Low-overhead space vector segment routing method for massive constellation networks

Abstract

The application discloses a large-scale constellation network low-overhead space vector segmentation routing method, and mainly solves the problems of high routing overhead and reduced network performance in the prior art, and the scheme is as follows: inter-satellite and cross-layer links of different constellations are respectively established; the earth surface is divided into regions, and the satellite constellations are grouped according to the regions; in the same group, a medium-orbit satellite is taken as a controller of a low-orbit satellite, and is responsible for routing control and management of the low-orbit satellite; the low-orbit satellite checks the inter-satellite link with a neighbor satellite and the queue length of the inter-satellite transmitter of the low-orbit satellite; the medium-orbit satellite receives the link fault feedback information and the queue information report of the low-orbit satellite in the management range of the medium-orbit satellite, obtains regional topological state information, and calculates three routing tables of a single-layer region, an inter-region and a multi-layer inter-layer; the low-orbit satellite utilizes the three routing tables to perform different on-satellite routing forwarding according to the position relationship of a source satellite and a destination satellite. The application reduces routing overhead, improves network performance, and can be used for a large-scale constellation network.

Description

Technical Field

[0001] This invention belongs to the field of satellite communication network technology, and in particular relates to a low-overhead space vector segmentation routing method that can be used in large-scale constellation networks. Background Technology

[0002] As a crucial component of mobile communications, low-Earth orbit (LEO) satellite communication systems, characterized by their innovative technology, rapid response, flexible application, and low cost, are attracting increasing attention. With the gradual maturation of a series of aerospace-related technologies and significant advancements in the research and manufacturing of LEO small satellites, the immense value of large-scale LEO constellation networks in the global communications field has become increasingly apparent. Large-scale LEO small satellites offer significant advantages such as rapid response, flexible use, and distributed deployment. They are also easily mass-produced, reducing aerospace costs. Through large-scale satellite networking, various complex communication tasks can be accomplished, meeting the transmission requirements of future high-capacity and diversified services. This makes large-scale constellation networks the main development trend of future satellite communication networks. A prime example is SpaceX's Starlink project. In 2015, Elon Musk's SpaceX announced its Starlink project, consisting of a multi-layered LEO constellation of 12,000 satellites, aiming to provide high-speed broadband access globally using a large-scale LEO satellite network. By 2019, the total number of satellites had increased to 42,000, with satellite launches expected to be completed by 2027. To date, SpaceX has completed 37 launches, with a total of 2,137 satellites in orbit, and a large-scale constellation network is being transformed from a conceptual idea into reality.

[0003] Routing is one of the key technologies for achieving network interconnection. Due to the dynamic nature of satellite networks, their topology changes frequently, making existing terrestrial network routing technologies unsuitable for direct application. Therefore, many researchers both domestically and internationally have studied satellite network routing technologies. For different satellite network configurations, existing satellite network technologies can be categorized into low-Earth orbit (LEO) satellite network routing technologies and hybrid orbit (HEO) satellite network routing technologies.

[0004] Existing low-Earth orbit (LEO) routing technologies commonly employ time-based virtualization (TVR) and location-based virtualization (RVR) technologies, respectively shielding the dynamic nature of satellite network topology from both temporal and spatial perspectives. A representative example of TVR-based satellite network routing is snapshot routing, which fixes the dynamic satellite network topology as a continuous static topology, calculating routes on this static topology using existing technologies. RVR-based routing aims to spatially shield the dynamic nature of the satellite network by mapping constantly moving satellites to one-to-one ground locations or virtual nodes, thus binding dynamic satellites to static ground or virtual nodes. In these routing technologies, ground-based systems typically perform route pre-calculation and periodically update routes, propagating the routing table to the entire satellite network. Satellites do not need to perform route calculations but require storing a large number of routing tables. As the scale of satellite networks continues to increase, the overhead of routing table propagation and updates increases dramatically, and static route calculation struggles to adapt to the dynamic changes in satellite networks, lacking the ability to monitor network congestion and link failures.

[0005] Existing hybrid orbit satellite network routing technologies mostly employ a model where higher-level constellations aggregate network-wide state information to calculate the entire network's routing table, and then distribute the routing table to lower-Earth orbit (LEO) constellations. The Hierarchical Satellite Routing Protocol (HSRP) proposed by Lee et al. uses distributed computing for routing calculation, requiring mid-Earth orbit (MEO) satellites to collect partial network state information, which is then aggregated into a complete network state information table and sent to LEO satellites, where the LEO satellites themselves are responsible for routing calculations. In contrast, Satellite Packet Routing Protocol (SGRP), Hierarchical QoS Routing Protocol (HQRP), and Adaptive QoS Routing Protocol (ARPQ) all use centralized computing. Higher-level satellites aggregate network-wide state information layer by layer, calculate the routing table, and send it to LEO satellites. In large-scale constellation networks, the interaction between higher-level satellites to obtain network-wide state information incurs significant routing overhead. The large distances between higher-level satellites mean that single-hop data transmission results in substantial propagation delays. Furthermore, the need to aggregate the state information collected by each satellite to ensure that every higher-level satellite receives the complete network state information necessitates frequent interactions between higher-level satellites, leading to substantial routing overhead. In summary, the most critical problem facing large-scale constellation network routing technologies is the high routing overhead. To generate satellite network topology and statistically analyze network load distribution for routing calculations, collecting network status information is essential. Furthermore, after routing calculations, the routing information needs to be distributed to each satellite node, all of which incur routing overhead. Taking snapshot routing technology as an example, the ground control center calculates the entire network routing table based on the predictability of the satellite network. Several ground gateway stations then upload the entire network routing table to the visible satellites, and finally, the entire network routing table is disseminated throughout the network. The amount of routing information data transmitted and the time consumed in the uploading and dissemination of the entire network routing table constitute the spatial and temporal overhead of routing. In a satellite network scenario with four inter-satellite links per satellite and a link rate of 1Gbps, when the number of satellites is only 72, the spatial overhead of a single routing table uploading and propagation is 9.8565Mb, and the time overhead is 0.0970 seconds. However, when the number of satellites reaches 756, the spatial overhead of a single routing table uploading and propagation reaches a staggering 10.8592Gb, and the time overhead also rises sharply to 0.3691 seconds, severely restricting the performance of the satellite network and increasing the network burden. Therefore, reducing routing overhead has become the most critical issue facing routing technology in large-scale constellation networks.

[0006] Meanwhile, large-scale constellation network routing technology also faces the problem of data forwarding path growth. Due to the dynamic nature of satellite networks, the relationship between users and satellites can change. When service data is transmitted, the destination user's access satellite may change. In this case, the traditional satellite node-based routing transmission mode will encounter a secondary transmission detour problem, requiring the service data to be forwarded to the original access satellite first, and then forwarded from the original access satellite to the new access satellite, causing service transmission detours. Moreover, as the number of satellites increases, the number of transmission hops for the same inter-satellite transmission distance will increase significantly, leading to increased data forwarding paths and reduced network performance. Therefore, how to solve the problem of increased data forwarding paths is also one of the important issues for large-scale constellation network routing technology.

[0007] Furthermore, how to utilize multi-layered low-Earth orbit (LEO) constellations in large-scale constellation networks is also an important issue to consider. Multi-layered LEO constellations are a development trend in large-scale constellation networks. SpaceX's Starlink Phase 1 project includes five layers of LEO constellations at different orbital altitudes, totaling 4,425 LEO satellites. Currently, the more than 2,000 Starlink satellites in orbit are distributed at altitudes of approximately 300km to 550km, and the entire constellation consists of many layers of LEO constellations. However, existing satellite routing technologies have not yet researched multi-layered LEO constellations. Therefore, how to establish relationships between multi-layered LEO constellations and utilize their characteristics to enable services to be forwarded across LEO constellation layers to improve network performance is also an important issue that large-scale constellation network routing technologies need to consider.

[0008] Therefore, with the trend of satellite networks gradually developing towards large scale, researching large-scale constellation network routing technology with low routing overhead and high network performance is of great practical significance for the future construction of my country's satellite communication network as well as the construction of national defense and civilian networks. Summary of the Invention

[0009] The purpose of this invention is to propose a low-overhead space vector segmentation routing method for large-scale constellation networks, in order to solve the problems of excessive routing overhead and reduced network performance caused by the application of existing routing technologies to large-scale constellation network routing technology, reduce routing overhead, improve network performance, and support routing operations in large-scale constellation networks.

[0010] To achieve the above objectives, the technical solution of the present invention includes the following:

[0011] (1) In a large-scale satellite constellation that includes multiple low-Earth orbit constellations, establish inter-satellite links within a single low-Earth orbit constellation, cross-layer links between multiple low-Earth orbit constellations, and cross-layer links between medium-Earth orbit constellations and low-Earth orbit constellations to complete large-scale constellation networking.

[0012] (2) Divide the Earth’s surface into several regions, group the satellite constellation according to the regions, and set up the access satellites for user terminals. In the same group, the medium-orbit satellite will act as the controller of the low-orbit satellite and will only be responsible for the routing control and management of the low-orbit satellites in the group.

[0013] (3) Low-Earth orbit satellites check inter-satellite links with neighboring satellites:

[0014] (3a) When the inter-satellite link between a low-Earth orbit satellite and its neighboring satellite has not communicated within a check period, the low-Earth orbit satellite sends a hello packet to its neighboring satellite to confirm whether the inter-satellite link is normal:

[0015] If a low-Earth orbit satellite receives a hello packet reply from a neighboring satellite within the specified time, it indicates that the inter-satellite link is normal.

[0016] Otherwise, it indicates that the inter-satellite link is faulty, and the low-Earth orbit satellite sends a fault report to the medium-Earth orbit satellite;

[0017] (3b) Low-orbit satellites dynamically change their inspection time based on their current activity level to reduce interaction overhead;

[0018] (4) Low-Earth orbit satellites check the length of their own inter-satellite transmitter queue:

[0019] (4a) Low-Earth orbit satellites periodically check whether their inter-satellite transmitter queue length exceeds the set threshold Qt:

[0020] If the length exceeds the limit, the low-Earth orbit satellite sends a queue information packet to the medium-Earth orbit satellite to report its own inter-satellite transmitter queue length.

[0021] Otherwise, low-Earth orbit satellites will not send queue information packets to medium-Earth orbit satellites;

[0022] (4b) Low-Earth orbit satellites dynamically change their inspection cycle based on the current length of their inter-satellite transmitter queue in order to reduce reporting overhead;

[0023] (5) The medium-orbit satellite receives queue information reports and fault feedback information from the low-orbit satellites within its management range to obtain regional topology status information;

[0024] (6) Based on the satellite position information stored locally and the network connection relationship, combined with the queue information report and fault feedback information of the low-orbit satellites within its management range, the medium-orbit satellites use the shortest path algorithm to periodically calculate the intra-area routing table and inter-area routing table of the single-layer routing area for each layer of the low-orbit constellation, and calculate the inter-layer routing table for the multi-layer low-orbit constellation, and simplify these three routing tables according to the actual needs of each low-orbit satellite, and then send them to the low-orbit satellites within its management range in the form of routing information packets;

[0025] (7) The low-orbit satellite receives the routing information packets sent by the medium-orbit satellite and stores the three routing tables locally;

[0026] (8) In large-scale satellite network routing scenarios, based on the positional relationship between the source and destination satellites, low-Earth orbit satellites utilize three routing tables for different on-board routing forwarding:

[0027] If the source satellite and the destination satellite are located in the same low-Earth orbit constellation and in the same location area, a single-level routing process within the routing area is executed, and a single-segment routing transmission is performed. That is, the low-Earth orbit satellite uses the routing table within the single-level routing area to forward the service data from the source satellite to the destination satellite through a single-segment minimum cost forwarding path.

[0028] If the source satellite and the destination satellite are in the same low-Earth orbit (LEO) constellation but in different location areas, a single-level routing area inter-interval routing process is executed, involving multi-segment vector routing transmission. Specifically, the LEO satellite first uses the single-level routing area's inter-interval routing table to find the relay satellite address for forwarding the service data to the destination location area, and then uses the single-level routing area's intra-area routing table to forward the service data to the relay satellite. Finally, it determines whether the relay satellite and the destination satellite are located within the same routing area.

[0029] If the relay satellite and the destination satellite are located in the same routing area, a single-layer routing process within the routing area is executed to forward the service data to the destination satellite;

[0030] Otherwise, use the single-layer routing area's inter-routing table to find the next relay satellite;

[0031] If the source satellite and the destination satellite are located in different LEO constellation layers, a cross-layer routing process is executed, involving cross-layer routing transmission and multi-segment vector routing transmission. Specifically, the LEO satellite first uses the inter-layer routing table to forward the service data to the cross-layer relay satellite in the LEO constellation layer where the destination satellite resides; then, it determines whether the cross-layer relay satellite and the destination satellite are located within the same routing area.

[0032] If the cross-layer relay satellite and the destination satellite are located in the same routing area, the single-layer routing area routing process is executed to forward the service data to the destination satellite;

[0033] Otherwise, execute the single-layer routing area inter-routing procedure.

[0034] Compared with the prior art, the present invention has the following advantages:

[0035] 1. Low routing overhead

[0036] Traditional satellite network routing technologies are unsuitable for large-scale networks due to their limited applicability, neglect of routing overhead, and inadequate handling of data transmission path growth caused by satellite dynamism. Furthermore, they lack cross-LEO constellation layer routing capabilities to accommodate multi-layered LEO constellations. This invention employs a low-overhead network state update method, leveraging the wide coverage and high computing power of medium-Earth orbit (MEO) satellites for distributed management of LEO satellites. MEO satellites only need to collect regional topology information for routing calculations, eliminating the need for collecting full network topology information, thus reducing routing overhead. Additionally, this invention utilizes an activity-based LEO satellite interaction mechanism and a congestion-state-based LEO satellite queue reporting mechanism, dynamically adjusting the inspection cycle based on different satellite and link states. Moreover, the simplification of the three routing tables sent to LEO satellites by MEO satellites further reduces routing overhead.

[0037] 2. High network performance

[0038] This invention, in its efficient space vector segmentation forwarding method, first establishes space vector forwarding channels with fixed geographical locations and connects multiple forwarding channels through a segmentation mechanism. Data forwarding is then performed based on geographical location, shielding the dynamic nature of satellite networks, avoiding detours in service transmission, and reducing the path length of service data transmission, thereby improving network performance. Simultaneously, because this invention adds a cross-low-Earth orbit constellation layer forwarding method, it achieves cross-layer traffic offloading of services, alleviating congestion problems in single-layer satellite networks and further optimizing network performance. Attached image description:

[0039] Figure 1 This is a large-scale satellite network routing scenario;

[0040] Figure 2 This is a flowchart illustrating the overall implementation of the present invention;

[0041] Figure 3 This is a sub-flowchart of the medium-orbit satellite's calculation of the routing table within a single-layer routing area in this invention;

[0042] Figure 4 This is a sub-flowchart of the calculation of the single-layer routing table between routing areas by a medium-orbit satellite in this invention;

[0043] Figure 5 This is a sub-flowchart of the calculation of the inter-layer routing table for medium-orbit satellites in this invention;

[0044] Figure 6 This is a comparison chart of end-to-end delay simulations between the present invention and existing technologies;

[0045] Figure 7 This is a simulation comparison chart of packet loss rates between the present invention and existing technologies;

[0046] Figure 8This is a comparison chart of the throughput simulations of the present invention and existing technologies;

[0047] Figure 9 This is a simulation comparison chart of the routing space overhead of the present invention and the prior art;

[0048] Figure 10 This is a simulation comparison chart of the routing time cost of the present invention and the prior art. Detailed Implementation

[0049] Routing technology is fundamental to end-to-end service forwarding, but existing satellite network routing technologies lead to increased routing overhead and reduced network performance when applied to large-scale constellation networks. To address this, this invention proposes a low-overhead space vector segmentation routing technique suitable for large-scale constellation networks, aiming to reduce routing overhead and improve network performance.

[0050] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings:

[0051] Large-scale satellite network routing scenarios such as Figure 1 As shown, it includes a medium-Earth orbit (MEO) constellation, a multi-layer low-Earth orbit (LEO) constellation, and user terminals. The MEO constellation calculates three routing tables—intra-area, inter-area, and inter-layer—and distributes them to the LEO satellites within its management range. The LEO constellation executes the on-board routing and forwarding process from uplink satellites to downlink satellites. User terminals select satellites above their location area as their access satellites, including source user terminals and destination user terminals. The access satellite for the source user terminal is called the uplink satellite, and the access satellite for the destination user terminal is called the downlink satellite. After initiating a service transmission, the source user terminal first sends the service data to the uplink satellite, then executes the on-board routing process. Through hop-by-hop routing and forwarding in the satellite network, the service data is forwarded to the downlink satellite above the destination user terminal's location area. The downlink satellite then sends the service data to the destination user terminal, completing the communication process from source to destination.

[0052] The large-scale constellation network configuration parameters for this example are shown in Table 1, comprising a Medium Earth Orbit (MEO) constellation and a three-layer Low Earth Orbit (LEO) constellation, totaling 2112 satellites. The MEO constellation primarily handles routing control and management for the LEO constellation. In the three-layer LEO constellation, LEO-1 is a polar orbit constellation, while LEO-2 and LEO-3 are inclined orbit constellations. LEO-1 has a higher orbital altitude, with satellites distributed globally, supporting service access in various regions and providing global coverage. LEO-2 and LEO-3 have lower orbital altitudes, primarily distributed in low- and mid-latitude regions, mainly serving high-density service areas globally and supporting high-capacity and real-time service transmission.

[0053] Table 1 Configuration parameters of large-scale constellation networks

[0054]

[0055] Reference Figure 2 The implementation steps for spatial vector segmentation routing in the aforementioned large-scale satellite network routing scenario are as follows:

[0056] Step 1: Establish multiple inter-satellite links to complete large-scale constellation networking.

[0057] 1.1) Establish inter-satellite links within a single-layer low-Earth orbit constellation according to the numbering rules:

[0058] Two inter-satellite links are established between the satellite and two adjacent visible satellites in the same orbit.

[0059] Inter-orbit links can be established between satellites in adjacent orbits that simultaneously meet the following three conditions: same orbital number, same latitude direction of travel, and visible relationship. A maximum of two inter-orbit links can be established. Among them, the LEO-1 layer is a polar orbit constellation, and there is a reverse gap between the last orbit and the first orbit, so inter-orbit links cannot be established. The LEO-2 and LEO-3 layers are inclined orbit constellations, and there is no reverse gap.

[0060] When a satellite moves into a high-latitude region, that is, when the satellite is north of 66 degrees north latitude or south of 66 degrees south latitude, the inter-satellite link is broken and re-established when the satellite moves out of the high-latitude region.

[0061] 1.2) Establish inter-layer links between multiple low-Earth orbit constellations using the shortest distance:

[0062] Cross-layer links between multiple LEO constellations are only established between adjacent LEO constellation layers. Based on the orbital altitude distribution of LEO-1 > LEO-2 > LEO-3, cross-layer links are established between LEO-1 and LEO-2, and between LEO-2 and LEO-3. At most, only 1 / 7 of the satellites in each LEO constellation layer have cross-layer links.

[0063] From the perspective of construction and maintenance costs, cross-layer links are only established for some low-Earth orbit (LEO) satellites, and in adjacent LEO constellation layers, lower-layer LEO satellites select the nearest upper-layer LEO satellite within their line of sight to establish cross-layer links.

[0064] 1.3) Establish cross-layer links between the MEO and LEO constellations using a partial node approach:

[0065] Medium-Earth orbit (MEO) satellites do not establish cross-layer links with low-Earth orbit (LEO) satellites outside their managed area. They only assume management responsibility for LEO satellites within their managed area, including collecting network status information and distributing routing tables.

[0066] Medium-Earth orbit (MEO) satellites establish cross-layer links with a maximum of one-third of the low-Earth orbit (LEO) satellites within their management range. These LEO satellites are called super nodes, while the other LEO satellites are called general nodes. If MEO satellites need to exchange data with general nodes, they need to use super LEO nodes for relay transmission.

[0067] Step 2: Divide the Earth's surface into regions.

[0068] The process of dividing the Earth's surface into regions involves first dividing the Earth's surface into several large regions, or routing regions, based on latitude and longitude, and then further dividing each routing region into multiple smaller regions, or location regions.

[0069] This example divides the global Earth's surface into 12 routing zones based on the latitude lines of 24.23° and -24.23° and the longitude lines of -90°, 0°, and 90°. Since the equator is the longest latitude line, and latitude lines become shorter the farther from the equator, routing zones closer to the equator have smaller latitudinal spans to maintain a similar size for each zone. Each routing zone is further divided into 63 location zones, resulting in a total of 12 routing zones and 756 location zones globally.

[0070] Each location area has a globally unique identifier that indicates its geographical location. This identifier is called the location area identifier and is 16 bits in size. It includes three fields: extension bit, routing area, and location area. The size and function of each field are shown in Table 2.

[0071] Table 2

[0072]

[0073] Satellites located above this location area and users within this location area have logical addresses equivalent to their location area identifiers. In addition to the logical address, each satellite also has a unique and unchanging identifier called the satellite physical address, which is 16 bits in size and includes four fields: extension bits, satellite position, orbital bits, and orbital number bits. The size and function of each field are shown in Table 3.

[0074] Table 3

[0075]

[0076] Step 3: Group the satellite constellations according to the region and set the access satellites for user terminals.

[0077] 3.1) Satellites located in the same routing area will be grouped together. Both medium-Earth orbit (MEO) and low-Earth orbit (LEO) satellites will directly determine their routing area address and grouping through their own latitude and longitude location information.

[0078] This method uses the satellite's own position information to determine its group, eliminating the need for pre-stored satellite group information. Medium-Earth orbit (MEO) and low-Earth orbit (LEO) satellites only need their onboard positioning equipment to determine their latitude and longitude positions. Then, based on the latitude and longitude boundary rules for route area division, they directly determine their route area and group. This method avoids complex group relationship calculations, reduces satellite storage pressure, and enables low-overhead group determination and maintenance of group relationships.

[0079] 3.2) Within the same group, a medium-Earth orbit satellite located over a certain routing area is designated as the routing control manager for all low-Earth orbit satellites over that routing area. It is only responsible for collecting regional topology status information and distributing routing information within its own routing area.

[0080] Due to limited satellite resources, it would be too costly and difficult to implement if the medium-Earth orbit controller were to equip all low-Earth orbit satellites within its management range with inter-satellite radio transceivers. Therefore, all low-Earth orbit satellites are divided into super nodes and general nodes, with a node ratio of approximately 1:2. The medium-Earth orbit controller only needs to equip the super nodes with inter-satellite radio transceivers.

[0081] When distributing routing information, the medium-Earth orbit (MEO) controller can communicate directly with the supernode via the MEO-Low Earth orbit (LEO) inter-layer link, sending the routing information to the supernode. The supernode then forwards the routing information to the general nodes via the inter-satellite links within the single-layer LEO constellation. Similarly, when collecting regional topology status information, the general nodes first aggregate the information to the supernode, which then uploads it to the MEO controller. MEO satellites only need to be equipped with inter-satellite transceivers for 1 / 3 of the LEO satellites, reducing the construction cost and difficulty of MEO satellites.

[0082] After collecting regional topology status information, the medium-Earth orbit (MEO) satellite does not need to perform MEO constellation-level interactions to obtain the overall network topology status information. It can directly perform route calculations and distribute simplified routing tables (without redundant information) to each low-Earth orbit (LEO) satellite within its managed area. Obtaining overall network status information through mid-level interactions incurs significant routing overhead. MEO satellites experience substantial single-hop propagation delays and require multiple interactions to obtain the complete network topology status information, resulting in substantial routing costs. Therefore, collecting only regional topology status information and distributing simplified routing tables to each LEO satellite effectively reduces routing overhead.

[0083] Step 4: Low-Earth orbit satellites check inter-satellite links with neighboring satellites.

[0084] 4.1) When a low-Earth orbit (LEO) satellite interacts with its neighbors in the same LEO constellation layer and with neighbors in adjacent LEO constellations, it checks its inter-satellite links. If no communication activity occurs between a LEO satellite and its neighboring satellites within a check period, it checks whether its own inter-satellite links with its neighbors are normal.

[0085] If a low-Earth orbit satellite receives a hello packet reply from a neighboring satellite within the specified time, it indicates that the inter-satellite link is normal.

[0086] If low-Earth orbit satellite 1 does not receive a hello packet from low-Earth orbit satellite 2 within the specified time, it is determined that the inter-satellite link has failed, and the failure status of the link is reported to the medium-Earth orbit satellite.

[0087] In this example, each LEO satellite has 2-6 neighboring satellites, including 2 neighboring satellites in the same LEO constellation layer, 0-2 neighboring satellites in adjacent orbits within the same LEO constellation layer, and 0-2 neighboring satellites in different layers of adjacent LEO constellations. Let LEO satellite 1 and LEO satellite 2 be neighbors. LEO satellite 1 checks the status of its inter-satellite link with its neighbor. If it finds that no data transmission has occurred on the "LEO satellite 1-LEO satellite 2" inter-satellite link for a certain period, it initiates inter-satellite neighbor information exchange. LEO satellite 1 sends a "hello" packet to LEO satellite 2 to confirm the inter-satellite link is functioning correctly. Upon receiving this "hello," LEO satellite 2 immediately sends a reply "hello" packet to LEO satellite 1.

[0088] 4.2) Low-Earth orbit satellites dynamically adjust their inspection times based on current activity levels:

[0089] The check time is related to satellite activity, ranging from 30 seconds to 60 seconds. When satellite activity is high, it means the satellite will provide services frequently, requiring constant assurance of normal inter-satellite link operation. In this case, the check time is shorter to ensure the satellite can provide services under high-pressure conditions. When satellite activity is low, it means the satellite provides services less frequently and does not require as many inter-satellite link checks. In this case, the check time is longer to reduce the number of interactions with neighboring satellites, thereby saving link bandwidth and reducing interaction overhead.

[0090] 4.2.1) Determine the current activity level based on the satellite's location and the current time:

[0091] The satellite's activity level is higher when it is located in a densely populated area with high traffic; it is lower when it is located in a sparsely populated area with low traffic; the activity level is higher when the current time is during the daytime when traffic is high; and lower when the previous time was during the nighttime when traffic is low.

[0092] 4.2.2) Set the check time according to the activity level. When the satellite activity is high, set the check time to 1 / 2 of the routing table calculation cycle. When the satellite activity is low, set the check time to the routing table calculation cycle.

[0093] Step 5: Low-Earth orbit satellites check the length of their own inter-satellite transmitter queue.

[0094] 5.1) Set the queue length threshold Qt, and the low-Earth orbit satellite periodically checks its own queue status:

[0095] If a queue length exceeds the threshold Qt, a queue information packet is sent to the medium-Earth orbit satellite to report its own queue congestion status. In this example, the maximum queue size is set to 6, meaning the queue can store a maximum of 6 data packets. Different inter-satellite link queue lengths correspond to different congestion states, and corresponding check periods are set. When the queue length is no more than 2, the inter-satellite link is in an idle state and no queue information packet is sent; the check period is set to 90 seconds. When the queue length is greater than 2 but no more than 4, the inter-satellite link is in a lightly congested state and a queue information packet is sent; the check period is set to 60 seconds. When the queue length is greater than 4, the inter-satellite link is in a heavily congested state and a queue information packet is sent; the check period is set to 30 seconds.

[0096] 5.2) Supernodes can directly send queue information packets to medium-Earth orbit satellites, while ordinary nodes need to first send queue information packets to supernodes in the same routing area as themselves, and then the supernodes will forward the queue information packets to medium-Earth orbit satellites.

[0097] Step 5: The medium-Earth orbit satellite receives queue information reports and fault feedback information from the low-Earth orbit satellites within its management range to obtain regional topology status information.

[0098] Step 6: The medium-Earth orbit satellite calculates the intra-area and inter-area routing tables for a single layer, as well as the inter-layer routing tables.

[0099] Based on locally stored satellite position information and overall network connectivity, combined with queue information reports and fault feedback information from low-Earth orbit (LEO) satellites within their management range, and using the shortest path algorithm at the cost of propagation delay and queuing delay, medium-Earth orbit (MEO) satellites calculate intra-area routing tables and inter-area routing tables for each LEO constellation layer every minute, and calculate inter-layer routing tables for multiple LEO constellations. The specific implementation is as follows:

[0100] 6.1) Based on the locally stored satellite position information and the connection relationships within the routing area, combined with the low-Earth orbit satellite queue information reports and fault feedback within its management range, the medium-Earth orbit satellites generate the routing area topology for each layer of the low-Earth orbit constellation and calculate the routing table within the single-layer routing area.

[0101] like Figure 3As shown, MEO satellites generate a single-layer regional network topology based on pre-stored single-layer routing area connectivity relationships. They then calculate inter-satellite link propagation delay using satellite position information and inter-satellite link queuing delay using low-Earth orbit (LEO) satellite queue information reports. Faulty links are eliminated based on LEO satellite fault feedback, generating a directed graph within the single-layer routing area. The link cost is propagation delay + queuing delay. Finally, the shortest path algorithm is used to calculate the routing table within the single-layer routing area, representing the next-hop satellite for LEO satellites at the same level above the routing area to forward service data to each other.

[0102] 6.2) Based on the locally stored satellite position information and single-layer connection relationship, and combined with the low-Earth orbit satellite queue information reports and fault feedback within its management range, the medium-Earth orbit satellites calculate the single-layer routing table for each layer of the low-Earth orbit constellation.

[0103] like Figure 4 As shown, MEO satellites generate a single-layer regional network topology based on pre-stored single-layer connection relationships. Then, they calculate the inter-satellite link propagation delay using satellite position information and the inter-satellite link queuing delay within the routing area using low-Earth orbit (LEO) satellite queue information reports. The inter-satellite link queuing delay outside the routing area is unknown and is defaulted to 0. Faulty links are removed based on LEO satellite fault feedback, generating a single-layer LEO constellation directed graph. The link cost is propagation delay + queuing delay. Finally, the shortest path algorithm is used to calculate the inter-inter-interval routing table for the single-layer routing area, representing relay satellites that forward service data from LEO satellites above the routing area to satellites at the same layer above the destination location area outside the routing area.

[0104] 6.3) The medium-orbit satellite controller calculates the inter-layer routing table based on the locally stored network connectivity and satellite position information, combined with the low-orbit satellite queue information reports and fault feedback within its management range.

[0105] like Figure 5 As shown, MEO satellites generate the entire network topology based on pre-stored network connectivity relationships, then calculate the inter-satellite link propagation delay using satellite position information, and calculate the inter-satellite link queuing delay within the routing area using low-Earth orbit (LEO) satellite queue information reports; the inter-satellite link queuing delay outside the routing area is unknown and defaults to 0. Faulty links are removed based on LEO satellite fault feedback, generating a multi-layer LEO constellation directed graph, where the link cost is propagation delay + queuing delay; finally, the shortest path algorithm is used to calculate the inter-layer routing table, indicating the next-hop satellite from which LEO satellites above the routing area forward service data to different LEO satellite layers.

[0106] Step 7: The medium-Earth orbit satellite simplifies the three routing tables and sends routing information packets to the low-Earth orbit satellite.

[0107] The medium-Earth orbit satellite takes the corresponding row from three routing tables for each low-Earth orbit satellite within its management range, adds the satellite address table header to simplify the routing table, puts it into the routing information packet, and sends it to the low-Earth orbit satellite.

[0108] The low-Earth orbit satellite receives routing information packets from the medium-Earth orbit satellite and stores the three routing tables locally.

[0109] Step 8: After receiving the service data packet uploaded by the user, the source satellite queries the medium-Earth orbit satellite for the physical and logical addresses of the destination satellite. Based on the positional relationship between the source and destination satellites, the low-Earth orbit satellite uses three routing tables to perform different on-board routing.

[0110] 8.1) Execution of routing procedures within a single-layer routing area:

[0111] For routing forwarding where the source satellite and the destination satellite are in the same low-Earth orbit constellation layer and the same routing area, the single-layer routing area intra-area routing process is executed: if the data packet does not reach the destination satellite, the single-layer routing area intra-area routing table is queried to find the inter-satellite next-hop satellite corresponding to the destination satellite and forward the data packet until the data packet reaches the destination.

[0112] 8.2) Execution of the inter-regional routing process in a single-layer routing area:

[0113] If the source satellite and the destination satellite are in the same low-Earth orbit constellation layer but not in the same routing area, then a single-layer routing area inter-routing procedure is performed:

[0114] If the data packet does not reach the routing area, the relay satellite is found using the single-level routing area's inter-interval routing table, and then the data packet is forwarded to the relay satellite using the single-level routing area's intra-interval routing table.

[0115] If the relay satellite and the destination satellite are still not in the same routing area, continue searching for the next relay satellite and continue segmenting the route until the data packet reaches the destination routing area.

[0116] Once the data packet arrives at the destination routing area, the single-level routing area intra-area routing process is executed, querying the single-level routing area intra-area routing table to forward the data packet until the data packet reaches the destination satellite.

[0117] 8.3) Execution of cross-layer routing process:

[0118] If the source satellite and the destination satellite are not in the same low-Earth orbit (LEO) constellation layer, a cross-layer routing process is performed. This involves first forwarding the service data packet to the cross-layer relay satellite in the LEO constellation layer where the destination satellite resides using the cross-layer routing table, and then determining whether the relay satellite and the destination satellite are located in the same routing area.

[0119] If the cross-layer relay satellite and the destination satellite are located in the same routing area, the single-layer routing area routing process is executed to forward the service data to the destination satellite;

[0120] Otherwise, execute the single-layer routing area inter-routing procedure.

[0121] The effects of this invention can be further illustrated by the following simulation results:

[0122] I. Simulation Conditions

[0123] A large-scale constellation routing simulation scenario was built in OPNET, with the constellation configuration shown in Table 1. The simulation lasted for 4.5 minutes. Service data packets were generated starting at the 10th second of the simulation and stopping at the 4th minute. The size of each service data packet was 1 Mb. The source of the service data packets was each LEO satellite, and the rate at which each LEO satellite generated service data packets followed a Poisson distribution with the same mean. Assuming the source satellite was located in the LEO-i layer, the destination location area of ​​the service data packets was the location area where LEO-i layer satellites were located, and the destination satellite was a random LEO-i layer satellite located above these location areas. Considering the running capacity of the OPNET simulation software, the inter-satellite transceiver rate was reduced, and the inter-satellite link transceiver rate was set to 20 Mbps. Signaling information used independent bandwidth.

[0124] II. Simulation Content and Results

[0125] Simulation 1: Under the above conditions, the low-overhead vector segmentation routing method of the present invention and the existing snapshot routing method are compared through service end-to-end latency simulation. The results are as follows: Figure 6 As shown.

[0126] from Figure 6 As can be seen, the end-to-end latency of this invention is consistently lower than that of the existing snapshot routing method. With increasing service rates, the difference between the two gradually increases and eventually stabilizes. When the service rate is approximately 0.8 Gb / s, a significant difference in end-to-end latency between the two routing technologies begins to appear, with the end-to-end latency of this invention starting to be lower than that of the snapshot routing method. When the service rate exceeds 2 Gb / s, the end-to-end latency of both routing technologies tends to stabilize, and the end-to-end latency of this invention is approximately 0.05 seconds lower than that of the existing snapshot routing technology.

[0127] Simulation 2: Under the above conditions, a comparative simulation of packet loss rate was performed between the low-overhead vector segmentation routing method of the present invention and the existing snapshot routing method. The results are as follows: Figure 7 As shown.

[0128] from Figure 7 As can be seen, the packet loss rate of this invention is consistently lower than that of the snapshot routing method, and the difference between the two gradually increases with the increase of service rate. When the service rate is approximately 0.8 Gb / s, a significant difference in end-to-end latency between the two routing methods begins to appear, and the packet loss rate of this invention initially becomes lower than that of the snapshot routing method. Subsequently, with the increase of service rate, the growth trend of the packet loss rate of this invention is relatively slow.

[0129] Simulation 3: Under the above conditions, the throughput of the low-overhead vector segmentation routing method of the present invention and the existing snapshot routing method are compared through simulation. The results are as follows: Figure 8 As shown.

[0130] from Figure 8 As can be seen, the service data throughput of this invention is consistently higher than that of the snapshot routing method, and the difference between the two gradually increases with the increase of the service rate. At the same service rate, a higher packet loss rate results in fewer received service data packets and lower throughput. Because the packet loss rate of this invention is always lower than that of the snapshot routing method, and the difference in packet loss rate gradually increases with the increase of the service rate, the service data throughput is always higher than that of the snapshot routing method, and the throughput improvement is greater than that of the snapshot routing technology with increasing service rate.

[0131] Simulation 4: Under the above conditions, a simulation comparison of the spatial overhead between the low-overhead vector segmentation routing method of the present invention and the existing snapshot routing method is performed. The results are as follows: Figure 9 As shown.

[0132] from Figure 8 As can be seen, the routing space overhead of snapshot routing technology is about 26Gb / time, while the routing space overhead of the present invention is only about 0.048Gb / time, which reduces the routing time overhead by 99.8% compared with snapshot routing method. Figure 8 The routing space overhead of the present invention increases slightly with the increase of service rate, which is due to the increase in the number of congested inter-satellite links and the increase in the number of queue information packets sent.

[0133] Simulation 5 compares the time overhead of the low-overhead vector segmentation routing method of the present invention with that of the existing snapshot routing method under the above conditions. The results are as follows: Figure 9 As shown.

[0134] from Figure 9 As can be seen, the routing time cost of the snapshot routing method is about 15 seconds per cycle, while the routing time cost of the present invention is only 0.06 seconds per cycle, which is 99.6% lower than that of the snapshot routing method.

[0135] The simulation results above demonstrate that this invention outperforms snapshot routing technology in three network performance metrics: end-to-end latency, packet loss rate, and throughput, resulting in a significant improvement in network performance. This is because the geolocation-based forwarding mechanism in this invention reduces the data routing forwarding length, effectively lowering the end-to-end latency of service data. This shortens the transmission time of service data packets in the satellite network, thereby reducing the number of data packets in the satellite network, alleviating congestion problems, and effectively reducing packet loss caused by transmitter queue overflow, thus improving network throughput. Furthermore, this invention effectively reduces routing overhead while maintaining superior network performance, specifically reducing time overhead by 99.6% and routing space overhead by 99.8%, demonstrating scalability.

[0136] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them. Although the present invention has been described in detail with reference to the foregoing examples, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of their technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A large-scale constellation network low-overhead space vector segment routing method, characterized in that, Comprise as follows: (1) In a large-scale satellite constellation comprising multiple low earth orbit constellations, establish inter-satellite links within a single low earth orbit constellation, cross-layer links between multiple low earth orbit constellations, and cross-layer links between medium earth orbit constellations and low earth orbit constellations, to complete large-scale constellation networking; (2) Divide the earth's surface into several regions, group the satellite constellations according to the regions, and set the access satellites for user terminals, in the same group, the medium earth orbit satellite will serve as the controller of the low earth orbit satellite, and only be responsible for the routing control and management of the low earth orbit satellite within the group; (3) The low earth orbit satellite checks the inter-satellite link with the neighbor satellite: (3a) When the inter-satellite link between the low earth orbit satellite and the neighbor satellite does not occur within a check time, the low earth orbit satellite sends a hello packet to the neighbor satellite to confirm whether the inter-satellite link is normal: If the low earth orbit satellite receives a hello packet reply from the neighbor satellite within a specified time, it indicates that the inter-satellite link is normal; Otherwise, it indicates that the inter-satellite link is faulty, and the low earth orbit satellite feeds back the fault to the medium earth orbit satellite; (3b) The low earth orbit satellite dynamically changes the check time according to the current activity, to reduce the interaction overhead; (4) The low earth orbit satellite checks the length of its inter-satellite transmitter queue: (4a) The low earth orbit satellite periodically checks whether the length of its inter-satellite transmitter queue exceeds a set threshold Qt: If it exceeds, the low earth orbit satellite sends a queue information packet to the medium earth orbit satellite to report the length of its inter-satellite transmitter queue; Otherwise, the low earth orbit satellite does not send a queue information packet to the medium earth orbit satellite; (4b) The low earth orbit satellite dynamically changes the check period according to the current length of its inter-satellite transmitter queue, to reduce the reporting overhead; (5) The medium earth orbit satellite receives the queue information reports and fault feedback information of the low earth orbit satellites within its management range, and obtains the regional topology state information; (6) The medium earth orbit satellite uses the shortest path algorithm to periodically calculate the intra-zone routing table of the single-layer routing zone, the inter-zone routing table of the single-layer routing zone, and the inter-layer routing table of the multi-layer low earth orbit constellation, using the propagation delay and queuing delay as the link cost, based on the locally stored satellite position information and the global connection relationship, combined with the queue information reports and fault feedback information of the low earth orbit satellites within its management range, and simplifies the three routing tables according to the actual needs of each low earth orbit satellite, and then distributes them to the low earth orbit satellites within its management range in the form of routing information packets; (7) The low earth orbit satellite receives the routing information packets distributed by the medium earth orbit satellite, and stores the three routing tables locally; (8) In the large-scale satellite network routing scenario, according to the position relationship between the source satellite and the destination satellite, the low earth orbit satellite uses the three routing tables for different on-satellite routing forwarding: If the source satellite and the destination satellite are located in the same low earth orbit constellation and in the same position zone, the intra-zone routing process of the single-layer routing zone is performed, and single-segment routing transmission is performed, that is, the low earth orbit satellite uses the intra-zone routing table of the single-layer routing zone to forward the business data from the source satellite to the destination satellite through the minimum cost transfer path of a single segment; If the source satellite and the destination satellite are located in the same layer of the low-orbit constellation but in different location areas, a single-layer routing area inter-area routing process is performed for multi-segment vector routing transmission, i.e., the low-orbit satellite uses the inter-area routing table of the single-layer routing area to query the address of the relay satellite for forwarding the service data to the destination location area, and uses the intra-area routing table of the single-layer routing area to forward the service data to the relay satellite; then it is determined whether the relay satellite and the destination satellite are located in the same routing area: If the relay satellite and the destination satellite are located in the same routing area, an intra-area routing process of the single-layer routing area is performed to forward the service data to the destination satellite; Otherwise, the inter-area routing table of the single-layer routing area is used to find the next relay satellite. If the source satellite and the destination satellite are located in different layers of the low-orbit constellation, a cross-layer routing process is performed for cross-layer routing transmission and multi-segment vector routing transmission, i.e., the low-orbit satellite uses the inter-layer routing table to forward the service data to the cross-layer relay satellite of the low-orbit constellation layer where the destination satellite is located; then it is determined whether the cross-layer relay satellite and the destination satellite are located in the same routing area: If the cross-layer relay satellite and the destination satellite are located in the same routing area, an intra-area routing process of the single-layer routing area is performed to forward the service data to the destination satellite; Otherwise, a single-layer routing area inter-area routing process is performed.

2. The routing method of claim 1, wherein, The (1) in the large-scale satellite constellation comprising multiple layers of low-orbit constellations establishes inter-satellite links within a single layer of low-orbit constellations, cross-layer links between multiple layers of low-orbit constellations, and cross-layer links between medium-orbit constellations and low-orbit constellations, and achieves the following: For the inter-satellite links within a single layer of low-orbit constellations, the links are established according to a numbering rule, i.e., a satellite establishes a same-orbit inter-satellite link with the two adjacent satellites on the same orbit, and establishes a different-orbit inter-satellite link with the satellite on the adjacent orbit that has the same orbit number and the same latitude running direction. For the cross-layer links between multiple layers of low-orbit constellations, the links are established by the shortest distance method, i.e., in the adjacent low-orbit constellation layer, 1 / 7 of the lower low-orbit satellites select the upper low-orbit satellite closest in distance within the visible range to establish a cross-layer link. For the cross-layer links between medium-orbit constellations and low-orbit constellations, the links are established by the partial node method, i.e., a medium-orbit satellite establishes an inter-layer link with at most 1 / 3 of the low-orbit satellites within its management range.

3. The routing method of claim 1, wherein, The (2) divides the earth's surface into regions, which are divided into routing areas according to longitude and latitude, each routing area containing a plurality of location areas. The routing area is a concept based on routing calculation and management, which represents the actual management range of a medium-orbit satellite. The location area is a concept based on users, and a low-orbit satellite located above a certain location area will be responsible for the service access and downlink of users in that location area.

4. The routing method of claim 1, wherein, The (2) groups the satellite constellations according to the regions, which divides the satellites located above the same routing area into a group. The medium-orbit satellite and the low-orbit satellite will directly determine their routing area and group through their longitude and latitude position information.

5. The routing method of claim 1, wherein, The routing control management of the medium orbit satellite to the low orbit satellite in the (2) is that the medium orbit satellite located above a certain routing area acts as a control manager of the routing of all low orbit satellites in the routing area, is responsible for the collection of regional topology state information in the routing area and the distribution of routing information packets, and divides the low orbit satellites into super nodes and general nodes for different ways of communication: The super node can directly communicate with the medium orbit satellite; The general node cannot directly communicate with the medium orbit satellite, and the collection of the queue information packet of the general node and the sending of the routing information packet to the general node can be relayed through the super node.

6. The routing method of claim 1, wherein, The low orbit satellite in the (3b) dynamically changes the checking time according to the current activity, and realizes the following: (3b1) The current activity is determined according to the position of the satellite and the current time: If the satellite is located in a high-traffic area with high population density, the activity is high, If the satellite is located in a low-traffic area with low population density, the activity is low, If the current time is a high-traffic daytime period, the activity is high, If the current time is a low-traffic nighttime period, the activity is low; (3b2) The length of the checking time is set according to the high and low of the activity, that is, when the activity of the satellite is high, the checking time is set to 1 / 2 of the routing table calculation period, and when the activity of the satellite is low, the checking time is set to the routing table calculation period.

7. The routing method of claim 1, wherein, The low orbit satellite in the (4) dynamically changes the checking period according to the current length of the inter-satellite transmitter queue, which is divided into three states of idle state, light congestion state and heavy congestion state according to the length of the queue, and different checking periods are set according to the state of the inter-satellite link: The idle state does not need to send the queue information packet, and the checking period is set to be the longest, that is, 3 / 2 of the routing table calculation period; The light congestion state needs to send the queue information packet, and the checking period is short, that is, the routing table calculation period; The heavy congestion state needs to send the queue information packet, and the checking period is the shortest, that is, 1 / 2 of the routing table calculation period.

8. The routing method of claim 1, wherein, The (6) obtains the intra-zone routing table, the inter-zone routing table and the inter-layer routing table, and the contents are as follows: The intra-zone routing table of the single-layer routing area indicates the next hop satellite for the mutual forwarding of business data by the low orbit satellites in the same layer above the routing area; The inter-zone routing table of the single-layer routing area indicates the relay satellite for the low orbit satellites above the routing area to forward the business data to the same layer satellites above the destination position area outside the routing area; The inter-layer routing table indicates the next hop satellite for the low orbit satellites above the routing area to forward the business data to different low orbit satellite layers.

9. The routing method of claim 1, wherein, The (6) simplifies the three routing tables according to the actual needs of each low orbit satellite, that is, for each low orbit satellite, the corresponding row in the routing table is taken, and the satellite address table header is added.

10. The routing method of claim 1, wherein, The (8) large-scale satellite network routing scenario is composed of a medium orbit constellation, a multi-layer low orbit constellation and user terminals, wherein: The medium orbit constellation is used to calculate the three routing tables and distribute them to the low orbit satellites within its management range; The user terminal is used to select the satellite above the location area as the access satellite of itself, wherein the access satellite of the source user terminal is called the upload satellite, and the access satellite of the destination user terminal is called the download satellite; The low earth orbit constellation is configured to perform on-board routing and forwarding procedures from an uplink satellite to a downlink satellite.

Images