reconstructing logical cellular rings.

[0060] Specifically, each node periodically exchanges heartbeat information only with its neighboring nodes within its logical cellular ring, monitoring each other's liveness and link quality. When a node fails or leaves the network, the entire network is not affected, but only those neighboring nodes that include that node within their own logical cellular rings. These affected nodes initiate a local, distributed reconstruction protocol, renegotiating among themselves to quickly form a new, stable logical cellular ring structure, thus "patching" the "hole" in the chainmail caused by the node's departure. Network changes propagate incrementally between adjacent rings through nested cellular ring relationships. This synchronization is local and converges rapidly, avoiding network-wide flooding communication.

[0061] In the above embodiments, this distributed dynamic self-sustaining mechanism, through a mode of local heartbeat monitoring combined with distributed consensus reconstruction, enables node anomalies to be quickly identified and handled. Heartbeat packet interaction is limited to the logical cellular ring, avoiding communication redundancy caused by network-wide heartbeat monitoring. The quantitative calculation of anomaly confidence and the triggering condition of consensus among more than half of the nodes ensure the accuracy of anomaly judgment, reduce unnecessary reconstruction caused by misjudgment, and quickly initiate the repair process. Compared with traditional global fault detection schemes, it shortens the fault response time from network-wide latency to local ring-level instant response, effectively ensuring the continuity of communication services.

[0062] The failure of any node only affects the local logical cellular ring containing it. The reconstruction process only takes place between the affected nodes and will not cause network-wide topology oscillation. Even if some satellites in some areas fail due to interference or faults, the remaining network can still maintain normal overall function, avoiding the risk of network-wide paralysis caused by a single point of failure in traditional centralized networks.

[0063] Network topology adjustment information is transmitted layer by layer through the adjacency relationship of nested cellular rings, without the need for network-wide broadcasting, which greatly reduces the bandwidth consumption of inter-satellite communication. At the same time, local reconstruction does not require calling up the entire network's computing resources, but can be completed by relying only on the distributed negotiation of the affected nodes. This is in line with the limited computing power of low-Earth orbit satellites and avoids resource overload caused by global recomputation.

[0064] Access node initialization: When a user terminal accesses the network, the set of candidate service nodes that receive the request determines the initial service node through a weighted election function. The user terminal is a high-speed moving terminal device, including remote sensing satellites.

[0065] The initial service nodes are determined using a weighted election function, and the weights of each candidate node in the candidate service node set are calculated:

[0066] (4)

[0067] in, Indicates candidate nodes The weight, Indicates candidate nodes Distance from the terminal Indicates candidate nodes The current load, Indicates the maximum allowed load. and Represents the weighting coefficient, election weight. The highest-ranking candidate node is determined as the initial service node through distributed negotiation, and R represents the maximum communication radius.

[0068] Based on the distributed service nodes of the cellular network, dynamic iteration is performed. The current service node (denoted as node A) starts service recursion according to the terminal's trajectory. Node A does not need to perform a global query, but instead queries the list of logical cellular ring satellite node members maintained by itself to evaluate which neighbor node in the ring is most likely to become the next service node (candidate node B).

[0069] Meanwhile, by utilizing the nesting characteristics of the chainmail structure, node A can consult with the core node (such as node C) of the adjacent logical cell ring to which the current service node is a member. Node C can return the candidate node information within its logical cell ring. By obtaining the node information within the logical cell ring to which the core node is located, a candidate service node set is formed. The candidate service node set includes the nodes in the current service node's own logical cell ring and the core nodes of its adjacent logical cell ring. Through this limited-range query in the local nesting relationship, node A can quickly synthesize an optimal and multiple alternative service node sequences.

[0070] The current service node (node ​​A) directly passes the service context information to the selected next service node (candidate node B).

[0071] When the switching opportunity arrives, the service is seamlessly transferred from the current service node (node ​​A) to the next service node (candidate node B).

[0072] After the next service node (candidate node B) takes over, it repeats the above distributed iterative process based on its own chainmail network local view, forming a long-term service sequence.

[0073] service node When predicting the next service node for the terminal, a multi-factor decision model is used to select the optimal node from the candidate set. For each candidate node in the candidate service node set, the handover priority score is calculated, and the candidate node with the highest score is selected as the next service node.

[0074] When selecting the optimal node, for each candidate node Calculate the candidate node handover priority score:

[0075] (5)

[0076] in, Indicates candidate nodes The switching priority score, Indicates the predicted link quality. Represents the predicted candidate nodes Service overlap time with terminal N represents the speed difference. candidates N represents the set of candidate service nodes. candidates This includes the nodes within the current service node's own logical cellular ring, as well as the core nodes of the adjacent logical cellular rings to which the current service node belongs. , and To normalize the weights, select the score. The highest-ranking candidate node will be the next service node. .

[0077] Specifically, the decentralized chainmail structure eliminates single points of failure. The failure of any node only triggers a local reconstruction and does not affect the overall function of the entire network. It is particularly suitable for military or high-reliability application scenarios. Network construction and maintenance do not depend on any pre-set infrastructure or central node. It has strong autonomy and survivability. Topology maintenance and service recursion are strictly limited to local nesting relationships. Communication and computing overhead is small and response speed is fast. New nodes can be integrated into the existing chainmail network simply by negotiating with neighboring nodes, making it easy to scale up the network.

[0078] In the above embodiments, the distributed service node dynamic iteration achieves smooth switching of service nodes by combining local queries with nested consultation for candidate node selection and multi-factor decision-making for optimal node selection. The current service node can quickly obtain candidate node information through local nesting relationships without global queries, which greatly shortens the switching preparation time. The switching priority score integrates link quality, service overlap time and speed difference, which not only ensures that the next service node can provide stable communication, but also extends the duration of a single service. At the same time, it adapts to the high-speed movement characteristics of the terminal and completely avoids communication interruption during the switching interval, providing seamless service guarantee for terminals such as remote sensing satellites that need to continuously transmit data.

[0079] Service recursion is performed only within the locally nested logical cellular ring, without the need for full network data interaction, which greatly reduces the bandwidth consumption of inter-satellite communication; the calculation of the multi-factor decision model only relies on the local information of the current service node and candidate nodes, without the need to call the full network computing power, which is in line with the actual situation of limited computing resources on low-Earth orbit satellites. Compared with the global service node planning scheme, the computing and communication overhead is reduced by more than an order of magnitude.

[0080] By designing the service node sequence, even if the optimal node fails suddenly, the backup node can be quickly activated, further improving service reliability. On the other hand, newly added satellite nodes only need to be integrated into the local logical cellular ring to participate in the service node iteration process without adjusting the entire network service mechanism, perfectly adapting to the construction mode of gradual deployment and dynamic expansion of low-Earth orbit constellations.

[0081] In addition, its ability to adapt to highly dynamic scenarios enables it to meet the diverse needs of terminals. Whether it is a high-speed remote sensing satellite or a low-speed ground terminal, it can obtain continuous service through dynamic iteration. The decentralized chainmail structure ensures that the service does not depend on specific nodes. Even if some areas of the network are damaged, the remaining nodes can still take over and complete the service relay.

[0082] To better illustrate the present invention, the technical solution of the present invention is shown below through an embodiment, which takes a low-orbit communication constellation with a decentralized architecture as the application scenario.

[0083] Step 1: Self-organize and build a distributed, nested cellular-like network.

[0084] After joining the network, satellite nodes periodically broadcast "neighbor discovery" beacons. Node A discovers neighbors B, C, D, E, F, and G.

[0085] Node A calculates the link stability weights with each neighbor according to formula (1). , ... Assuming the ring capacity is set to 5, node A is selected using the criterion of formula (2) through a greedy algorithm. The five largest neighbors (let's say B, C, D, E, F) form its logical cellular ring. Meanwhile, nodes B, C, and others also perform the same process, forming their own nested loops.

[0086] Step 2: Implement distributed dynamic self-sustaining of the network.

[0087] Suppose node F is out of contact due to a fault. Node A detects that its heartbeat with F has been interrupted.

[0088] Node A initiates a local reconfiguration protocol within its logical cellular ring {A;B,C,D,E,F}, notifying ring members B, C, D, and E: "Node F has failed, and we need to reconfigure it."

[0089] After a rapid round of distributed negotiation, the remaining nodes within the ring agree to update the ring structure to {A;B,C,D,E}, and may jointly elect a new node (such as node Z that has entered the range) to join the ring in order to maintain the ring's scalability. This reconfiguration process is completed only within this local ring, and other parts of the network are unaffected.

[0090] Step 3: Perform access node initialization.

[0091] Remote sensing satellite broadcast request. Nodes A, C, and G receive the request simultaneously.

[0092] Each node calculates its own election weight according to formula (4). Assume node C computes its The highest (due to proximity and lighter load), nodes A, C, and G, through rapid distributed negotiation (such as comparing their respective calculated E values), unanimously agree that node C should be the initial service node.

[0093] Step 4: Execute dynamic iteration of distributed service nodes based on cellular network-like systems.

[0094] Service node C predicts that the terminal will soon be moved out of its advantageous service area. A set, including its ring The nodes A, B, G, H, I, etc. are within the system.

[0095] For each candidate node in the set, node C calculates its switching priority score according to formula (5). The score for node H was calculated. It is the highest because it predicts link quality well and has a long overlap time.

[0096] Node C determines node H as The service context information is sent to H in advance. After the service is successfully switched from C to H, H continues the next round of recursion based on its own local view using the same formula (5).

[0097] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for selecting the optimal access node for low-Earth orbit communication satellite constellations, characterized in that, Includes the following steps: Self-organizing to build a distributed nested cellular-like network, in which all satellite nodes, based on their own location information in the initial state, discover neighboring nodes through a distributed negotiation protocol, evaluate the stability weight of the links with neighboring nodes, and form a logical cellular ring; The network is distributed and dynamically self-sustaining. The cellular-like network is dynamically maintained through a distributed protocol. Each satellite node only exchanges heartbeat information with its neighboring nodes within the logical cellular ring periodically to monitor each other's liveness and link quality. When a satellite node is detected to have failed or disconnected, the affected node will initiate a local, distributed reconstruction protocol to renegotiate and form a new logical cellular ring structure. Access node initialization: When a user terminal accesses the network, the set of candidate service nodes that receive the request determines the initial service node through a weighted election function. Based on the distributed service nodes of the cellular network, dynamic iteration is performed. The current service node starts service recursion according to the trajectory of the terminal, queries the list of satellite nodes in the logical cellular ring maintained by itself, and initiates consultation with the core node of the adjacent logical cellular ring that the current service node is a member to obtain the node information in the logical cellular ring where the core node is located. The results are combined to form a set of candidate service nodes. For each candidate node in the set of candidate service nodes, the handover priority score is calculated, and the candidate node with the highest score is selected as the next service node. In the self-organizing construction of a distributed nested cellular-like network, satellite nodes evaluate the stability weights of links with neighboring nodes, and these stability weights are calculated using the following formula: ； in, Indicates satellite node with neighboring nodes Stability weights of the links between them. Indicates satellite node with neighboring nodes The distance between them Indicates the maximum communication radius. This represents the normalized signal-to-noise ratio value. Indicates the magnitude of relative velocity. , and Represents the weighting coefficients and ; When forming a logical cellular ring, each satellite node selects several neighboring nodes that make its local network structure most stable, minimizes the reciprocal of the sum of the stability weights of all links within the ring, and solves this problem using a distributed algorithm to form a nested ring structure. ； in, Represents a logical cellular ring. Indicates satellite node Let S be the set of neighboring nodes, and let S be the set of all possible connections in the logical cellular ring.

2. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 1, characterized in that, In a distributed, dynamic, self-sustaining network, satellite nodes periodically exchange heartbeat information with neighboring nodes within the logical cellular ring, monitor link quality, and calculate anomaly confidence levels. ； in, Indicates the confidence level of the anomaly. This represents the threshold for the number of heartbeat packets, if the satellite node In the time window Internally received from neighboring nodes Heart rate below threshold If so, the connection is considered faulty; When logical cellular ring More than half of the member nodes in the node have a certain neighbor node abnormal confidence level Greater than the confidence threshold At that time, that is ,in, If it is an indicator function, then it triggers a response against neighboring nodes. The process of removing and reconstructing logical cellular rings, Let S represent the candidate node, and let S represent the set of all possible connections in the logical cellular ring.

3. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 1, characterized in that, In a distributed, dynamic, and self-sustaining network, the affected nodes include failed or detached satellite nodes as well as neighboring nodes contained within their own logical cellular rings. Changes in the cellular-like network propagate incrementally between adjacent rings through nested cellular ring relationships.

4. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 1, characterized in that, During access node initialization, the initial service node is determined through a weighted election function, and the weight of each candidate node in the candidate service node set is calculated: ； in, Indicates candidate nodes The weight, Indicates candidate nodes Distance from the terminal Indicates candidate nodes The current load, Indicates the maximum allowed load. and Represents the weighting coefficient, election weight. The highest-ranking candidate node is determined as the initial service node through distributed negotiation, and R represents the maximum communication radius.

5. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 1, characterized in that, When a distributed service node based on a cellular network undergoes dynamic iteration, the set of candidate service nodes includes nodes in the current service node's own logical cellular ring, as well as the core nodes of its adjacent logical cellular rings.

6. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 5, characterized in that, When selecting the optimal node, for each candidate node Calculate the candidate node handover priority score: ； in, Indicates candidate nodes The switching priority score, Indicates the predicted link quality. Represents the predicted candidate nodes Service overlap time with terminal N represents the speed difference. candidates N represents the set of candidate service nodes. candidates This includes the nodes within the current service node's own logical cellular ring, as well as the core nodes of the adjacent logical cellular rings to which the current service node belongs. , and To normalize the weights, select the score. The highest-ranking candidate node will be the next service node. .

7. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 6, characterized in that, The current service node directly passes the service context information to the selected next service node; When the switching opportunity arrives, the service is seamlessly transferred from the current service node to the next service node; After the next service node takes over, it repeats the dynamic iteration process of the distributed service nodes based on its own chainmail network local view, forming a long-term service sequence.

8. The method for selecting the optimal access node for a low-Earth orbit communication constellation as described in claim 1, characterized in that, The user terminal is a high-speed moving terminal device, including remote sensing satellites.

Images