Topological and spatio-temporal routing fast emergency recovery method under navigation satellite network failure

Abstract

The application discloses a kind of topological structure and space route quick emergency recovery method under navigation satellite network failure, it is suitable for single star configuration quick switching inter-satellite measurement terminal, single time slot to the time slot navigation satellite network of multiple one planned chain object.Failure, obtain fault state information and the node failure and inter-satellite link failure are mapped as the planned inter-satellite link unavailability of relevant time slot;Healthy navigation satellite participating in emergency recovery executes route emergency disposal and deterministic distributed topology incremental reconstruction in parallel, prohibits failure-related forwarding and caches data for the affected time slot, while maintaining the original planned link unaffected by failure.In the presence of candidate nodes that meet the conditions, form emergency new inter-satellite link, only when the opposite end is the destination node, form one-hop direct relationship, and enable at the same effective time;The rest of the cached data continues to hop by hop forwarding according to the original routing table item in subsequent time slots.

Description

Technical Field

[0001] This invention relates to the field of dynamic topology control and time-varying routing technology for satellite navigation networks, specifically to a rapid emergency recovery method for topology and spatiotemporal routing under navigation satellite network failures.

[0002] This invention is particularly applicable to the following navigation satellite networks: each navigation satellite is equipped with an inter-satellite measurement terminal that can quickly switch link establishment targets between different time slots, with time slots as the basic time unit for establishing inter-satellite links, and the same navigation satellite can establish such inter-satellite links with at most one planned target in the same time slot; each navigation satellite can dynamically establish links with different adjacent satellites in multiple time slots, and form a spatiotemporal route that utilizes link opportunities in different time slots based on a pre-determined link timing plan. Background Technology

[0003] Global Navigation Satellite System (GNSS) is a crucial space infrastructure providing positioning, navigation, and timing services. To reduce reliance on continuous visibility of ground stations and enhance the autonomous operation of the constellation, modern navigation satellite systems are increasingly equipped with inter-satellite links (ISL). ISLs are used to perform tasks such as precise ranging between satellites, time synchronization, and exchange of operational status and service data. The quantity and geometric diversity of inter-satellite measurement data directly affect the orbit determination and overall constellation time synchronization performance of navigation satellites. Therefore, navigation satellites typically need to establish inter-satellite links with as many different neighboring satellites as possible within a certain time frame to acquire abundant inter-satellite measurement data.

[0004] Constrained by resources such as onboard payload, antenna aperture, radio frequency channels, power consumption, and pointing control capabilities, the navigation satellite network targeted in this invention is configured with only one inter-satellite measurement terminal capable of rapidly switching link establishment targets per navigation satellite. A time-slot-based dynamic link establishment method is employed: within a time slot, a navigation satellite establishes an inter-satellite link with at most one target; in adjacent or subsequent time slots, the navigation satellite can switch to other visible satellites to establish links. By dynamically switching between different link establishment targets in consecutive time slots, navigation satellites can increase the number and geometric diversity of inter-satellite measurement targets, thereby providing richer observational information for precise orbit determination and full constellation time synchronization.

[0005] The aforementioned link establishment mechanism gives navigation satellite networks a significant time-varying characteristic, distinguishing them from continuous multi-adjacent communication networks. Since a single navigation satellite has at most one link establishment target within any given time slot, the network topology formed within a single time slot typically contains only a few discrete inter-satellite links. It does not require the formation of an instantaneous end-to-end multi-hop connectivity path covering all source and destination nodes within that time slot. End-to-end data transmission can be completed progressively across multiple consecutive time slots: data is forwarded from the current satellite to its current link establishment target in a certain time slot, and then continues to be transmitted to the destination node in subsequent time slots as the link establishment target switches. Therefore, the feasible spatiotemporal path from the same source node to the same destination node changes with time slots.

[0006] In such time-varying navigation satellite networks, link timing plans are typically pre-generated based on satellite orbits, visibility relationships, ranging requirements, communication needs, and terminal resource constraints, and further, spatiotemporal routes are pre-calculated accordingly. Since the planned link establishment target for the current time slot may not correspond to a suitable forwarding direction for a specific destination node, even if the current inter-satellite link can be established normally, data destined for a specific destination node may not have a suitable forwarding opportunity in that time slot. In this case, the data can be stored in the onboard cache, waiting for subsequent time slots to switch to a planned link establishment target that can continue to propel the data to the destination node before forwarding, forming a store-carry-forward data transmission method.

[0007] When a navigation satellite node fails, all inter-satellite links planned to establish links with that failed satellite cannot be established normally. Similarly, when an inter-satellite link fails, the planned link establishment cannot be executed during the duration of the failure. For the single fast-switching inter-satellite measurement terminal scenario addressed by this invention, there is no second parallel inter-satellite link of this type available in the same time slot for a healthy navigation satellite to continue the original plan. Therefore, whether the failure is due to a peer node failure or a failure of the planned inter-satellite link itself, the network layer's immediate handling of the affected healthy navigation satellite manifests as the unavailability of the planned inter-satellite link in the relevant time slot. Consequently, data originally planned to be forwarded via this link cannot be transmitted as intended; if the original forwarding plan continues, it may result in invalid transmission, data loss, or continuous data backlog in the direction adjacent to the failure.

[0008] Existing satellite network fault recovery methods typically employ two approaches:

[0009] The first type involves replanning the topology globally or over a larger area after a failure occurs, and then recalculating the routes from the source node to the destination node on the reconstructed topology. This type of method can re-obtain a topology and routes with better overall performance based on the network state after a failure, but it usually requires large-scale state collection, topology optimization, and route calculation, resulting in a relatively long processing time. It is more suitable as a long-term recovery method after a failure than as a rapid emergency response to be implemented immediately after a failure occurs.

[0010] The second type involves immediately finding a new multi-hop forwarding path to bypass the faulty object after detecting a link or node failure, through methods such as local fast rerouting, backup path switching, or incremental path search. This type of method typically relies on the ability of a single node to maintain multiple adjacent links simultaneously and high instantaneous network connectivity, thus making it highly applicable in communication satellite networks with multiple parallel adjacent links.

[0011] However, for the navigation satellite network targeted by this invention, the above-described processing method does not fully utilize its cross-time-slot dynamic link establishment characteristics formed by diverse inter-satellite measurement needs. On the one hand, since each navigation satellite is only equipped with one fast-switching inter-satellite measurement terminal, there is at most one planned link establishment object in the same time slot. Therefore, when the original planned inter-satellite link in the relevant time slot fails, the healthy navigation satellite does not have a second parallel inter-satellite link of the same type available for alternative forwarding in that time slot, and there is no physical basis to immediately search for a new parallel adjacent link in that time slot. On the other hand, the failure of the original planned inter-satellite link in the relevant time slot does not mean that the same source node to the destination node will remain unreachable in subsequent time slots. As the navigation satellite switches to other measurement objects in subsequent time slots, feasible forwarding opportunities that do not pass through the failed object may reappear in the originally pre-calculated spatiotemporal route. Therefore, if the spatiotemporal route of the entire network is searched again immediately after a failure occurs, it will introduce a large on-board computing overhead, and these calculations are not always necessary in the emergency phase.

[0012] Furthermore, if a centralized ground-based approach is used to complete global topology and route reconstruction, it requires processes such as fault status feedback, network-wide status aggregation, global optimization calculation, result uploading, and network-wide synchronous switching. If global optimization and full-source-full-destination route recalculation are performed directly on-board, it will be constrained by the on-board processor's computing power, storage capacity, and power consumption. Therefore, there is an objective emergency response phase between the occurrence of a fault and the long-term availability of global optimization results, which requires rapid maintenance of network operation.

[0013] Therefore, it is necessary to design a fault emergency recovery method that is computationally efficient, fast-responding, can be distributed and executed by each navigation satellite, and can maintain consistent results across the entire network, taking into account the single-terminal time-slotted dynamic link establishment characteristics and cross-time-slot spatiotemporal routing characteristics of navigation satellite networks. Without immediately pursuing the globally optimal topology and optimal route after the fault, it prioritizes blocking non-executable fault-related forwarding behaviors, utilizes on-board buffers to wait for the opportunity of normal spatiotemporal paths in subsequent time slots, and performs lightweight topology compensation for the link holes caused by the fault, thereby maintaining the normal operation of the navigation satellite network as soon as possible before the long-term global recovery is completed. Summary of the Invention

[0014] (a) Purpose of the invention

[0015] The purpose of this invention is to overcome the problems of high computational overhead, long recovery process, and insufficient utilization of subsequent time slot spatiotemporal path opportunities in existing satellite network fault recovery methods under the scenario of time slotted inter-satellite links of navigation satellites, and to provide a rapid emergency recovery method for topology and spatiotemporal routing under navigation satellite network faults.

[0016] This invention does not aim to obtain the globally optimal topology and globally optimal route immediately after a failure occurs. Instead, it focuses on the emergency phase from the occurrence of a failure until the long-term global recovery results are available. Under the premise of ensuring that the existing links and routing results unaffected by the failure remain as unchanged as possible, it enables the navigation satellite network to recover to an operational state as quickly as possible with lower on-board computing and control overhead by suppressing failure-related forwarding behavior in real time, caching data using a "store-carry-forward" mechanism, performing deterministic distributed incremental topology reconstruction in parallel, and activating new links at a unified time. It also compensates for the performance loss caused by the failure to a certain extent.

[0017] (II) Technical Solution

[0018] To achieve the above objectives, the present invention adopts the following technical solution. The present invention provides a rapid emergency recovery method for topology and spatiotemporal routing under navigation satellite network failure conditions. Each navigation satellite in the navigation satellite network is configured with an inter-satellite measurement terminal capable of rapidly switching link establishment targets. Time slots are used as the unit for establishing inter-satellite links, ensuring that each navigation satellite has at most one planned link establishment target in the same time slot. Under normal conditions, the navigation satellite network has a pre-determined link timing plan. This plan enables the same navigation satellite to establish inter-satellite links with different adjacent navigation satellites in multiple time slots, and a spatiotemporal route utilizing link opportunities in different time slots is pre-formed based on the link timing plan. Each navigation satellite has a local routing execution table indexed by time slot and destination node, and the local routing execution table consists of multiple routing table entries.

[0019] Furthermore, a single routing table entry in the local routing execution table includes a time slot identifier, a destination node identifier, a planned link establishment object identifier, and a forwarding flag. For the same time slot of the same navigation satellite, when a planned link establishment object exists, different destination nodes correspond to the same planned link establishment object, and the forwarding flag indicates whether data destined for each destination node is sent via the planned link establishment object. When the forwarding flag uses binary encoding, a forwarding-prohibited state is represented by 0, and a forwarding-allowed state is represented by 1.

[0020] S1. Obtain fault status information and map node faults and inter-satellite link faults to the unavailability of planned inter-satellite links in the relevant time slots.

[0021] Furthermore, the fault status information is propagated hop-by-hop to the healthy navigation satellites participating in emergency recovery using the inter-satellite links in the link timing plan that was still available before the fault occurred; each healthy navigation satellite participating in emergency recovery determines a consistent fault input based on the fault object identifier carried in the fault status information, as well as at least one of the fault event reference time and fault status version number.

[0022] Furthermore, for node failures, planned inter-satellite links with the failed node as the target for planned link establishment are marked as unavailable, and the failed node is removed from the set of effective relay nodes, the set of candidate link establishment targets, and the set of effective destination nodes during the emergency phase; for inter-satellite link failures, planned inter-satellite links that have failed are marked as unavailable during the corresponding failure duration time slot; when the peer of the planned inter-satellite link in the relevant time slot is a failed navigation satellite, or when the planned inter-satellite link itself fails, the same failure suppression processing is applied to the forwarding behavior of each effective destination node in that time slot.

[0023] S2. After the health navigation satellite participating in emergency recovery obtains the fault status information, it performs route emergency handling and deterministic distributed topology incremental reconstruction in parallel.

[0024] The routing emergency handling includes: for affected time slots where planned inter-satellite links are unavailable due to node failures and inter-satellite link failures, without recalculating new multi-hop spatiotemporal routes, setting the corresponding forwarding flags for valid destination nodes within the affected time slots to a prohibited forwarding state, storing the data to be sent in a local cache, and keeping the original routing table entries unaffected by the node failures and inter-satellite link failures unchanged.

[0025] The deterministic distributed topology incremental reconstruction includes: healthy navigation satellites participating in emergency recovery independently calculate based on consistent link timing plans, fault status information, and candidate link feasibility information, according to the same deterministic reconstruction rules, keeping the originally planned inter-satellite links unaffected by the node faults and inter-satellite link faults unchanged, performing alternative link selection for healthy navigation satellites that have lost their originally planned link establishment targets due to faults, and forming emergency new inter-satellite links when there are candidate navigation satellites that meet the link establishment conditions, thereby obtaining a consistent emergency topology.

[0026] Furthermore, for healthy navigation satellites that have lost their original planned link establishment targets due to node failure and inter-satellite link failure, an alternative link establishment candidate set is constructed. The alternative link establishment candidate set consists of candidate navigation satellites that meet all of the following conditions: they have the conditions to establish an inter-satellite link with the healthy navigation satellite in the corresponding time slot; they are not faulty nodes; their configured fast-switching inter-satellite measurement terminals are not occupied in the corresponding time slot; and they are not the healthy navigation satellite itself.

[0027] Furthermore, each healthy navigation satellite participating in emergency recovery follows the same affected time slot processing order and affected navigation satellite processing order. When the alternative link establishment candidate set is not empty, a deterministic candidate selection rule is applied to the alternative link establishment candidate set to obtain a unique alternative link establishment object. The deterministic candidate selection rule is based on the following performance compensation factors: increasing ranging neighbor diversity, increasing inter-satellite links across ground capability categories, and restoring effective inter-satellite links in the original fault time slot. When the performance compensation factors are the same, deterministic disambiguation is performed using a pre-agreed node identifier order.

[0028] Furthermore, after selecting a candidate navigation satellite from the alternative link establishment candidate set as the alternative link establishment object and forming an emergency new inter-satellite link with the healthy navigation satellite, both the healthy navigation satellite and the alternative link establishment object are marked as having their corresponding time slots occupied, so that subsequent deterministic distributed topology incremental reconstruction will no longer consider them as idle candidate link establishment objects for the corresponding time slots.

[0029] S3. For the affected time slots that form emergency new inter-satellite links, establish link objects according to the emergency topology update plan; when the peer node of the emergency new inter-satellite link is the destination node in the local routing execution table, set the corresponding forwarding flag to the forwarding allowed state, and perform one-hop direct forwarding through the emergency new inter-satellite link; when the peer node of the emergency new inter-satellite link is not the destination node, do not use the emergency new inter-satellite link to derive new multi-hop routes, and maintain the cached state of the corresponding data.

[0030] S4. When the emergency topology unification takes effect, the newly added emergency inter-satellite link and the corresponding one-hop direct forwarding relationship are activated, and the emergency topology unification takes effect at the time of alignment with the predetermined synchronization switching boundary of the navigation satellite network.

[0031] Furthermore, the emergency topology unification take effect time is determined based on the fault event reference time and the preset upper bound of fault state propagation, upper bound of topology incremental reconstruction calculation, and synchronization protection margin; the predetermined synchronization switching boundary includes the unified time slot boundary and the frame boundary; the operation of setting the forwarding flag to the prohibited forwarding state takes effect immediately after each participating healthy navigation satellite obtains the corresponding fault state information, and the emergency newly added inter-satellite link and the corresponding one-hop direct forwarding relationship take effect at the emergency topology unification take effect time.

[0032] S5. After the emergency topology takes effect, for data that is still in the cache state, when the time progresses to the subsequent time slot, and the planned link establishment object indicated by the original routing table entry of the subsequent time slot in the local routing execution table is in the normal state and the forwarding flag of the original routing table entry is in the allowed forwarding state, the data to be sent stored in the local cache in step S2 is forwarded hop by hop according to the original routing table entry.

[0033] Furthermore, while executing the rapid emergency recovery method, the fault status information is provided to at least one of the ground computing equipment and the on-board computing node with global computing capabilities, and long-term global link timing plan reconstruction and full spatiotemporal route recalculation based on the reconstructed link timing plan are performed; after the long-term global link timing plan and the full spatiotemporal route calculation are completed and distributed, the navigation satellite network is switched from emergency state to long-term global recovery state at the unified effective time of long-term global recovery.

[0034] (III) Beneficial Effects

[0035] (1) This invention fully utilizes the operational characteristics of navigation satellites dynamically switching link establishment objects between consecutive time slots in order to obtain diverse inter-satellite measurement data. When the planned inter-satellite link of a relevant time slot fails, it is not equated with the source node and the destination node being unreachable for a long time. Instead, it is cached and waits for the normal spatiotemporal path opportunity in subsequent time slots, thereby avoiding path recalculation during the emergency phase.

[0036] (2) On the routing side, the present invention only needs to match the fault-related planned link establishment objects, modify the forwarding flag and manage the cached data, without re-executing the full source-full destination multi-hop spatiotemporal route search; on the topology side, it only adds inter-satellite links to the incremental range of fault impact, while keeping the unaffected links unchanged, thereby significantly reducing the on-board computation of fault emergency recovery.

[0037] (3) After the navigation satellite obtains the fault status information, the present invention can immediately set the fault-related forwarding flag to prohibit forwarding, which can block invalid transmission in time before the topology supplement chain is unified and effective and before the long-term global reconstruction is completed, thereby reducing the risk of data loss and continuous backlog of fault direction data.

[0038] (4) The present invention adopts consistent input, unified processing order and deterministic candidate selection rules, enabling each navigation satellite to independently perform topology incremental reconstruction to obtain the same emergency topology of the whole network, reducing the control overhead required for negotiating new links one by one after a fault, and is suitable for on-board distributed operation.

[0039] (5) The incremental topology reconstruction of this invention, based on the failure of fault-related links, keeps all planned inter-satellite links unaffected by the fault unchanged, and only adds replacement inter-satellite links for the link gaps caused by the fault. Therefore, the original spatiotemporal routes corresponding to the links unaffected by the fault do not need to be changed with the topology reconstruction; the routing side only needs to suppress forwarding in the fault-related routing table entries and incrementally update the directly determinable one-hop forwarding relationship between the two ends of the newly added emergency inter-satellite links, without using the newly added links to re-derive complex multi-hop routes in the emergency phase. Thus, the topology reconstruction itself will not trigger a large-scale change in the entire network routing, and while compensating for the loss of connectivity caused by the fault to a certain extent, it further reduces the computational load and modification scope of route emergency recovery.

[0040] (6) This invention separates the "instant routing protection after receiving fault information" and the "emergency topology switching after unified effective time", so that each satellite can stop erroneous forwarding as early as possible, while ensuring that both ends of the newly added emergency inter-satellite link perform consistent link establishment actions at the same time boundary.

[0041] (7) The rapid emergency recovery described in this invention does not conflict with subsequent global topology optimization and complete spatiotemporal route recalculation. It can serve as a transitional operation mechanism before the long-term global recovery results are available. While prioritizing the network to resume normal operation as soon as possible, it reserves an interface for subsequent global performance optimization. Attached Figure Description

[0042] Figure 1 This is a schematic diagram illustrating the principle of time-slotted link establishment and the change of source-destination spatiotemporal path over time in the navigation satellite network to which this invention applies.

[0043] Figure 2 This is a schematic diagram of the local routing execution table for navigation satellites under normal conditions and the mechanism of storage-carry-forwarding based on the execution table.

[0044] Figure 3 This is a schematic diagram of the overall process of the rapid emergency recovery method for topology and spatiotemporal routing under navigation satellite network failure according to the present invention.

[0045] Figure 4 This diagram illustrates how, after a fault occurs, the routing forwarding flag of the affected time slot is set to 0, the data is buffered, and subsequent time slots continue to forward data according to the original spatiotemporal route.

[0046] Figure 5 This is a schematic diagram of distributed topology incremental reconstruction and non-faulty topology preservation based on consistent input and deterministic rules.

[0047] Figure 6 This diagram illustrates the time relationship between immediate route protection after a failure, unified emergency topology activation, and subsequent long-term global recovery. Detailed Implementation

[0048] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the following embodiments are used to explain the technical principles and implementation process of the present invention, and are not intended to limit the scope of protection of the present invention. Without departing from the core technical concept of the present invention, equivalent substitutions for node numbering, time slot length, link operating frequency band, fault duration, candidate sorting parameters, global recovery algorithm, and specific message format can all adopt the same or equivalent implementation methods.

[0049] Implementation prerequisites, node set, and basic operating conditions:

[0050] Let V be the set of navigation satellite nodes in the navigation satellite network. Each navigation satellite has a unified time reference for the entire network and uses discrete time slots. As the basic time unit for inter-satellite link establishment and switching in this invention, each navigation satellite is equipped with an inter-satellite measurement terminal capable of rapidly switching the link establishment target according to time slots. Therefore, for any navigation satellite... and any time slot It has at most one planned link establishment object of this type of inter-satellite link in that time slot.

[0051] The fast-switching inter-satellite measurement terminal can be a phased array terminal or other inter-satellite measurement terminal that can meet the link establishment constraint of at most one object per time slot. In the general implementation of the present invention, the carrier frequency band, beamforming method or specific time slot length is not limited; in the typical implementation of the navigation satellite network described later, a Ka-band phased array inter-satellite link terminal and a typical 3-second time slot can be used.

[0052] Each navigation satellite obtains a link timing plan in advance during normal operation. This link timing plan can be generated based on satellite orbits, inter-satellite geometric visibility, terminal capabilities, inter-satellite measurement requirements, communication requirements, and other resource constraints. The generation method for the link timing plan can employ matching algorithms, integer programming, heuristic methods, or other topology planning algorithms; this implementation does not limit the specific generation algorithm.

[0053] like Figure 1 As shown, to acquire diverse inter-satellite measurement data for precise orbit determination and full constellation time synchronization, the same navigation satellite can switch to different adjacent satellites to establish links in different time slots. Since each navigation satellite has at most one link establishment target within the same time slot, the instantaneous topology in a single time slot does not require end-to-end connectivity across the entire network; a complete data transmission path can be gradually formed across multiple consecutive time slots. For example, data from source node S to destination node D can be transmitted in time slots. Transmitted from S to A, in time slot Stored in A, in time slot Transmitted from A to C, and then from C to D in a subsequent time slot. If sent from a different starting time slot, the same source-destination node pair may also be transmitted via another set of relay nodes.

[0054] Therefore, in the navigation satellite network targeted by this invention, "the current time slot cannot continue to forward" and "the source-destination node will remain unreachable in the future" are not the same thing. The aforementioned diversified link establishment characteristics across time slots are the network basis for this invention to maintain the original spatiotemporal route and cache and wait for subsequent normal forwarding opportunities after a failure, without immediately re-searching for new multi-hop obstacle bypass paths.

[0055] Local routing execution table and hop-by-hop execution in normal state:

[0056] Under normal conditions, the spatiotemporal route from the source node to the destination node can be calculated based on a pre-determined link timing plan. This spatiotemporal route can be obtained using contact graph routing, time-spread graph search, time-varying shortest path, or other algorithms that can utilize future link timing information. This invention does not require that the normal spatiotemporal route be generated by a specific algorithm.

[0057] To adapt to on-board execution, the pre-computed spatiotemporal route can be converted into a local routing execution table indexed by time slot and destination node. This local routing execution table consists of multiple routing entries, each containing information including: the index time slot, the destination node, the planned link-building object under the index time slot, and a forwarding flag indicating whether to forward to that planned link-building object. For navigation satellites... Time slot and destination node Local routing table entries Represented as: in, Indicates navigation satellite In the time slot According to the link timing plan, the planned link establishment objects are to be established for inter-satellite links; Indicates the destination node Data in time slots Whether to allow forwarding via the target chain of the aforementioned plan, This indicates that forwarding is allowed. This indicates that the data will not be forwarded and will be retained. Since the local node identifier is inherent information of the device, the local execution table does not need to repeatedly save the source node identifier in every routing table entry.

[0058] The forwarding flag using 0 / 1 is only one encoding method; other equivalent encodings that can distinguish between the allowed and prohibited forwarding states can also be used.

[0059] When there is no planned link building object in a certain time slot, it can be Record it as an empty value, and direct this time slot to each destination node. Set to 0. This definition allows for idle terminal time slots in the link timing plan and maintains consistency with the resource constraint of "at most one planned link construction object in the same time slot".

[0060] like Figure 2 As shown, when the same navigation satellite faces different destination nodes in the same time slot, its planned link establishment object field is the same, but the forwarding flag can be different. For example, satellite A in time slot If the planned link establishment target is B, then the data directed to the destination node D can correspond to flag=1, while the data directed to another destination node can correspond to flag=0. In the latter case, although the physical link between A and B can be established normally, B is not the appropriate forwarding direction that the current data should take according to the pre-calculated spatiotemporal route, so the data is still stored in the cache of A.

[0061] On-board forwarding employs a hop-by-hop execution method. The current satellite does not need to recalculate the complete end-to-end path from the current node to the final destination node in real time for each time slot; instead, it only needs to read the local routing table entry corresponding to the current time slot and the current destination node. When flag=1, data is sent to the planned link establishment object; when flag=0, data is stored in the local cache. After the data reaches the next navigation satellite, that next navigation satellite continues to query the local routing execution table and execute the same rules in its corresponding time slot. This mechanism can be implemented using the store-and-forward mechanism in a delay / interruption tolerant network. The field definitions of a single routing table entry in the local routing execution table are shown in Table 1.

[0062] Table 1. Field definitions of individual route entries in the local route execution table.

[0063] S1. Obtain fault status information and uniformly map the impact of faults.

[0064] This invention considers navigation satellite node failures and inter-satellite link failures that render a planned inter-satellite link unusable within a corresponding time slot. Fault detection itself can be implemented by the physical layer, link layer, payload health management unit, or other existing fault detection mechanisms, and is not a requirement of this invention.

[0065] Upon detection of a fault, fault status information is generated. To enable different navigation satellites to identify and handle the same fault event, a fault status message can at least include a fault event identifier, a fault object identifier, a fault category, a fault event reference time, and a fault status version number; for inter-satellite link faults, it can also include the nodes at both ends of the failed link and the time slot or time interval affected by the fault. Exemplary field definitions for fault status messages are shown in Table 2.

[0066] Table 2 Fault Status Message Field Definitions

[0067] Fault status can be propagated hop-by-hop using the original link timing schedule that was still available before the fault occurred, or it can be propagated through a dedicated control link or other available network mechanisms. This invention does not limit specific broadcast or routing protocols, only requiring that navigation satellites participating in the same emergency reconfiguration process ultimately obtain consistent fault status input.

[0068] Let the set of faulty nodes be... The set of failed inter-satellite links is For health navigation satellites and its time slot If the planned blockchain object belong Or plan inter-satellite links belong If so, the planned inter-satellite link for that navigation satellite in that time slot is defined as unavailable. This is represented as: ⇔ or in Indicates health navigation satellite exist Affected by faults in time slots 0 indicates a healthy navigation satellite exist The time slot was unaffected by the fault. Since each navigation satellite is equipped with only one fast-switching inter-satellite measurement terminal, and at most one planned link establishment target exists in the same time slot, for the affected healthy navigation satellites, although the physical causes of the peer node failure and the planned inter-satellite link failure are different, both manifest as the unavailability of the planned inter-satellite link in the relevant time slot in real-time network layer processing. If the faulty node was originally a destination node for a certain service, then that faulty node is removed from the set of valid destination nodes during the emergency phase and is no longer considered as a recovery target in this invention.

[0069] S2 performs parallel routing emergency handling and deterministic distributed topology incremental reconstruction.

[0070] (1) Routing emergency handling: fault forwarding suppression and cache waiting.

[0071] like Figure 3 As shown, once any navigation satellite obtains a fault status, it can immediately perform emergency routing measures without waiting for the entire network topology incremental reconstruction to be completed, or for the long-term global topology and routing recalculation to be completed.

[0072] For navigation satellites time slot First, check whether the planned link establishment objects and corresponding planned inter-satellite links in that time slot are affected by a fault. At that time, the time slot will be directed to all valid destination nodes. Forwarding flag Set all values ​​to 0: .

[0073] The above processing only changes the execution state of time slots affected by node failures and inter-satellite link failures; it does not delete the pre-calculated spatiotemporal routing results under normal conditions, nor does it run the complete routing algorithm that generated the original spatiotemporal routes. Data to be transmitted is stored in the local cache of the current navigation satellite according to the store-carry-forward mechanism. For time slots and original routing table entries not affected by the aforementioned node failures and inter-satellite link failures, their planned link establishment objects and forwarding flags remain unchanged.

[0074] Figure 4 A timing example is given. This applies to time slots unaffected by the fault. Satellite A forwards data normally according to its original local routing table. When a planned inter-satellite link in a subsequent relevant time slot becomes unavailable due to node failure or inter-satellite link failure, Satellite A forcibly updates the forwarding flag for the valid destination node to 0 and stores the data in Satellite A's cache. Time then continues to advance; even if the physical link establishment object in an intermediate time slot undergoes a normal switch, as long as the original pre-calculated spatiotemporal routing table still gives flag=0 for the current destination node, the data continues to be cached. Only in a subsequent time slot unaffected by the failure, when the planned link establishment object given by the original local routing table is normal and flag=1, does the cached data re-enter the forwarding process according to the original pre-calculated spatiotemporal route.

[0075] Therefore, the rapid emergency routing handling of this invention is not "immediately searching for another multi-hop path after a failure," but rather transforms the path re-search after a failure into local fault matching, Boolean forwarding flag updates, and cross-timeslot buffering. This is based on the fact that navigation satellite networks will continue to switch to other normal neighbors in subsequent time slots according to diverse measurement needs, potentially allowing executable forwarding opportunities that do not pass through the failed object to reappear in the original pre-calculated spatiotemporal routes.

[0076] (2) Deterministic distributed topology incremental reconstruction.

[0077] In parallel with the aforementioned emergency routing measures, each navigation satellite performs deterministic distributed incremental topology reconstruction. This reconstruction is based on the normal link timing plan prior to node and inter-satellite failures, without re-optimizing the overall network topology, but only compensating for planned link gaps caused by the failures. Figure 5 The process is illustrated.

[0078] (2.1) Non-faulty topology preservation and affected node identification.

[0079] Set normal time slot The planned inter-satellite link set is The set of planned inter-satellite links that failed due to faults is: The emergency addition of inter-satellite links is as follows The reconstructed emergency topology It can be represented as: in, All planned inter-satellite links unaffected by faults remain unchanged. For node faults, healthy endpoints with planned inter-satellite links to the faulty node in the corresponding time slot are identified as affected healthy nodes. For inter-satellite link faults, if both ends of the link are healthy, both ends can be identified as affected healthy nodes that lost their planned link-building targets in that time slot. When multiple faults exist simultaneously, the sets of affected healthy nodes generated by each fault can be combined and processed according to a unified deterministic order.

[0080] like Figure 5 As shown in the lower section, in an example normal topology, AB, CD, SE, and FG are the originally planned inter-satellite links. When node B fails, AB naturally fails, while the originally planned inter-satellite links CD, SE, and FG, which are unaffected by the failure, remain unchanged. A becomes an affected healthy node awaiting reconstruction due to losing its originally planned link-building target.

[0081] (2.2) Construction of alternative chain candidate sets.

[0082] For affected healthy nodes and time slot Construct a candidate set of alternative chain building methods One implementation method is as follows: in, Indicates in time slot With nodes A set of nodes that meet the conditions for establishing a chain; Represents the set of faulty nodes; This indicates the set of healthy nodes in which the fast-switching inter-satellite measurement terminal configured in that time slot is already occupied by the originally planned inter-satellite link that has not been affected by the fault or the confirmed emergency new inter-satellite link.

[0083] The feasibility of candidate links can be judged based on inter-satellite geometric visibility, antenna pointing range, link budget, terminal operating status, or information that can equivalently determine the above conditions. Candidate nodes must not be faulty nodes or nodes that are not active nodes. It itself, and its configured fast-switching inter-satellite measurement terminal is in an idle state in the current time slot that can be used to add an emergency inter-satellite link. If If the value is empty, an alternative link will not be forcibly established in that time slot to avoid violating the constraint that a single satellite can only establish a link in a single time slot; in this case, the routing side will continue to execute flag=0 and wait in the cache.

[0084] An empty candidate set does not affect the execution of the routing emergency response of this invention, nor does it constitute a failure of the emergency recovery method.

[0085] (2.3) Deterministic candidate selection.

[0086] To ensure that each navigation satellite receives the same emergency topology without requiring pairwise negotiation of new links, the candidate selection process must be deterministic. Each navigation satellite participating in emergency reconfiguration must have pre-stored or be able to deterministically obtain the same link timing plan, fault status, candidate feasibility information, time slot processing order, affected node processing order, resource update rules, and candidate ranking rules.

[0087] The consistent input does not require the use of exactly the same data storage format, as long as each health navigation satellite participating in emergency recovery can determine the same candidate set, processing order and candidate selection result based on the information obtained.

[0088] Candidate ranking can employ hierarchical rules, lexicographical comparison, weighted scoring, or other deterministic methods that guarantee a unique and identical output for the same input. Performance compensation factors considered may include, but are not limited to: prioritizing candidates that have not yet established measurement links with affected nodes during the current evaluation period to increase the diversity of ranging targets; prioritizing the formation of cross-class links when the system distinguishes between navigation satellites with direct Earth reachability and those without, to increase opportunities for Earth communication or reference transfer; assigning basic link recovery benefits to candidates capable of restoring a valid inter-satellite link; and determining the final candidate when key indicators are the same through fixed node numbers, preset node order, or other unique disambiguation rules.

[0089] For example, a deterministic score can be defined solely for implementation specification. : in, Indicates candidate Is it possible to provide a node Add new ranging neighbors, Indicates candidate With nodes Whether cross-class links have been formed between nodes with direct ground-to-ground capabilities and other nodes. Indicates candidate With nodes The fundamental benefits of establishing effective alternative links This represents the deterministic disambiguation term corresponding to a fixed node order. Weights , , , This can be adjusted according to specific system settings. This implementation does not require the use of the above-described linear scoring; it only requires that the candidate selection result has a unique and definite output under consistent input.

[0090] In one implementation of a navigation satellite network, the priorities of "fresh ranging, cross-ground capability category, effective non-idle link, and fixed node order" can be set sequentially with significant hierarchical differences, for example, using example score values ​​of 1500, 1000, 100, and 1 respectively. This set of values ​​is only used to illustrate a reproducible implementation and does not constitute a necessary limitation of the present invention.

[0091] (2.4) Time slot resource update and consistent output.

[0092] When affected healthy nodes Select candidate node And generate emergency new inter-satellite links After that, node and nodes All inter-satellite measurement terminals in this time slot are occupied during the rapid switching, therefore the node... and nodes Join the collection now When processing other affected nodes subsequently, the nodes... and nodes It will no longer be considered an idle candidate for the current time slot. If necessary, the ranging neighbor history, link recovery status, or other state variables that may affect the subsequent candidate ranking will be updated simultaneously.

[0093] Each navigation satellite performs the above operations sequentially according to the same time slot order and the same order of affected nodes, thus obtaining the same result under consistent input conditions. The process can be summarized as follows:

[0094] Consistent input + consistent processing order + deterministic rules ⇒ Consistent emergency topology

[0095] Therefore, the "distributed" nature of this invention refers to each navigation satellite independently executing the same deterministic reconstruction algorithm on-board, rather than each satellite making uncoordinated decisions for its own local links; its output is a consistent emergency topology across the entire network. This avoids the need for request-confirmation negotiation for each new link after a failure occurs.

[0096] This invention can handle single node failures, single inter-satellite link failures, or multiple simultaneous or sequential failures after obtaining a consistent set of failures. When multiple failures exist, all failed nodes are incorporated. Incorporate all failed inter-satellite links into Based on this, the set of affected healthy nodes is identified in each affected time slot. Each navigation satellite still performs deterministic incremental reconstruction according to a unified time slot order, fault object order, and affected node order.

[0097] For a given affected healthy node, if there are no candidate nodes in the current time slot that meet the conditions of link feasibility, non-faultiness, and fast handover of idle inter-satellite measurement terminals, then no alternative link will be established for that time slot. The node will not be considered an algorithm failure, nor will it be forcibly repaired by destroying other unaffected links. Instead, the node will maintain its forwarding flag at 0 and cache data on the routing side, awaiting the original routing opportunity in subsequent time slots or the results of a long-term global recovery.

[0098] In the event of an inter-satellite link failure, if both ends of the failed link are healthy nodes, then both ends lose their originally planned link establishment targets in that time slot. Both ends can be added to the set of affected nodes and processed according to a unified deterministic order. If the preceding endpoint selects a candidate and forms an emergency new inter-satellite link, then according to... The resource occupancy status is updated immediately, and the next endpoint will automatically exclude the occupied node during subsequent processing, thus continuing to satisfy the constraint that there can be at most one chain object per satellite and time slot.

[0099] S3. Incrementally update the local routing table based on the emergency topology.

[0100] After the deterministic distributed topology incremental reconstruction is completed, only the local routing tables related to the affected time slots and newly added emergency inter-satellite links are incrementally updated. Since the originally planned inter-satellite links that were not affected by the fault remain unchanged, their corresponding original routing table entries do not need to be modified.

[0101] by Figure 5 For example, if A originally planned to be in time slot If a planned inter-satellite link is established with B, and B fails, A deterministically selects H to form a new AH link. In this case, the planned link establishment object field in all routing table entries of A in that time slot will be updated from B to H. This is handled separately for different destination nodes.

[0102] (1) If the destination node If the peer H of the newly added inter-satellite link in the emergency is itself the destination node of the data, then the one-hop forwarding relationship from A to H can be directly determined by the physical link endpoint relationship without any new multi-hop path search. Therefore, the forward_flag of the corresponding routing table entry is set to 1.

[0103] (2) If the destination node Therefore, the new multi-hop path A→H→…→d cannot be derived solely based on the newly added AH link. During the emergency phase, the forward_flag corresponding to the destination node should be kept at 0, and the data should continue to be cached, waiting for subsequent time slots to have a normal forwarding opportunity according to the original spatiotemporal routing.

[0104] (3) For time slots and routing table entries that are not affected by the fault, the planned link establishment object and forwarding flag shall remain unchanged.

[0105] Therefore, the emergency routing table is not a completely new multi-hop routing table regenerated on the new topology, but rather consists of "fault-related routing table entries set to 0 + newly added direct-to-the-endpoint routing table entries set to 1 + unaffected routing table entries maintained". This method limits the impact of local topology changes on the routing layer to the minimum directly determinable range.

[0106] by Figure 5 For example, satellite A is in the time slot. The state changes of the routing table entries before, after and after the fault are shown in Table 3.

[0107] Table 3. Satellite A in time slots Example of routing table entry state change

[0108] S4. Activate the emergency topology at the predetermined synchronization switching boundary.

[0109] Although all navigation satellites can ultimately achieve the same emergency topology using deterministic rules, the actual time at which each satellite receives the fault status may differ. If a navigation satellite immediately activates a new link after completing its local calculations, but the other end of its proposed link has not yet received the fault status or entered emergency mode, inconsistencies may occur between the two ends of the link. Therefore, this invention employs a dual-time action mechanism of "real-time route protection + unified topology switching".

[0110] Fault-related forwarding suppression is implemented immediately after the navigation satellite obtains the corresponding fault status information, without waiting for the entire network to propagate completely. Emergency additions of inter-satellite links and one-hop direct forwarding relationships determined directly by their endpoints are implemented at the moment the emergency topology becomes unified. Enabled.

[0111] Let the reference time of the fault event be... The upper bound of the propagation required for the fault state to reach consistency among the nodes participating in emergency reconfiguration is: The upper bound for local topology incremental reconstruction calculation is The synchronous protection margin is Then one The method for determining it is as follows: , actual The nearest unified time slot boundary, frame boundary, or other predetermined synchronization handover boundary, no earlier than the aforementioned time, can be selected. Each navigation satellite uses the same fault event reference time or version information and the same... The calculation rules yield the same effective time, thus ensuring synchronous switching at both ends of the link without the need for additional link negotiation.

[0112] like Figure 6 As shown, the time from when a satellite receives a fault status report. arrive During this period, the satellite's fault-related routes are already under protection, and related data will no longer be sent to the faulty object; simultaneously, the topology incremental reconstruction results are awaiting unified activation. At that time, the entire network simultaneously activated emergency inter-satellite links and corresponding one-hop direct forwarding relationships.

[0113] S5. Continue forwarding cached data using the original routing table entries in subsequent time slots.

[0114] exist Subsequently, the navigation satellite network enters emergency topology operation mode. For one-hop direct relationships formed because the peer of the newly added link happens to be the destination node, the corresponding data can be sent directly when the new link becomes available. For other data that is still in a cached state, this invention does not re-search for new multi-hop spatiotemporal paths, but continues to utilize the spatiotemporal routes pre-calculated before the failure.

[0115] Time advances to subsequent time slots At that time, the navigation satellite reads the original routing table entry for the subsequent time slot in the local routing execution table. If the planned link establishment object indicated by the original routing table entry does not belong to a faulty node, the corresponding planned inter-satellite link is in a normal state, and the forwarding flag of the original routing table entry is in a forwarding-allowed state, then the cached data continues to be forwarded hop by hop according to the original routing table entry; if the above conditions are still not met, then it continues to be cached and waits for the subsequent time slot.

[0116] This process and Figure 1 and Figure 4 The spatiotemporal path changes described in the common description correspond to the following: data is not required to immediately form an alternative multi-hop path within the current time slot after a failure occurs; instead, transmission can be gradually restored using other normally established link objects that already exist in subsequent time slots. The cache duration can be managed by upper-layer business time limits, storage resources, data priorities, or other task strategies; this invention does not limit the fixed maximum cache duration.

[0117] Further implementation methods for long-term global recovery:

[0118] The rapid emergency recovery mechanism of this invention is used to fill the operational gap between the occurrence of a fault and the availability of long-term global recovery results, and does not replace long-term global optimization. Once a fault state is established, long-term global recovery can be initiated in parallel with the rapid emergency recovery process.

[0119] Long-term global recovery can be performed by ground-based computing equipment or by on-board computing nodes with sufficient computing power. Its processing may include re-collecting the overall network status after the failure, re-optimizing the timing plans for a larger area or all links, and recalculating the complete source-destination spatiotemporal routes on the new global topology. Any suitable topology optimization and routing algorithm can be used during the long-term global recovery phase.

[0120] Once the new long-term global link timing plan and complete spatiotemporal routing calculations are completed and distributed to the relevant navigation satellites, the unified effective time for long-term global recovery will be set. Each navigation satellite in Switch from emergency state to long-term global recovery state and begin executing the long-term global recovery topology and corresponding routes.

[0121] like Figure 6 As shown, the emergency plan is in It can start running immediately, while long-term global recovery can last for a longer period of time, and there is no waiting dependency between the two. Figure 6 The "approximately 1-2 minutes" and "approximately 20-30 minutes" marked in the text are merely examples of engineering time scales used to intuitively distinguish the time roles of fault propagation / emergency unified activation and long-term global topology and route recalculation, and do not constitute fixed time parameters that must be met by this invention.

[0122] Engineering application examples:

[0123] The invention is further illustrated below using a specific navigation satellite network as an application example. This application example provides feasible engineering parameters, fault scenarios, and handling actions, and does not constitute a limitation on the aforementioned general technical solutions.

[0124] In one implementation, navigation satellites may include medium Earth orbit satellites, inclined geosynchronous orbit satellites, and geostationary orbit satellites. Each navigation satellite participating in this method is equipped with a Ka-band phased array inter-satellite link terminal, employing a time-slotted polling link establishment method, with at most one inter-satellite link established with an object in the same time slot. The time slot length is typically 3 seconds; 20 time slots constitute a 60-second superframe; in a frozen geometry implementation, the visibility relationship is updated every 300 seconds, and the 60-second time slot topology can be reused or periodically executed within this 300-second state. The above 60-second and 300-second organization methods are only one engineering implementation.

[0125] During normal operation, a normal time-slot topology can be generated based on inter-satellite visibility, satellite positions, satellite-to-ground link status, and ranging and communication requirements at a 5-minute granularity or other granularities. Normal topology planning can employ fair contact planning, differentiated fair contact planning, integer planning, or other methods. Based on the obtained time-slot link timing plan, routes can be pre-calculated using contact graph routing or other spatiotemporal routing algorithms, and the results can be stored in a local execution table for each satellite.

[0126] S1: Generate fault status input and identify affected time slots.

[0127] The fault scenario can be set as follows: an inter-satellite link terminal of a certain satellite becomes continuously unavailable from a certain time slot or superframe; or a planned inter-satellite link becomes unavailable for a period of time. Upon detecting a fault, a fault status is generated, including the fault object identifier, the fault event reference time, and the status version information, and propagated hop-by-hop using the inter-satellite links that were still available before the fault occurred. Taking the affected healthy satellite A as an example, if it was originally scheduled to establish a link with the faulty node B in a certain time slot, the planned inter-satellite link between A and B in that relevant time slot is marked as unavailable, and B is simultaneously removed from the set of valid relays, candidate link establishment objects, and valid destination nodes during the emergency phase.

[0128] S2: Performs forwarding suppression and deterministic topology incremental reconstruction in parallel.

[0129] Upon obtaining a consistent fault status, A immediately sets the forwarding flag for the relevant affected time slots towards the valid destination node to 0 and buffers the data to be transmitted. Simultaneously, each healthy navigation satellite participating in emergency recovery maintains its original planned inter-satellite links unaffected by the fault and performs deterministic distributed topology incremental reconstruction on affected healthy nodes such as A, according to a unified fault set, a unified order of affected nodes, and unified candidate rules. In one candidate selection implementation, priority can be given to fresh ranging objects, followed by cross-ground capability category links between nodes with direct ground reachability and other nodes, then the basic recovery benefits of establishing any valid alternative inter-satellite link, and finally disambiguation using a fixed satellite number order. To enhance priority separation, scoring levels of 1500, 1000, 100, and 1 can be used as an example; after a candidate is selected, both ends of the newly added emergency inter-satellite link are immediately marked as occupied in that time slot.

[0130] S3: Update the local routing table for the affected time slots based on the emergency topology.

[0131] If the originally planned link B fails and a healthy candidate H is deterministically selected to form an emergency new AH link, then the planned link establishment object in the relevant local routing table entry of A in the affected time slot will be updated from B to H. For destination node H, the corresponding forwarding flag is set to 1, thus forming a one-hop direct relationship from A to H; for other destination nodes D, the multi-hop path after H is not derived based on the new AH link, and the corresponding data continues to be cached.

[0132] S4: Activate emergency inter-satellite links according to unified time boundaries.

[0133] In engineering practice, a conservative upper bound can be set for the propagation of fault conditions across the entire network. For example, 120 s; setting an upper bound for computation for lightweight topology refactoring. For example, 10 seconds; and on this basis, leave a synchronization protection margin, Set as follows: The unified time slot boundary occurs approximately 180 seconds later. The above values ​​are merely configurable parameters for a specific navigation satellite network embodiment. The core of this invention is that each satellite can immediately suppress routing forwarding upon receiving a fault status, while new topology additions await unification. Effective simultaneously.

[0134] S5: In Then, the cached data is forwarded using subsequent entries in the original routing table.

[0135] For data whose endpoint on the newly added link happens to be the destination node, it can be sent directly using the newly added emergency inter-satellite link; for other cached data, a new full multi-hop path search is not performed, but the original local routing table is read in subsequent time slots. When the corresponding planned link establishment object is in a normal state and the original forwarding flag allows forwarding, hop-by-hop transmission continues according to the original routing table entry; otherwise, it continues to be cached and waits for a later time slot.

[0136] Long-term overall recovery will be further implemented.

[0137] Long-term global recovery can be initiated simultaneously. For example, the fault status is transmitted back to the ground-based global computing node, the complete inter-satellite link timing plan is re-optimized under the new faulty network state, and the spatiotemporal routes of all sources and all destinations are recalculated. The results are then annotated and... A unified replacement of the emergency state is implemented. This long-term process can employ contact graph routing, time-spread graph search, or other routing algorithms; this invention does not limit the specific algorithm.

[0138] Technical boundary description of the application example:

[0139] In the above application examples, the 3s time slot, Ka band, 60s superframe, 300s state, 120s upper bound for fault propagation, 10s upper bound for topology calculation, approximately 180s unified effective time, specific weights of candidate scores, and contact graph routing are all used to illustrate an feasible navigation satellite network engineering configuration and should not be construed as necessary technical conditions of the present invention.

[0140] Each navigation satellite is equipped with a fast-switching inter-satellite measurement terminal participating in time-slotted link establishment, with at most one planned link establishment object in the same time slot; according to S1, consistent fault input is generated and node faults or inter-satellite link faults are mapped to the unavailability of planned inter-satellite links in the relevant time slots; according to S2, fault-related forwarding suppression and deterministic distributed topology incremental reconstruction are performed in parallel, where the original planned inter-satellite links unaffected by the fault are maintained, and incremental link supplementation is only performed on the fault-affected domain; according to S3, incremental updates are performed only on the planned link establishment objects and emergency new inter-satellite link endpoint one-hop forwarding relationships in the affected time slots; according to S4, emergency new inter-satellite links are uniformly activated at the predetermined synchronization handover boundary; according to S5, the original local routing table entries in subsequent time slots are used to continue hop-by-hop forwarding of cached data; long-term global recovery is performed independently and in parallel, and... Switch to long-term global recovery mode.

[0141] Those skilled in the art can make corresponding adjustments to the candidate feasibility determination, performance compensation indicators, caching strategies, and unified switching boundaries based on the orbital configuration, terminal system, link budget, service model, and on-board computing resources of a specific navigation satellite network. As long as the above-mentioned technical relationships are still used to achieve rapid emergency recovery after a failure, they are all equivalent implementations of the embodiments of the present invention.

Claims

1. A method for rapid emergency recovery of topology and spatiotemporal routing under navigation satellite network failure, characterized in that, The navigation satellites in the navigation satellite network are configured to quickly switch between inter-satellite measurement terminals for link establishment targets, using time slots as the unit for establishing inter-satellite links, ensuring that each navigation satellite has at most one planned link establishment target in the same time slot; the navigation satellite network has a pre-determined link timing plan, and based on the link timing plan, it pre-forms spatiotemporal routes utilizing link opportunities in different time slots, and each navigation satellite has a local routing execution table corresponding to the spatiotemporal routes; the method includes: S1. Obtain fault status information and map node faults and inter-satellite link faults to the unavailability of planned inter-satellite links in the relevant time slots; S2. After the health navigation satellite participating in emergency recovery obtains the fault status information, it performs route emergency handling and deterministic distributed topology incremental reconstruction in parallel. The routing emergency handling includes: without recalculating new multi-hop spatiotemporal routes, prohibiting the forwarding of corresponding data through the planned inter-satellite links affected by the fault and caching the data to be sent, while keeping the original routing table entries unaffected by the fault unchanged; The deterministic distributed topology incremental reconstruction includes: keeping the original planned inter-satellite links that are not affected by the fault unchanged, incrementally supplementing the links of healthy navigation satellites that have lost the original planned link establishment targets due to the fault, and forming emergency new inter-satellite links when there are available alternative link establishment targets, so as to obtain a consistent emergency topology. S3. Update the affected local routing execution table according to the emergency topology increment, wherein the emergency newly added inter-satellite link is only used for one-hop direct forwarding when its peer node is the destination node, and no new multi-hop spatiotemporal route is recalculated based on the emergency newly added inter-satellite link; S4. When the emergency topology unification takes effect, activate the newly added emergency inter-satellite link and the corresponding one-hop direct forwarding relationship; S5. For data that is still in the cache state, if there is an available forwarding opportunity for the original pre-calculated spatiotemporal route in a subsequent time slot, continue to forward it hop by hop according to the original pre-calculated spatiotemporal route.

2. The method according to claim 1, characterized in that, The local routing execution table consists of multiple routing entries. Each routing entry includes a timeslot identifier, a destination node identifier, a planned link establishment object identifier, and a forwarding flag. For the same timeslot of the same navigation satellite, when a planned link establishment object exists, different destination nodes correspond to the same planned link establishment object, and the forwarding flag indicates whether data destined for each destination node is sent via the planned link establishment object. When the forwarding flag uses binary encoding, a prohibited forwarding state is represented by 0, and a permitted forwarding state is represented by 1. For the affected timeslots that form emergency new inter-satellite links, the planned link establishment object is updated according to the emergency topology. When the peer node of the emergency new inter-satellite link is the destination node, the corresponding forwarding flag is set to the permitted forwarding state. When the peer node of the emergency new inter-satellite link is not the destination node, the corresponding forwarding flag is kept in the prohibited forwarding state. For the data cached in step S2, when the planned link establishment object indicated by the original entry in the subsequent timeslot is in a normal state and the corresponding forwarding flag is in the permitted forwarding state, forwarding continues hop-by-hop according to the original entry.

3. The method according to claim 1, characterized in that, In step S1, the fault status information is propagated hop-by-hop to the healthy navigation satellites participating in emergency recovery using the inter-satellite links in the link timing plan that was still available before the fault occurred. Each participating health navigation satellite determines a consistent fault input based on the fault object identifier carried in the fault status information, as well as at least one of the fault event reference time and fault status version number.

4. The method according to claim 1, characterized in that, Step S1 includes: for node failure, marking the planned inter-satellite link with the faulty node as the planned link establishment target as unavailable, and removing the faulty node from the set of effective relay nodes, the set of candidate link establishment targets, and the set of effective destination nodes in the emergency phase; for inter-satellite link failure, marking the faulty planned inter-satellite link as unavailable during the corresponding fault duration time slot; when the peer of the planned inter-satellite link in the relevant time slot is a faulty navigation satellite, or when the planned inter-satellite link itself fails, the same fault suppression processing is applied to the forwarding behavior of each effective destination node in that time slot.

5. The method according to claim 1, characterized in that, In step S2, an alternative link establishment candidate set is constructed for healthy navigation satellites that have lost their original planned link establishment targets due to malfunctions. The alternative link establishment candidate set consists of candidate navigation satellites that meet all of the following conditions: they have the conditions to establish an inter-satellite link with the healthy navigation satellite in the corresponding time slot; they are not faulty nodes; their configured fast-switching inter-satellite measurement terminals are not occupied in the corresponding time slot; and they are not the healthy navigation satellite itself.

6. The method according to claim 5, characterized in that, In step S2, each healthy navigation satellite participating in emergency recovery follows the same affected time slot processing order and affected navigation satellite processing order. When the alternative link establishment candidate set is not empty, a deterministic candidate selection rule is used to obtain a unique alternative link establishment object. The deterministic candidate selection rule is based on the following performance compensation factors: increasing ranging neighbor diversity, increasing inter-satellite links across ground capability categories, and restoring effective inter-satellite links in the original fault time slot. When the performance compensation factors are the same, deterministic disambiguation is performed using a pre-agreed node identifier order.

7. The method according to claim 5, characterized in that, When a candidate navigation satellite is selected from the alternative link establishment candidate set as the alternative link establishment object, and an emergency new inter-satellite link is formed with the healthy navigation satellite, both the healthy navigation satellite and the alternative link establishment object are marked as having their corresponding time slots occupied, so that subsequent deterministic distributed topology incremental reconstruction will no longer consider them as idle candidate link establishment objects for the corresponding time slots.

8. The method according to claim 1, characterized in that, In step S4, the emergency topology unification effective time is determined based on the fault event reference time and the preset upper bound of fault state propagation, upper bound of topology incremental reconstruction calculation, and synchronization protection margin; the emergency topology unification effective time is aligned to the unified time slot boundary, frame boundary, or other predetermined synchronization switching boundary; the operation of setting the forwarding flag to the prohibited forwarding state takes effect immediately after each participating healthy navigation satellite obtains the corresponding fault state information, and the emergency newly added inter-satellite link and the corresponding one-hop direct forwarding relationship take effect at the emergency topology unification effective time.

9. The method according to claim 1, characterized in that, While executing the rapid emergency recovery method, the fault status information is provided to at least one of the ground computing equipment and the on-board computing node with global computing capabilities. Long-term global link timing plan reconstruction and full spatiotemporal route recalculation based on the reconstructed link timing plan are performed. After the long-term global link timing plan and full spatiotemporal route calculation are completed and distributed, the navigation satellite network is switched from emergency state to long-term global recovery state at the unified effective time of long-term global recovery.

Images