An orbit prediction based on-board core network offline autonomous system and method

Abstract

The application provides a satellite-borne core network offline autonomous system and method based on orbit prediction, relates to the technical field of satellite communication, and predicts user access probability by combining satellite ephemeris and historical access logs, and differentiates pre-fetching of user states according to probability levels, so that the satellite can complete autonomous access of users of different levels based on local states when the satellite-ground link is interrupted. The application not only reduces the dependence on real-time queries of the ground core network, improves the service continuity in the disconnected scene, but also reduces the state synchronization bandwidth and the cost of on-board storage resources.

Description

Technical Field

[0001] This application relates to the field of satellite communication technology, and in particular to an offline autonomous system and method for a spaceborne core network based on orbit prediction. Background Technology

[0002] With the development of low-Earth orbit satellite constellations and 5G non-terrestrial network (5G-NTN) technologies, more and more core network functions are migrating to the satellite side to reduce service access latency, improve service continuity, and enhance network coverage. By deploying lightweight core network functions on satellites, services such as user access, session management, and data forwarding can be completed in orbit, thereby reducing dependence on the terrestrial core network and improving network autonomy.

[0003] However, the high-speed motion of low-Earth orbit satellites leads to frequent switching between satellite-to-ground and inter-satellite links. Furthermore, satellite-to-ground links are susceptible to intermittent interruptions due to factors such as ground station visibility, obstruction, and interference. Existing technologies still rely on the terrestrial core network for user authentication, session establishment, and state management. When the satellite-to-ground link is interrupted, users may be unable to access the service or services may be interrupted. On the other hand, although core network functions are deployed on the satellite side, real-time backhaul to the source is still required to obtain user state information, making it difficult to guarantee service continuity in unstable link scenarios. Another approach uses a full user state mirroring method to ensure offline service capabilities on the satellite side, but this suffers from high satellite-to-ground transmission costs, high onboard storage resource consumption, and untimely state updates.

[0004] Therefore, this application proposes a novel offline autonomous system for a spaceborne core network based on orbit prediction. Summary of the Invention

[0005] In view of the above problems, embodiments of this application provide an offline autonomous system and method for a spaceborne core network based on orbit prediction, so as to overcome the above problems or at least partially solve the above problems.

[0006] In a first aspect, this application provides an offline autonomous system for a spaceborne core network based on orbit prediction, comprising: a ground prediction subsystem, a ground pre-distribution subsystem, and a first satellite; The ground prediction subsystem determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; and predicts the access probability of each candidate access user in the ground cell based on the historical access log of the ground cell for the first satellite. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window. The ground pre-distribution subsystem performs hot prefetching on high-probability users based on their access probabilities within the ground cell, obtaining a complete session state snapshot of high-probability users; performs warm prefetching on medium-probability users, obtaining the session context skeleton and subscription digest of medium-probability users; and registers the user identifier and default policy pointer of low-probability users to obtain prefetch packets. The prefetch packets are then distributed to the first satellite via the satellite-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: Based on a snapshot of the complete session state of a high-probability user, establish a complete session for that user. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users; Based on the user identifier of low-probability users and the default policy pointer, establish restricted sessions for low-probability users.

[0007] Optionally, the first satellite includes at least: a first onboard autonomous strategy engine, a first hot buffer, a first warm buffer, and a first cold buffer; The ground pre-distribution subsystem distributes prefetch packets to the first onboard autonomous strategy engine via the satellite-to-ground link; The first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot state cache, writes the session context skeleton and subscription summary for medium-probability users in the prefetch packet into the first warm state cache, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold state cache. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

[0008] Optionally, the ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the access time window of each candidate access user in the ground cell. The first satellite also includes: an on-board cache manager, which binds the lifespan of each candidate access user in the ground cell to a time window aligned with the beam dwell time window; When the expected access time window of a high-probability user arrives and the high-probability user has not yet accessed the first satellite, the on-board cache manager will downgrade the high-probability user from hot state to warm state, release the complete session state snapshot of the high-probability user in the first hot state cache, and retain only the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache. When the end of the safe time window arrives and the high-probability user has not yet connected to the first satellite, the on-board cache manager will further downgrade the downgraded high-probability user from the warm state to the cold state, release the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache area, and retain the user identifier and default policy pointer of the downgraded high-probability user only in the first cold state cache area. When the lifespan expires and the beam dwell window ends, the on-board cache manager deletes the user identifiers and default policy pointers of each candidate access user in the first cold buffer.

[0009] Optionally, the ground pre-distribution subsystem sets high-probability and low-probability thresholds based on the currently available satellite-to-ground bandwidth budget; The ground prefetching subsystem lowers the high-probability and low-probability thresholds when the currently available satellite-to-ground bandwidth budget changes to a level higher than the satellite-to-ground bandwidth budget threshold, in order to expand the coverage of thermal prefetching and warm prefetching. The ground pre-distribution subsystem, when the currently available satellite-to-ground bandwidth budget changes to below the satellite-to-ground bandwidth budget threshold, increases the high-probability threshold and the low-probability threshold to ensure the distribution of complete session state snapshots for high-probability users, and to ensure the distribution of session context skeletons and subscription summaries for medium-probability users. The ground pre-distribution subsystem determines users with access probabilities higher than the high probability threshold as high-probability users, users with access probabilities between the high and low probability thresholds as medium-probability users, and users with access probabilities lower than the low probability thresholds as low-probability users, based on the access probabilities of each candidate access user in the ground cell.

[0010] Optionally, the ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the service type or service quality level of each candidate access user in the ground cell. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users, including: Select the corresponding session context skeleton based on the business type or service quality level predicted for users with a medium probability of usage. Based on the subscription summary of medium-probability users, the personalized information of medium-probability users is populated into the selected session context skeleton to enable the establishment of quick sessions for medium-probability users.

[0011] Optionally, during the offline autonomy of the onboard core network, the first satellite records the actual number of users accessing the ground cell, the actual access time window, the actual service type, and the actual service quality level. Based on the differences between the actual number of users accessing the network, the actual access time window, the actual service type, the actual service quality level, and the access probability, access time window, service type, and service quality level predicted by the prediction model of the ground prediction subsystem, the model parameters of the prediction model of the ground prediction subsystem are updated. The updated prediction model is used for the next execution of offline autonomy of the spaceborne core network.

[0012] Optionally, the system further includes: a second satellite; Based on the ephemeris of the first satellite and the second satellite, as well as the current location of the target user who has already connected to the first satellite, the first satellite predicts that the target user will switch from the first satellite to the second satellite at the expected switching time. Before the handover occurs, the first satellite will transfer the complete session state of the target user to the second satellite via an inter-satellite link. The second satellite writes the target user's complete session state into the second hot buffer. When a target user enters the beam coverage area of ​​the second satellite and initiates a handover, the second satellite hits the target user's complete session state in the second hot buffer and completes the session continuation locally.

[0013] A second aspect of this application provides an offline autonomous method for a spaceborne core network based on orbit prediction, the method comprising: The ground station determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; based on the historical access logs of the ground cell for the first satellite, it predicts the access probability of each candidate access user in the ground cell. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window. Based on the access probability of each candidate access user in the ground cell, the ground station performs hot prefetching on high-probability users to obtain a complete session state snapshot of high-probability users, performs warm prefetching on medium-probability users to obtain the session context skeleton and subscription digest of medium-probability users, and registers the user identifier and default policy pointer of low-probability users to obtain prefetch packets; the prefetch packets are then sent to the first satellite via the satellite-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: Based on a snapshot of the complete session state of a high-probability user, establish a complete session for that user. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users; Based on the user identifier of low-probability users and the default policy pointer, establish restricted sessions for low-probability users.

[0014] Optionally, the method further includes: The first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot state cache, writes the session context skeleton and subscription summary for medium-probability users in the prefetch packet into the first warm state cache, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold state cache. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

[0015] Optionally, based on the historical access logs of the ground cell, in addition to predicting the access probability of each candidate access user in the ground cell, the access time window of each candidate access user in the ground cell is also predicted. Bind a lifetime to each candidate access user in the ground cell that is aligned with the beam dwell time window; When the expected access time window of a high-probability user arrives and the high-probability user has not yet accessed the first satellite, the on-board cache manager will downgrade the high-probability user from hot state to warm state, release the complete session state snapshot of the high-probability user in the first hot state cache, and retain only the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache. When the end of the safe time window arrives and the high-probability user has not yet connected to the first satellite, the on-board cache manager will further downgrade the downgraded high-probability user from the warm state to the cold state, release the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache area, and retain the user identifier and default policy pointer of the downgraded high-probability user only in the first cold state cache area. When the lifespan expires and the beam dwell window ends, the on-board cache manager deletes the user identifiers and default policy pointers of each candidate access user in the first cold buffer.

[0016] The beneficial effects of this application are: This application provides an offline autonomous system for a spaceborne core network based on orbit prediction, comprising: a ground prediction subsystem, a ground pre-distribution subsystem, and a first satellite; the ground prediction subsystem determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; and predicts the access probability of each candidate access user in the ground cell based on the historical access logs of the ground cell for the first satellite, wherein the access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window; the ground pre-distribution subsystem performs hot prefetching on high-probability users based on the access probabilities of each candidate access user in the ground cell, obtaining a complete session state snapshot of the high-probability users. For medium-probability users, warm prefetching is performed to obtain their session context skeleton and subscription digest. The user identifier and default policy pointer of low-probability users are registered to obtain prefetch packets. These prefetch packets are then sent to the first satellite via the satellite-to-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to access requests initiated by each candidate access user within the beam dwell time window, implementing offline autonomy of the onboard core network according to the following steps: Based on the complete session state snapshot of high-probability users, a complete session is established for high-probability users; based on the session context skeleton and subscription digest of medium-probability users, a fast session is established for medium-probability users; and based on the user identifier and default policy pointer of low-probability users, a restricted session is established for low-probability users.

[0017] This application predicts user access probability by combining satellite ephemeris and historical access logs, and performs differentiated prefetching of user status based on probability levels. This enables autonomous access for different levels of users based on local status when the satellite-to-ground link is interrupted. This application not only reduces reliance on real-time queries from the ground core network and improves service continuity in disconnection scenarios, but also reduces state synchronization bandwidth and onboard storage resource costs. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a schematic diagram of an offline autonomous system architecture for a spaceborne core network based on orbit prediction, provided in an embodiment of this application. Figure 2 This is a schematic diagram of another offline autonomous system architecture for a spaceborne core network based on orbit prediction provided in the embodiments of this application; Figure 3 This is a timing diagram of a cross-satellite state relay provided in an embodiment of this application; Figure 4 This is a flowchart illustrating the steps of an offline autonomous method for a spaceborne core network based on orbit prediction, provided in an embodiment of this application. Figure 5 This is a flowchart of a state pre-distribution and offline autonomy process provided in an embodiment of this application. Detailed Implementation

[0020] Exemplary embodiments of this application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of this application are shown in the drawings, it should be understood that this application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art.

[0021] In a first aspect, this application provides an offline autonomous system for a spaceborne core network based on orbit prediction, such as... Figure 1 As shown, it includes: a ground prediction subsystem, a ground pre-distribution subsystem, and a first satellite; The ground prediction subsystem determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; and predicts the access probability of each candidate access user in the ground cell based on the historical access log of the ground cell for the first satellite. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window. The ground pre-distribution subsystem performs hot prefetching on high-probability users based on their access probabilities within the ground cell, obtaining a complete session state snapshot of high-probability users; performs warm prefetching on medium-probability users, obtaining the session context skeleton and subscription digest of medium-probability users; and registers the user identifier and default policy pointer of low-probability users to obtain prefetch packets. The prefetch packets are then distributed to the first satellite via the satellite-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: Based on a snapshot of the complete session state of a high-probability user, establish a complete session for that user. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users; Based on the user identifier of low-probability users and the default policy pointer, establish restricted sessions for low-probability users.

[0022] In this application, such as Figure 1 As shown, the spaceborne core network offline autonomous system based on orbit prediction includes a ground prediction subsystem, a ground pre-distribution subsystem, and the first satellite.

[0023] The ground prediction subsystem is used to determine the beam dwell time window of the first satellite for the target ground cell within its future orbital period, based on the ephemeris information of the first satellite. Since the orbital trajectory of low-Earth orbit satellites is deterministic, the coverage time range of the first satellite for the target ground cell can be obtained in advance through ephemeris information.

[0024] The ground prediction subsystem is also used to acquire historical access logs of the target ground cell for the first satellite. Based on information such as user access time and access frequency in the historical access logs, it predicts candidate access users who may initiate access requests to the first satellite within the beam dwell time window, obtaining the access probability corresponding to each candidate access user. The user's access probability represents the probability that the user will initiate an access request to the first satellite within the beam dwell time window of the target ground cell covered by the first satellite.

[0025] The ground pre-distribution subsystem performs differentiated state prefetching for users with different probability levels based on their access probabilities. Specifically, for users with high access probabilities, hot prefetching is performed to generate a complete session state snapshot containing the user's full session information; for users with access probabilities in the middle range, warm prefetching is performed to generate information including the session context skeleton and subscription summary; for users with low access probabilities, only the user identifier and default policy pointer are registered. The ground pre-distribution subsystem generates prefetch packets from the above different levels of prefetch information and sends them to the first satellite via the satellite-ground link when it is available, enabling the first satellite to obtain user state information that may be used later in advance.

[0026] In this application, when the satellite-to-ground link between the first satellite and the ground is interrupted, the first satellite does not need to rely on the ground core network to query the user status. Instead, it responds to user access requests based on locally stored prefetched packets. Specifically, when a high-probability user initiates an access request within the beam dwell time window, the first satellite establishes a complete session based on the corresponding complete session state snapshot; when a medium-probability user initiates an access request, the first satellite establishes a fast session based on the corresponding session context skeleton and subscription digest; when a low-probability user initiates an access request, the first satellite establishes a restricted session based on the user identifier and default policy pointer.

[0027] This application predicts user access probability by combining satellite ephemeris and historical access logs, and performs differentiated prefetching of user status based on probability levels. This enables autonomous access for different levels of users based on local status when the satellite-to-ground link is interrupted. This application not only reduces reliance on real-time queries from the ground core network and improves service continuity in disconnection scenarios, but also reduces state synchronization bandwidth and onboard storage resource costs.

[0028] In one embodiment, reference continues... Figure 1 As shown, the first satellite includes at least: a first onboard autonomous strategy engine, a first hot buffer, a first warm buffer, and a first cold buffer; The ground pre-distribution subsystem distributes prefetch packets to the first onboard autonomous strategy engine via the satellite-to-ground link; The first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot state cache, writes the session context skeleton and subscription summary for medium-probability users in the prefetch packet into the first warm state cache, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold state cache. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

[0029] In this embodiment, the first satellite is equipped with a first onboard autonomous strategy engine, a first hot-state cache, a first warm-state cache, and a first cold-state cache, which are used to classify and store user status information of different levels issued by the ground pre-distribution subsystem, and to realize onboard core network autonomous services based on local cache information during satellite-to-ground link interruption.

[0030] Specifically, the ground pre-distribution subsystem generates prefetch packets based on the access probability of candidate users and sends these packets to the first onboard autonomous policy engine when the satellite-ground link is available. Upon receiving the prefetch packets, the first onboard autonomous policy engine stores the user state information in a tiered manner according to different types of user status information within the packets. Specifically, a complete session state snapshot corresponding to high-probability users is written to the first hot-state buffer, enabling the first satellite to directly obtain complete state information upon user access; the session context skeleton and subscription summary corresponding to medium-probability users are written to the first warm-state buffer, allowing for rapid session construction based on existing information upon user access; and the user identifier and default policy pointer corresponding to low-probability users are written to the first cold-state buffer, providing a basic access basis when high-level state information is not hit.

[0031] When the satellite-to-ground link is interrupted, the first satellite cannot access user status information in the ground core network via the satellite-to-ground link. At this time, the first satellite-borne autonomous policy engine responds to the access request initiated by the candidate access user within the beam dwell time window by sequentially querying the first hot-state buffer, the first warm-state buffer, and the first cold-state buffer. When a complete session state snapshot of the corresponding user is found in the first hot-state buffer, user access and session establishment are completed based on the complete session state snapshot; when no complete session state snapshot is found but the session context skeleton and subscription digest of the corresponding user are found in the first warm-state buffer, the user session is constructed based on the above information; when no corresponding user status is found in any of the above buffers, restricted access is performed according to the user identifier and default policy pointer in the first cold-state buffer.

[0032] This embodiment utilizes a three-level caching mechanism (hot, warm, and cold) to manage the prefetch status hierarchically, enabling the first satellite to complete access processing for different levels of users based on the local cache status during satellite-to-ground link interruptions. This reduces the onboard storage resource consumption while improving the autonomous capability and service continuity of the onboard core network in disconnection scenarios.

[0033] For example, for high-probability users whose access probability exceeds a high threshold, hot prefetching is adopted, that is, prefetching a complete state snapshot, including subscription data, security context, session context, quality of service flow rules and billing initial context, so that the satellite can directly complete access and session establishment when disconnected.

[0034] For users with a medium probability of access between high and low thresholds, a warm prefetch is adopted, which means that only the policy template and the subscription summary are prefetched. When the connection is lost, a session can be quickly constructed based on the template, which can save bandwidth by exchanging a small amount of establishment latency. Specifically, the policy template is a parameterized session context skeleton that pre-defines the common and reusable parts for users under the same service type or quality of service level (such as default PDU session type, default network slice identifier, default quality of service flow rules and 5QI, default data network name, and default user plane forwarding template), leaving only user-specific fields (such as user identifier, security context, session address and tunnel identifier, and initial billing count) blank. When the connection is lost, the satellite selects the corresponding template based on the access user's service type and fills the blank fields with the personalized fields from the accompanying subscription digest. In addition to the subscription data and quality of service information, the subscription digest also includes a lightweight authentication vector or temporary security context pre-generated by the ground to support limited authentication of users during the offline period. The user's complete security context is completed from the source after the link is restored. Based on this, a complete session context can be instantiated locally, thereby completing authentication, session establishment, quality of service flow installation, and billing initiation.

[0035] For low-probability users whose access probability is below a low threshold, a cold registration approach is adopted. This means that only the user identifier and default policy pointer are registered, without issuing a specific status. In the event of a disconnection, a limited default policy is used as a backup for access. The limited default policy refers to a set of minimum and restricted service configurations pre-configured for users whose specific status has not been pre-fetched. For example: only providing the lowest priority, best-effort default bearer, and limiting the maximum rate and concurrent sessions; only allowing basic or emergency restricted services, and temporarily not opening high-value services; using pre-configured restricted credentials or temporary security contexts for limited authentication; and recording call detail records with a short temporary session validity period and limited billing tiers. Correspondingly, backup access means that when neither the hot nor warm status is met, the user is not directly rejected, but their access is temporarily allowed locally using the aforementioned limited default policy. The complete authentication and billing will be completed back to the origin server after the link is restored. It should be noted that the restricted service configuration can use the restricted default policy as the last resort in the offline autonomous hierarchical hit access decision, linking with the hit decision of hierarchical prefetching, and forming a closed loop with the local incremental logs during the disconnection period and the supplementation and merging after the link is restored.

[0036] In this application, the prefetch packet is encrypted and digitally signed by the ground, and then pushed to the target satellite via the satellite-to-ground link when the link is available. The satellite verifies the signature and writes it to the corresponding cache level. The digital signature is used to ensure the legitimacy of the source of the prefetch state and the integrity of its content, preventing it from being tampered with or forged on the satellite-to-ground link, thereby ensuring that the state on which local authentication and session establishment are based is trustworthy during offline periods.

[0037] In one embodiment, the ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the access time window of each candidate access user in the ground cell. The first satellite also includes: an on-board cache manager, which binds the lifespan of each candidate access user in the ground cell to a time window aligned with the beam dwell time window; When the expected access time window of a high-probability user arrives and the high-probability user has not yet accessed the first satellite, the on-board cache manager will downgrade the high-probability user from hot state to warm state, release the complete session state snapshot of the high-probability user in the first hot state cache, and retain only the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache. When the end of the safe time window arrives and the high-probability user has not yet connected to the first satellite, the on-board cache manager will further downgrade the downgraded high-probability user from the warm state to the cold state, release the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache area, and retain the user identifier and default policy pointer of the downgraded high-probability user only in the first cold state cache area. When the lifespan expires and the beam dwell window ends, the on-board cache manager deletes the user identifiers and default policy pointers of each candidate access user in the first cold buffer.

[0038] In this embodiment, based on the above embodiments, the ground prediction subsystem not only predicts the access probability of each candidate access user based on historical access logs, but also further predicts the access time window when each candidate access user may initiate an access request to the first satellite, so that the first satellite can dynamically manage the user status information in the cache by combining the user access probability and the expected access time.

[0039] Specifically, the first satellite includes an on-board cache manager, which sets a lifespan for each candidate access user that matches the beam dwell time window. The lifespan characterizes the effective time range for the corresponding user status information to be cached on the first satellite, enabling the first satellite to dynamically adjust the caching method for different levels of status information based on the actual access situation of users.

[0040] For high-probability users, since these users are more likely to initiate an access request to the first satellite within the expected access time window, the first satellite pre-stores a complete session state snapshot of these users in the first hot-state buffer. It should be noted that the complete session state snapshot includes state information used to establish a complete session, as well as the session context skeleton and subscription digest used to construct the basic session. When the expected access time window for a high-probability user ends, but the user has still not initiated an access request to the first satellite, it indicates that the probability of the user actually accessing the satellite within the current beam dwell time window has decreased. At this time, the onboard buffer manager downgrades the user from hot to warm state, releases the complete session state snapshot stored in the first hot-state buffer, and retains only the session context skeleton and subscription digest corresponding to the user, in order to reduce the hot-state buffer resource consumption.

[0041] When the security window ends and the user is still not connected to the first satellite, the on-board cache manager further reduces the user's cache level, downgrading the user from warm state to cold state, releasing the session context skeleton and subscription digest in the first warm state cache area, and retaining only the user's user identifier and default policy pointer in the first cold state cache area to maintain a minimum amount of user identification and access policy information.

[0042] When the lifespan of a candidate access user and the beam dwell window of a satellite for the target ground cell expire, it indicates that the first satellite no longer covers the corresponding ground area, and the current prefetch state is no longer useful. At this time, the onboard cache manager deletes the user identifiers and default policy pointers of each candidate access user stored in the first cold state cache area, releasing the corresponding cache resources. It should be noted that in this application, the beam dwell time window is used to represent the time range during which the first satellite provides services to the target ground cell, the expected access time window is used to represent the time range during which a candidate access user is expected to initiate an access request to the first satellite, and the safety time window is used to represent the state retention buffer time range set after the expected access time window ends to avoid premature state release due to access prediction errors.

[0043] This embodiment dynamically adjusts the user state cache level based on the user's expected access time, ensuring that high-value state information maintains a high cache level only during the time the user is likely to access the network. Cache resources are released gradually when the user is not accessing the network. This improves the utilization efficiency of onboard storage resources while maintaining offline autonomous service capabilities, preventing expired states from occupying onboard storage for extended periods, thus balancing timeliness and resource constraints. Furthermore, during offline periods, user access is achieved locally, avoiding cross-satellite-to-ground source lookups. Access and session establishment are handled locally without relying on satellite-to-ground links, which helps meet the latency requirements of a single core network service.

[0044] For example, taking a user's expected access time window as T0+2min to T0+3min as an example: their status is issued at T0 and written to the hot state area, with a bound lifespan until the end of the beam dwell window (T0+6min); when T0+3min arrives, which is close to or past the end of the access time window and the user still has not accessed, their hot state is downgraded to a warm state, releasing the complete session context in high-speed storage while retaining only the policy template and summary; if they still have not accessed after a safe time window (e.g., until T0+4.5min), the warm state is further downgraded to a cold state, retaining only the identifier and default policy pointer; when the lifespan expires at T0+6min and the beam dwell window ends, the cold state is deleted. If the user has already accessed before the downgrade, their status is directly consumed and converted to an active session, no longer participating in the above downgrade process.

[0045] In one embodiment, the ground pre-distribution subsystem sets a high-probability threshold and a low-probability threshold based on the currently available satellite-to-ground bandwidth budget. The ground prefetching subsystem lowers the high-probability and low-probability thresholds when the currently available satellite-to-ground bandwidth budget changes to a level higher than the satellite-to-ground bandwidth budget threshold, in order to expand the coverage of thermal prefetching and warm prefetching. The ground pre-distribution subsystem, when the currently available satellite-to-ground bandwidth budget changes to below the satellite-to-ground bandwidth budget threshold, increases the high-probability threshold and the low-probability threshold to ensure the distribution of complete session state snapshots for high-probability users, and to ensure the distribution of session context skeletons and subscription summaries for medium-probability users. The ground pre-distribution subsystem determines users with access probabilities higher than the high probability threshold as high-probability users, users with access probabilities between the high and low probability thresholds as medium-probability users, and users with access probabilities lower than the low probability thresholds as low-probability users, based on the access probabilities of each candidate access user in the ground cell.

[0046] In this embodiment, the ground pre-distribution subsystem dynamically adjusts the user classification threshold based on the currently available satellite-to-ground bandwidth resources, so that the prefetching strategy can adapt to different satellite-to-ground link resource states.

[0047] Specifically, the ground pre-distribution subsystem first sets high-probability and low-probability thresholds based on the currently available satellite-to-ground bandwidth budget. The high-probability threshold is used to determine whether to classify candidate access users as high-probability users, while the low-probability threshold is used to determine whether to classify candidate access users as medium-probability or low-probability users.

[0048] When the available satellite-to-ground bandwidth budget exceeds the preset satellite-to-ground bandwidth budget threshold, it indicates that the satellite-to-ground link has sufficient data transmission capacity. At this time, the ground pre-distribution subsystem lowers the high-probability threshold and the low-probability threshold, allowing more candidate access users to enter the high-probability or medium-probability user range, thereby expanding the coverage of hot prefetching and warm prefetching, enabling more users to obtain more complete status information in advance, and improving service assurance capabilities in the event of satellite-to-ground link interruption.

[0049] When the currently available satellite-to-ground bandwidth budget is lower than the preset satellite-to-ground bandwidth budget threshold, it indicates that satellite-to-ground link transmission resources are limited. At this time, the ground pre-distribution subsystem raises the high-probability threshold and the low-probability threshold, narrows the range of high-probability and medium-probability users, prioritizes sending complete session state snapshots only to users with a higher probability of access, and ensures that the session context skeleton and subscription digest of a certain number of medium-probability users can be sent, thereby reducing the amount of prefetched data and reducing the load on the satellite-to-ground link.

[0050] Furthermore, the ground pre-distribution subsystem classifies each candidate access user within the ground cell based on the adjusted high-probability threshold and low-probability threshold. When the access probability of a candidate access user is higher than the high-probability threshold, the user is identified as a high-probability user; when the access probability of a candidate access user is between the high-probability threshold and the low-probability threshold, the user is identified as a medium-probability user; and when the access probability of a candidate access user is lower than the low-probability threshold, the user is identified as a low-probability user.

[0051] In this embodiment, by using access probability prediction and confidence-level prefetching, the system only issues necessary states that are predicted to be hit and are classified according to probability. The satellite-to-ground prefetching bandwidth is significantly reduced, effectively resolving the contradiction between state availability and satellite-to-ground bandwidth.

[0052] In one embodiment, the ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the service type or service quality level of each candidate access user in the ground cell. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users, including: Select the corresponding session context skeleton based on the business type or service quality level predicted for users with a medium probability of usage. Based on the subscription summary of medium-probability users, the personalized information of medium-probability users is populated into the selected session context skeleton to enable the establishment of quick sessions for medium-probability users.

[0053] In this embodiment, the ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user initiating an access request to the first satellite within the beam dwell time window, but also further predicts the service type or corresponding quality of service level that each candidate access user may use.

[0054] Specifically, for medium-probability users, since their access probability is in the middle range, sending a complete session state snapshot to them in advance would increase the transmission pressure on the satellite-to-ground link and the onboard storage usage. Therefore, the ground pre-distribution subsystem only pre-fetches the session context skeleton and subscription digest for medium-probability users. When the satellite-to-ground link is interrupted and a medium-probability user initiates an access request, the first satellite selects a session context skeleton that matches the service type or service quality level from multiple pre-stored session context skeletons, based on the pre-predicted service type or service quality level corresponding to the medium-probability user.

[0055] Furthermore, the first satellite, based on the subscription summary corresponding to the medium-probability user, populates the user's personalized information into the selected session context skeleton to generate a quick session for that user. In this way, it is not necessary to pre-store the complete session state of each medium-probability user; a suitable session configuration can be quickly generated based on business needs when the user actually accesses the system.

[0056] This embodiment achieves rapid session construction by predicting the user's service type or service quality level and combining the session context skeleton and subscription summary. While reducing the occupation of onboard cache resources, it improves the access efficiency and service adaptation capability of medium-probability users in the event of satellite-to-ground link interruption.

[0057] In one embodiment, during the offline autonomous operation of the onboard core network, the first satellite records the actual number of users accessing the ground cell, the actual access time window, the actual service type, and the actual service quality level. Based on the differences between the actual number of users accessing the network, the actual access time window, the actual service type, the actual service quality level, and the access probability, access time window, service type, and service quality level predicted by the prediction model of the ground prediction subsystem, the model parameters of the prediction model of the ground prediction subsystem are updated. The updated prediction model is used for the next execution of offline autonomy of the spaceborne core network.

[0058] In this embodiment, the first satellite records the actual user access situations during the offline autonomous operation of the onboard core network in order to optimize the prediction accuracy of the ground prediction subsystem.

[0059] Specifically, the first satellite records information about actual users accessing the target ground cell, including the corresponding actual access time window, actual service type, and actual quality of service (QoS) level. Once the satellite-to-ground link is restored, the first satellite compares this actual access information with the access probability, access time window, service type, and QoS level predicted by the ground prediction subsystem based on historical access logs to determine the difference between the prediction results and the actual access situation.

[0060] Furthermore, the ground-based prediction subsystem adjusts the model parameters of the prediction model based on the aforementioned differences, enabling the prediction model to continuously optimize prediction results according to actual operational conditions. The updated prediction model is used for user access prediction, service type prediction, and service quality level prediction in subsequent orbital cycles, thereby improving the accuracy of user state pre-fetching.

[0061] In this embodiment, the access data generated during actual satellite operation is used to perform closed-loop optimization of the prediction model, enabling the user state prefetching strategy to adapt to actual business changes. By correcting the prediction model and grading thresholds through online feedback of the deviation between prediction and reality, the prefetching hit rate continues to improve with operation, giving the system a self-learning ability that becomes more accurate with use.

[0062] In one embodiment, such as Figure 2 As shown, the system also includes: a second satellite; Based on the ephemeris of the first satellite and the second satellite, as well as the current location of the target user who has already connected to the first satellite, the first satellite predicts that the target user will switch from the first satellite to the second satellite at the expected switching time. Before the handover occurs, the first satellite will transfer the complete session state of the target user to the second satellite via an inter-satellite link. The second satellite writes the target user's complete session state into the second hot buffer. When a target user enters the beam coverage area of ​​the second satellite and initiates a handover, the second satellite hits the target user's complete session state in the second hot buffer and completes the session continuation locally.

[0063] In addition to orbit prediction-based on-board core network offline autonomous system Figure 1 The aforementioned includes: the ground prediction subsystem, the ground pre-distribution subsystem, and other components besides the first satellite, such as... Figure 2 As shown, it also includes a second satellite.

[0064] In this embodiment, the first satellite determines the coverage relationship and expected handover time between the first satellite and the second satellite based on the ephemeris information of the first satellite and the second satellite, and predicts the expected handover time when the target user will switch from the first satellite to the second satellite by combining the current location of the target user who has been connected to the first satellite.

[0065] Before the expected handover time arrives, the first satellite transmits the target user's complete session state to the second satellite via an inter-satellite link, enabling the second satellite to obtain the target user's session information in advance. After receiving the target user's complete session state, the second satellite writes the complete session state into a second hot buffer to complete the preloading of the target user's session state on the next serving satellite.

[0066] When a target user enters the beam coverage area of ​​the second satellite and initiates a satellite handover request, the second satellite does not need to query the user status through the ground core network again. Instead, it directly obtains the complete session status corresponding to the target user from the second hot buffer and completes the session continuation locally based on the complete session status.

[0067] This embodiment enables users to switch between satellites without service interruption and with zero backhaul by relaying inter-satellite status based on orbit prediction, significantly reducing handover latency and failure risk.

[0068] For example, such as Figure 3 The timing diagram for the cross-satellite relay shown below illustrates the specific steps: The first step is switchover prediction. Based on the orbital predictions of the ephemeris data of satellites S1 and S2, the ground-side system or satellite S1 determines in advance that user U will soon switch from satellite S1 to satellite S2, and identifies the approximate switchover time (approximately 2 minutes after user U connects). The inputs to this prediction are the ephemeris data of the two satellites and user U's current position; the outputs are satellite S2 and the predicted switchover time, which can be compared and verified with the actual time user U enters the satellite S2 beam.

[0069] The second step is session state relay and warm-up. Before the actual handover occurs, satellite S1 relays user U's session state (including session context, quality of service flow rules, and security context) to satellite S2 via an inter-satellite link. Upon receiving the data, satellite S2 warms it up to a hot state and binds the corresponding lifetime. The relayed content and the warm-up status on the satellite S2 side can be verified in the logs of satellite S1 and satellite S2, respectively.

[0070] The third step is hot-state hit detection and session continuation. When user U actually enters the S2 satellite beam and initiates a handover, the S2 satellite directly detects user U's session state in the hot-state zone, completing session continuation and service quality flow reconstruction locally without needing to query the origin ground. The hit level, continuation result, and handover processing latency can be observed through the S2 satellite's local logs to verify zero origin backhaul and uninterrupted service.

[0071] The fourth step is source-side elimination and differential recovery. After the handover is completed, the session status corresponding to satellite S1 is downgraded and eliminated according to its lifespan, while satellite S2 continues to provide services to user U. When satellite S2 subsequently passes over ground stations and the link is restored, it merges user U's session and billing change incremental differences back into the ground authoritative core network. It should be noted that the session status that is relayed to satellite S2 is also included in the hierarchical caching and elimination management of satellite S2: it is bound to satellite S2 with a lifespan aligned with the beam dwell window of satellite S2 for that grid and user. When user U's session ends, user U leaves the coverage of satellite S2, or the lifespan expires, satellite S2 also downgrades and ultimately eliminates the session according to the rules of hot state to cool state, warm state to cold state, and cold state deletion, consistent with the mechanism on satellite S1. It should be noted that the above-mentioned lifespan-driven elimination only applies to the relay state that has been pre-warmed but not yet consumed; once user U actually accesses satellite S2, its state is consumed and converted to an in-use session, and is now managed by the session lifecycle (normal session termination or user U leaving the coverage area of ​​satellite S2), no longer subject to the pre-fetch lifespan-driven degradation elimination process. The consistency between the elimination of satellite S1 and satellite S2 states and the merged state on the ground side can be verified.

[0072] In one embodiment, when the satellite-to-ground link is restored, the system described in this application performs synchronous processing on the data generated during the offline autonomy process to achieve state consistency maintenance and prediction model optimization.

[0073] Specifically, the first satellite transmits the local state increments generated during offline periods to the ground-based authoritative core network, and compares and merges them with the user state data in the ground-based authoritative core network. During the data merging process, conflicting data is handled according to preset data adjudication rules. For example, billing data is accumulated and merged, session data retains the latest valid state, and subscription data is based on the data in the ground-based authoritative core network, thereby ensuring the consistency of user session, billing, and subscription data.

[0074] Based on the same inventive concept, the second aspect of this application provides an offline autonomous method for a spaceborne core network based on orbit prediction, such as... Figure 4 As shown, the method includes: S101, the ground station determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; based on the historical access logs of the ground cell for the first satellite, it predicts the access probability of each candidate access user in the ground cell. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window.

[0075] In this step, the ground station determines the beam dwell time window for the target ground cell within the future orbital period of the first satellite based on the ephemeris information of the first satellite. Since low-Earth orbit satellites operate according to predetermined orbits, their coverage area and coverage time follow certain patterns. The satellite ephemeris allows for the prediction of the time range in which the first satellite will pass over the target ground cell.

[0076] Furthermore, the ground station acquires the historical access logs of the target ground cell for the first satellite, and based on information such as user access time and access frequency in the historical access logs, predicts the probability that each candidate access user in the target ground cell will initiate an access request to the first satellite within the beam dwell time window, thus obtaining the access probability corresponding to each candidate access user. The user's access probability represents the probability that the user will initiate an access request to the first satellite during the period when the first satellite covers the target ground cell.

[0077] S102, the ground station performs hot prefetching on high-probability users based on the access probability of each candidate access user in the ground cell, to obtain a complete session state snapshot of the high-probability users, performs warm prefetching on medium-probability users, to obtain the session context skeleton and subscription digest of the medium-probability users, and registers the user identifier and default policy pointer of low-probability users in order to obtain the prefetch packet; and sends the prefetch packet to the first satellite through the satellite-ground link.

[0078] In this step, the ground station performs differentiated state prefetching for users with different access probability levels based on the access probability corresponding to each candidate access user.

[0079] Specifically, for users with a high access probability, hot prefetching is performed to obtain a complete session state snapshot corresponding to the high-probability user; for users with an access probability in the middle range, warm prefetching is performed to obtain the session context skeleton and subscription summary corresponding to the medium-probability user; for users with a low access probability, the user identifier and default policy pointer corresponding to the low-probability user are registered.

[0080] The ground station encapsulates the status information corresponding to the different levels of users, generates a prefetch packet, and sends the prefetch packet to the first satellite through the satellite-ground link when the satellite-ground link is available, so that the first satellite can obtain the status information required by users who may access the satellite in the future beam dwell time window in advance.

[0081] S103, If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: S1031, Based on the complete session state snapshot of the high-probability user, establish a complete session for the high-probability user; S1032, establish a fast session for the medium probability user based on the session context skeleton and subscription summary of the medium probability user; S1033, based on the user identifier of the low-probability user and the default policy pointer, establish a restricted session for the low-probability user.

[0082] In this step, when the satellite-to-ground link between the first satellite and the ground station is interrupted, the first satellite cannot access the ground core network in real time to obtain user status information. At this time, in response to the access requests initiated by each candidate access user within the beam dwell time window, the first satellite implements offline autonomy of the onboard core network according to the following steps: S1031, establish a complete session for a high-probability user based on a complete session state snapshot.

[0083] When a high-probability user initiates an access request to the first satellite, the first satellite obtains the session state information corresponding to the user based on the complete session state snapshot in the prefetch packet, and completes the user session establishment based on the complete session state snapshot, so that the high-probability user can continue to obtain the full service without accessing the ground core network.

[0084] S1032, establishes a fast session for medium-probability users based on the session context skeleton and the contract summary.

[0085] When a medium-probability user initiates an access request to the first satellite, the first satellite obtains the basic session information corresponding to the user based on the session context skeleton and subscription digest in the prefetch packet, and quickly constructs the user session based on the above information, thereby enabling service access for medium-probability users while reducing the amount of state prefetching.

[0086] S1033, establish a restricted session for a low-probability user based on the user identifier and the default policy pointer.

[0087] When a low-probability user initiates an access request to the first satellite, the first satellite determines the user's identity and the corresponding default service policy based on the user identifier and default policy pointer in the prefetch packet, and establishes a restricted session based on the above information to provide basic access capabilities.

[0088] In one embodiment, the first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot buffer, writes the session context skeleton and subscription digest for medium-probability users in the prefetch packet into the first warm buffer, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold buffer. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

[0089] In this embodiment, after receiving the prefetch packet, the first onboard autonomous policy engine classifies and stores the state information corresponding to different user levels in the prefetch packet. Specifically, the complete session state snapshot corresponding to high-probability users is written to the first hot-state cache; the session context skeleton and subscription digest corresponding to medium-probability users are written to the first warm-state cache; and the user identifier and default policy pointer corresponding to low-probability users are written to the first cold-state cache.

[0090] When the satellite-to-ground link is interrupted, the first satellite cannot obtain user status information from the ground core network in real time. At this time, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window, and sequentially queries the first hot buffer, the first warm buffer, and the first cold buffer through the first onboard autonomous strategy engine to perform offline autonomy of the onboard core network.

[0091] In one embodiment, this application also provides a flowchart of the overall process for state pre-distribution and offline autonomy, such as... Figure 5 As shown: Obtain satellite ephemeris and historical access logs, and determine the beam dwell time window for the target ground cell within the future orbital period of the satellite based on the satellite ephemeris; predict the access probability of users in the target ground cell who may access the first satellite based on the historical access logs, and obtain a list of candidate access users with time windows.

[0092] Furthermore, based on the access probability of each candidate user, users are classified into confidence levels, namely high-probability users, medium-probability users, and low-probability users. Specifically, hot prefetching is performed for high-probability users; warm prefetching is performed for medium-probability users; and cold registration is performed for low-probability users.

[0093] Furthermore, the prefetch information corresponding to the different user levels is encapsulated into prefetch packets and sent to the first satellite via the satellite-to-ground link. After receiving the prefetch packets, the first satellite performs hierarchical caching according to the user level.

[0094] When a satellite-to-ground link interruption is detected, the first satellite performs offline autonomy of the onboard core network based on local cached information.

[0095] During the offline autonomous operation of the first satellite, operational data such as actual access users, actual access time windows, actual service types, and actual service quality levels are recorded. The actual operational data is then compared with the access probability, access time window, service type, and service quality level output by the prediction model. The prediction model parameters are then updated based on the results of the comparison analysis.

[0096] Each embodiment in this specification focuses on the differences from other embodiments. For the same or similar parts between the embodiments, please refer to each other.

[0097] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, embodiments of this application can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of this application can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0099] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0101] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0102] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0103] The above provides a detailed description of the offline autonomous system and method for a spaceborne core network based on orbit prediction. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An offline autonomous system for a spaceborne core network based on orbit prediction, characterized in that, include: Ground prediction subsystem, ground pre-distribution subsystem, and the first satellite; The ground prediction subsystem determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; and predicts the access probability of each candidate access user in the ground cell based on the historical access log of the ground cell for the first satellite. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window. The ground pre-distribution subsystem performs hot prefetching on high-probability users based on their access probabilities within the ground cell, obtaining a complete session state snapshot of high-probability users; performs warm prefetching on medium-probability users, obtaining the session context skeleton and subscription digest of medium-probability users; and registers the user identifier and default policy pointer of low-probability users to obtain prefetch packets. The prefetch packets are then distributed to the first satellite via the satellite-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: Based on a snapshot of the complete session state of a high-probability user, establish a complete session for that user. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users; Based on the user identifier of low-probability users and the default policy pointer, establish restricted sessions for low-probability users.

2. The offline autonomous system for spaceborne core network based on orbit prediction according to claim 1, characterized in that, The first satellite includes at least: a first onboard autonomous strategy engine, a first hot buffer, a first warm buffer, and a first cold buffer; The ground pre-distribution subsystem distributes prefetch packets to the first onboard autonomous strategy engine via the satellite-to-ground link; The first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot state cache, writes the session context skeleton and subscription summary for medium-probability users in the prefetch packet into the first warm state cache, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold state cache. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

3. The offline autonomous system for spaceborne core network based on orbit prediction according to claim 2, characterized in that, The ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the access time window of each candidate access user in the ground cell. The first satellite also includes: an on-board cache manager, which binds the lifespan of each candidate access user in the ground cell to a time window aligned with the beam dwell time window; When the expected access time window of a high-probability user arrives and the high-probability user has not yet accessed the first satellite, the on-board cache manager will downgrade the high-probability user from hot state to warm state, release the complete session state snapshot of the high-probability user in the first hot state cache, and retain only the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache. When the end of the safe time window arrives and the high-probability user has not yet connected to the first satellite, the on-board cache manager will further downgrade the downgraded high-probability user from the warm state to the cold state, release the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache area, and retain the user identifier and default policy pointer of the downgraded high-probability user only in the first cold state cache area. When the lifespan expires and the beam dwell window ends, the on-board cache manager deletes the user identifiers and default policy pointers of each candidate access user in the first cold buffer.

4. The offline autonomous system for spaceborne core network based on orbit prediction according to claim 1, characterized in that, The ground pre-distribution subsystem sets high-probability and low-probability thresholds based on the currently available satellite-to-ground bandwidth budget. The ground prefetching subsystem lowers the high-probability and low-probability thresholds when the currently available satellite-to-ground bandwidth budget changes to a level higher than the satellite-to-ground bandwidth budget threshold, in order to expand the coverage of thermal prefetching and warm prefetching. The ground pre-distribution subsystem, when the currently available satellite-to-ground bandwidth budget changes to below the satellite-to-ground bandwidth budget threshold, increases the high-probability threshold and the low-probability threshold to ensure the distribution of complete session state snapshots for high-probability users, and to ensure the distribution of session context skeletons and subscription summaries for medium-probability users. The ground pre-distribution subsystem determines users with access probabilities higher than the high probability threshold as high-probability users, users with access probabilities between the high and low probability thresholds as medium-probability users, and users with access probabilities lower than the low probability threshold as low-probability users, based on the access probabilities of each candidate access user in the ground cell.

5. The offline autonomous system for spaceborne core network based on orbit prediction according to claim 2, characterized in that, The ground prediction subsystem, based on the historical access logs of the ground cell, not only predicts the access probability of each candidate access user in the ground cell, but also predicts the service type or service quality level of each candidate access user in the ground cell. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users, including: Select the corresponding session context skeleton based on the business type or service quality level predicted for users with a medium probability of usage. Based on the subscription summary of medium-probability users, the personalized information of medium-probability users is populated into the selected session context skeleton to enable the establishment of quick sessions for medium-probability users.

6. The offline autonomous system for spaceborne core network based on orbit prediction according to claim 5, characterized in that, During the offline autonomous operation of the onboard core network, the first satellite records the actual number of users accessing the ground cell, the actual access time window, the actual service type, and the actual service quality level. Based on the differences between the actual number of users accessing the network, the actual access time window, the actual service type, the actual service quality level, and the access probability, access time window, service type, and service quality level predicted by the prediction model of the ground prediction subsystem, the model parameters of the prediction model of the ground prediction subsystem are updated. The updated prediction model is used for the next execution of offline autonomy of the spaceborne core network.

7. The offline autonomous system for spaceborne core network based on orbit prediction according to any one of claims 1-6, characterized in that, The system also includes: a second satellite; Based on the ephemeris of the first satellite and the second satellite, as well as the current location of the target user who has already connected to the first satellite, the first satellite predicts that the target user will switch from the first satellite to the second satellite at the expected switching time. Before the handover occurs, the first satellite will transfer the complete session state of the target user to the second satellite via an inter-satellite link. The second satellite writes the target user's complete session state into the second hot buffer. When a target user enters the beam coverage area of ​​the second satellite and initiates a handover, the second satellite hits the target user's complete session state in the second hot buffer and completes the session continuation locally.

8. A method for offline autonomous operation of a spaceborne core network based on orbit prediction, characterized in that, The method includes: The ground station determines the beam dwell time window for the ground cell within the future orbital period of the first satellite based on the ephemeris of the first satellite; based on the historical access logs of the ground cell for the first satellite, it predicts the access probability of each candidate access user in the ground cell. The access probability of a user is the probability that the user initiates an access request to the first satellite within the beam dwell time window. Based on the access probability of each candidate access user in the ground cell, the ground station performs hot prefetching on high-probability users to obtain a complete session state snapshot of high-probability users, performs warm prefetching on medium-probability users to obtain the session context skeleton and subscription digest of medium-probability users, and registers the user identifier and default policy pointer of low-probability users to obtain prefetch packets; the prefetch packets are then sent to the first satellite via the satellite-ground link. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user to the first satellite within the beam dwell time window, and implements offline autonomy of the onboard core network according to the following steps: Based on a snapshot of the complete session state of a high-probability user, establish a complete session for that user. Based on the session context skeleton and subscription summary of medium-probability users, a fast session is established for medium-probability users; Based on the user identifier of low-probability users and the default policy pointer, establish restricted sessions for low-probability users.

9. The offline autonomous method for spaceborne core network based on orbit prediction according to claim 8, characterized in that, The method further includes: The first onboard autonomous policy engine writes the complete session state snapshot for high-probability users in the prefetch packet into the first hot state cache, writes the session context skeleton and subscription summary for medium-probability users in the prefetch packet into the first warm state cache, and writes the user identifier and default policy pointer for low-probability users in the prefetch packet into the first cold state cache. If the satellite-to-ground link is interrupted, the first satellite responds to the access requests initiated by each candidate access user within the beam dwell time window by sequentially querying the first hot buffer, the first warm buffer, and the first cold buffer to perform offline autonomy of the onboard core network.

10. The offline autonomous method for spaceborne core network based on orbit prediction according to claim 9, characterized in that, Based on the historical access logs of the ground cell, in addition to predicting the access probability of each candidate access user in the ground cell, the access time window of each candidate access user in the ground cell is also predicted. Bind a lifetime to each candidate access user in the ground cell that is aligned with the beam dwell time window; When the expected access time window of a high-probability user arrives and the high-probability user has not yet accessed the first satellite, the on-board cache manager will downgrade the high-probability user from hot state to warm state, release the complete session state snapshot of the high-probability user in the first hot state cache, and retain only the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache. When the end of the safe time window arrives and the high-probability user has not yet connected to the first satellite, the on-board cache manager will further downgrade the downgraded high-probability user from the warm state to the cold state, release the session context skeleton and subscription summary of the downgraded high-probability user in the first warm state cache area, and retain the user identifier and default policy pointer of the downgraded high-probability user only in the first cold state cache area. When the lifespan expires and the beam dwell window ends, the on-board cache manager deletes the user identifiers and default policy pointers of each candidate access user in the first cold buffer.

Images