Method for dynamic routing optimization in satellite networks based on link state awareness

Abstract

This invention relates to the field of communication network technology, specifically to a dynamic routing optimization method for satellite networks based on link state awareness. This method determines the health of a link by combining its comprehensive transmission capacity with bit error rate and availability; determines a predicted health based on the trend of health changes; determines a routing metric based on the importance level of the service to be transmitted and the predicted health; selects a primary forwarding path and backup forwarding paths based on the cumulative routing metrics among candidate paths; determines a backup activation tendency value based on the data packet size and importance level of the service to be transmitted; and determines the final forwarding path for each service to be transmitted from the primary and backup forwarding paths based on the backup activation tendency value, using the final forwarding path for data transmission, thereby improving the stability of important service transmission and the utilization rate of link resources.

Description

Technical Field

[0001] This invention relates to the field of communication network technology, and more specifically to a dynamic routing optimization method for satellite networks based on link state awareness. Background Technology

[0002] With the development of low-Earth orbit satellite constellations, inter-satellite link communication, and integrated space-air-ground networks, satellite networks have gradually become an important component of long-distance communication, wide-area coverage communication, and emergency communication. Compared with fixed terrestrial networks, satellite nodes in satellite networks are typically in high-speed motion, and the network topology continuously changes with satellite orbital movement. Inter-satellite links are also affected by factors such as the relative positions of satellites, link orientation, space environment, and service load. To achieve continuous data service transmission in satellite networks, routing mechanisms are typically used to determine forwarding paths between source and destination satellite nodes, enabling data transmission via one or more inter-satellite links. Therefore, satellite network routing technology is a crucial technical direction for ensuring the stable operation of satellite communication networks and the efficiency of service transmission.

[0003] Existing satellite network routing typically relies on pre-set routing tables, fixed-period updates, or single indicators such as link connectivity and hop count as the basis for path selection. In scenarios involving high-speed satellite node movement and frequent changes in inter-satellite link status, this approach struggles to reflect the combined impact of changes in link bandwidth usage, transmission latency, bit error rate, and link availability in a timely manner. When the quality of some links deteriorates or service loads become concentrated, existing routing methods tend to continue using the original forwarding paths, leading to delayed path selection, uneven distribution of link resources, and decreased stability of some service transmissions. Furthermore, different services have varying requirements for transmission reliability, continuity, and link resource usage. Existing routing mechanisms often lack the ability to adjust paths based on service attributes, making it difficult to ensure stable transmission of critical services while simultaneously balancing the overall utilization of satellite network link resources. Summary of the Invention

[0004] To address the technical challenge of balancing the stability of critical service transmissions with efficient link resource utilization, this invention aims to provide a dynamic routing optimization method for satellite networks based on link state awareness. The specific technical solution adopted is as follows:

[0005] This invention provides a dynamic routing optimization method for satellite networks based on link-state awareness, the method comprising:

[0006] Acquire real-time status information for each link in the satellite network; the real-time status information includes the available bandwidth, total bandwidth, one-way latency, bit error rate, and availability of each link; acquire the set of services to be transmitted and its detailed information; the detailed information includes a set of candidate paths composed of links and the data packet size and importance level of each service to be transmitted;

[0007] Based on the available bandwidth, total bandwidth, and one-way latency of each link, determine the overall transmission capacity of each link; combine the overall transmission capacity with the bit error rate and availability of the link to determine the health of each link; and determine the predicted health of each link based on the trend of the health changes.

[0008] Based on the importance level of the service to be transmitted and the predicted health, determine the routing metric value for each link to each service to be transmitted; based on the routing metric values ​​corresponding to all links in the candidate path, filter to obtain the primary forwarding path and the backup forwarding path;

[0009] Based on the data packet size and importance level of the service to be transmitted, determine the backup activation tendency value for each service to be transmitted; based on the backup activation tendency value, determine the final forwarding path for each service to be transmitted in the primary forwarding path and the backup forwarding path, and use the final forwarding path for data transmission.

[0010] Furthermore, the method for obtaining the overall transmission capability of each link includes:

[0011] For any given link, the ratio of the available bandwidth of that link to the total bandwidth is denoted as the available bandwidth ratio of that link.

[0012] Obtain the inversely proportional normalized result of the one-way delay of the link, and multiply the inversely proportional normalized result by the bandwidth availability ratio of the link, which is recorded as the comprehensive transmission capacity of the link.

[0013] Furthermore, the method for obtaining the health status of each link includes:

[0014] For any link, obtain the product of the inverse proportional result of the link's bit error rate and its availability, and then take the linearly normalized result of the product of the link's overall transmission capacity and denote it as the health of the link.

[0015] Furthermore, the method for obtaining the predicted health status includes:

[0016] For any link, perform a linear fit on the health of the link within a preset recent window, and record the slope of the fitted line as the trend index of the link.

[0017] For any given moment within the preset prediction window, obtain the time distance between that moment and the current moment, obtain the product of that time distance and the trend index of the link, and sum the product with the health of the link at the current moment, which is recorded as the predicted local health of the link at that moment; the minimum value of the predicted local health of the link across all moments within the preset prediction window is recorded as the predicted health of the link.

[0018] Furthermore, the method for obtaining the routing metric value for each link for each service to be transmitted includes:

[0019] For any link and any service to be transmitted, obtain the difference between the preset health threshold and the predicted health of the link, and record the maximum value between the difference and 0 as the health adaptation cost of the link.

[0020] The inverse proportional result of the importance level of the service to be transmitted is recorded as the important adaptation cost of the service to be transmitted.

[0021] The sum of the health adaptation cost of the link and the critical adaptation cost of the service to be transmitted is recorded as the routing metric of the link for the service to be transmitted.

[0022] Furthermore, the method for obtaining the primary forwarding path and the backup forwarding path includes:

[0023] For any link, the average of the routing metrics of all services to be transmitted in the service set to be transmitted on that link is denoted as the overall routing metric of that link.

[0024] For each candidate path, the total path cost is obtained by summing the overall routing metrics of all links in the candidate path.

[0025] The primary forwarding path and backup forwarding paths are determined based on the total path cost of the candidate paths.

[0026] Furthermore, the specific method for determining the primary forwarding path and backup forwarding paths based on the total path cost of candidate paths includes:

[0027] The candidate paths in the candidate path set are sorted in ascending order of total path cost, and the candidate path with the lowest total path cost is determined as the primary forwarding path.

[0028] Among the candidate paths other than the primary forwarding path, the candidate path with the second smallest total cost and not exactly the same as the primary forwarding path is determined as the backup forwarding path.

[0029] Furthermore, the method for obtaining the standby activation tendency value for each service to be transmitted includes:

[0030] For any service to be transmitted, the inverse proportional normalization result of the data packet size of the service to be transmitted is denoted as the backup bearer adaptation index of the service to be transmitted.

[0031] The inverse proportional result of the importance level of the service to be transmitted is recorded as the low importance fit index of the service to be transmitted.

[0032] The product of the backup bearer adaptation index and the low importance adaptation index of the service to be transmitted is recorded as the backup activation tendency value of the service to be transmitted.

[0033] Furthermore, the specific method for determining the final forwarding path for each service to be transmitted in the primary forwarding path and the backup forwarding path based on the backup activation tendency value, and using the final forwarding path for data transmission, includes:

[0034] Based on the backup activation tendency value of each service to be transmitted, construct the traffic splitting characteristic sequence of the services to be transmitted;

[0035] The traffic splitting feature sequences are subjected to binary clustering to obtain a first service cluster and a second service cluster;

[0036] Calculate the mean standby activation tendency value of the services to be transmitted in the first service cluster and the second service cluster respectively;

[0037] The cluster of services with the smallest average backup activation tendency value is designated as the primary path service cluster, and the cluster of services with the largest average backup activation tendency value is designated as the backup path service cluster.

[0038] The primary forwarding path is determined as the final forwarding path for each service to be transmitted in the primary path service cluster;

[0039] The final forwarding path for each service to be transmitted in the backup path service cluster is determined based on the availability of the links in the backup forwarding path.

[0040] A service forwarding table is generated based on the final forwarding path corresponding to each service to be transmitted, and the data packets of each service to be transmitted are forwarded according to the service forwarding table.

[0041] Furthermore, the specific method for determining the final forwarding path of each service to be transmitted in the backup path service cluster based on the availability of links in the backup forwarding path includes:

[0042] The availability of the backup forwarding path is determined based on the availability of the links in the backup forwarding path.

[0043] If the backup forwarding path is available, then the backup forwarding path is determined as the final forwarding path for each service to be transmitted in the backup path service cluster.

[0044] If the backup forwarding path is unavailable, the primary forwarding path will be determined as the final forwarding path for each service to be transmitted in the backup path service cluster.

[0045] The present invention has the following beneficial effects:

[0046] To address the technical challenge of balancing the stability of critical service transmissions with efficient link resource utilization, this invention first obtains the predicted health status of each link. It then comprehensively characterizes the link status using available bandwidth, total bandwidth, one-way latency, bit error rate, and availability. Furthermore, it predicts the link's future state in the short term by considering the trend of health status changes. This allows routing decisions to avoid links with low predicted health or deteriorating trends in advance, rather than solely relying on current link connectivity, thus reducing the impact of link mutations on continuous service transmission. Based on the routing metrics corresponding to all links in the candidate paths, it selects primary and backup forwarding paths, ensuring that path selection simultaneously considers both the predicted link health status and the services to be transmitted. The system prioritizes services based on their importance and determines the primary forwarding path with the best overall quality through total path cost, while retaining alternative backup forwarding paths. This improves the reliability of path selection and the flexibility of route adjustment. Based on the backup activation preference value, the system determines the final forwarding path for each service to be transmitted within the primary and backup forwarding paths. This prioritizes services with higher importance and larger data packets for transmission on the primary forwarding path, while assigning services with lower importance and smaller data packets to backup forwarding paths when available. This differentiated traffic distribution between primary and backup paths avoids all services concentrating on the primary forwarding path, improving the transmission stability of important services and the utilization rate of link resources. Attached Figure Description

[0047] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart illustrating the steps of a link-state-aware dynamic routing optimization method for satellite networks provided in an embodiment of the present invention. Detailed Implementation

[0049] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the link-state-aware dynamic routing optimization method for satellite networks proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0051] The specific scheme of the link-state-aware dynamic routing optimization method for satellite networks provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0052] Example of a link-state-aware dynamic routing optimization method for satellite networks:

[0053] Please see Figure 1 This illustrates a flowchart of a link-state-aware dynamic routing optimization method for satellite networks provided by an embodiment of the present invention. Figure 1 As shown, the method specifically includes the following steps:

[0054] Step S1: Obtain real-time status information of each link in the satellite network; obtain the set of services to be transmitted and its detailed information.

[0055] The purpose of this embodiment is to dynamically determine the final forwarding path of each service to be transmitted based on the real-time transmission status of the links in the satellite network and the service attributes of the services to be transmitted, thereby improving the utilization rate of the satellite network's link resources while ensuring the stable transmission of important services.

[0056] Specifically, the real-time status information of each link in the satellite network is obtained; the link is a communication link between adjacent satellite nodes in the satellite network used for data transmission, and the real-time status information includes the available bandwidth, total bandwidth, one-way delay, bit error rate and availability of each link.

[0057] The available bandwidth is the remaining bandwidth of the link at the current acquisition time that can be used to carry newly added services to be transmitted; the total bandwidth is the maximum transmission bandwidth supported by the link under the current communication configuration; the one-way delay is the time required for a data packet to be transmitted from one satellite node to the other satellite node of the link; the bit error rate is the proportion of data transmission errors of the link at the current acquisition time; and the availability is used to characterize whether the link is in a state that can be used for data forwarding, with an availability value of 1 indicating that the link is available and an availability value of 0 indicating that the link is unavailable.

[0058] In one embodiment, the available bandwidth and total bandwidth of each link are obtained by the link resource management module of the satellite node; the one-way delay of each link is determined based on the transmission and reception times of the probe messages between the satellite nodes at both ends of the link; the bit error rate of each link is determined based on the number of check errors and the total amount of received data during the link transmission process; and the availability of each link is determined based on the link connectivity status.

[0059] The real-time status information of each link is continuously collected according to the preset collection period; the preset collection period is described using 1 second as an example.

[0060] Furthermore, the set of services to be transmitted and its detailed information are obtained. The set of services to be transmitted includes multiple services in the current satellite network that require path allocation; for any service to be transmitted, the service identifier, data packet size, importance level, and a set of candidate paths composed of links are obtained.

[0061] The data packet size represents the amount of data that the service to be transmitted needs to transmit within the current transmission cycle. The importance level represents the degree of requirement of the service to be transmitted for transmission stability and continuity. The higher the importance level, the more likely the service to be transmitted needs to be transmitted through a path with higher stability. In this embodiment, the importance level of emergency services is 0.8, the importance level of ordinary services is 0.5, and the importance level of general services is 0.2. Emergency services include, but are not limited to, control commands and voice, ordinary services include, but are not limited to, video, and general services include, but are not limited to, files. The candidate path set includes multiple candidate paths that can be used to transmit the service to be transmitted, and each candidate path is formed by connecting at least one link in sequence.

[0062] In one embodiment, the candidate path set is obtained as follows: a link topology map is constructed based on the link connection relationship between each satellite node in the satellite network, and unavailable links are removed from the link topology map based on the availability of each link to obtain a transmittable link topology map; based on the transmittable link topology map, multiple transmittable paths capable of carrying the corresponding service to be transmitted are obtained, and the links connected sequentially in each transmittable path are taken as a candidate path, and all the obtained candidate paths are combined into a candidate path set.

[0063] Step S2: Determine the overall transmission capacity of each link based on the available bandwidth, total bandwidth, and one-way latency; combine the overall transmission capacity with the link's bit error rate and availability to determine the health of each link; determine the predicted health of each link based on the trend of the health changes.

[0064] It should be noted that in satellite networks, the suitability of a link for data forwarding cannot be determined solely by its connectivity. Even if a link is connected, low remaining available bandwidth can easily lead to queuing and congestion when carrying new services. High one-way latency will increase end-to-end transmission delays, especially for real-time services such as control commands and voice, directly reducing transmission quality. Therefore, this embodiment first comprehensively evaluates the basic transmission capabilities of the link from the perspectives of bandwidth sufficiency and latency transmission efficiency.

[0065] Specifically, for any given link, the ratio of the available bandwidth of that link to the total bandwidth is denoted as the available bandwidth ratio of that link.

[0066] Obtain the inversely proportional normalized result of the one-way delay of the link, and multiply the inversely proportional normalized result by the bandwidth availability ratio of the link, which is recorded as the comprehensive transmission capacity of the link.

[0067] As an example, the overall transmission capacity of this link is calculated as follows:

[0068]

[0069] In the formula, This refers to the overall transmission capacity of the link; This represents the available bandwidth of the link. This represents the total bandwidth of the link. This represents the one-way delay of the link; It is a linear normalization function, and the normalization object is the one-way delay of all links; This represents the proportion of available bandwidth for that link.

[0070] It should be noted that the overall transmission capability is used to characterize the basic transmission advantage of the link at the current moment. The larger the value, the higher the available bandwidth ratio and the lower the one-way latency of the link, which can carry new services to be transmitted with less transmission waiting cost. The smaller the value, the more likely the link is to have bandwidth constraints and high latency. Even if the link is not completely interrupted, it is not suitable as a priority forwarding link.

[0071] It should be further explained that the overall transmission capability mainly reflects the transmission conditions of the link in terms of bandwidth and latency. However, in satellite network scenarios, link quality is also affected by factors such as space environment interference, link obstruction, and changes in inter-satellite link pointing, which can lead to increased bit error rate or link unavailability. If routing is based solely on bandwidth and latency, services may be assigned to links with high bit error rates or those about to be interrupted, resulting in data retransmission, packet loss, or even service interruption. Therefore, this embodiment further introduces bit error rate and link availability to correct the overall transmission capability of the link, so as to obtain a health level that can simultaneously reflect transmission efficiency, transmission accuracy, and link connectivity.

[0072] Specifically, for any link, the product of the inverse proportional result of the link's bit error rate and its availability is obtained, and the linearly normalized result of the product of the link's overall transmission capacity is recorded as the health of the link.

[0073] As an example, the health of this link is calculated as follows:

[0074]

[0075] In the formula, The health status of this link; This refers to the overall transmission capacity of the link; This represents the bit error rate of the link. For the availability of this link; This is a linear normalization function, and the normalization object is all links. .

[0076] Regarding comprehensive transmission capabilities The calculation is determined by the product of the available bandwidth ratio and the inversely proportional normalized result of the one-way delay, with a value range of [value missing]. Availability To characterize whether a link is in a state where it can be used for data forwarding, a binary quantization method is used: when the link is connected and can be used for data forwarding... When the link is disconnected, faulty, or unavailable, Therefore, the health calculation formula will take into account the overall transmission capacity. As a fundamental factor, it reflects the sufficiency of bandwidth resources and the efficiency of latency transmission; As a correction factor for transmission accuracy, among which Reflecting transmission accuracy, This reflects the link's availability. Even with high bandwidth and low latency, a high bit error rate or unavailability can significantly reduce actual transmission efficiency; finally, the value is mapped to a linear normalization function. The interval ensures that the health of different links is directly comparable, and this product structure effectively reflects the overall availability of the link at the current moment.

[0077] It should be noted that health status is used to characterize the overall availability of the link at the current moment. The larger the value, the better the bandwidth and latency conditions of the link, the lower the bit error rate, and the more usable the link is. The smaller the value, the more likely the link may have problems such as decreased transmission efficiency, increased bit error rate, or link unavailability, and the higher the risk of carrying the service to be transmitted.

[0078] It should be noted that in low-Earth orbit satellite networks, satellite nodes are in high-speed motion, and the latency, bandwidth, bit error rate, and availability of inter-satellite links may continuously change over time. If routing decisions are made solely based on the current health status, the system may still misclassify a link as available when it is rapidly deteriorating, leading to link quality degradation or interruption immediately after service allocation. Therefore, this embodiment estimates the link health status based on recent trends in link health, thereby identifying links that may deteriorate in the near future.

[0079] Specifically, for any link, the health of the link within a preset recent window is fitted with a straight line, and the slope of the fitted line is recorded as the trend index of the link. The preset recent window is obtained by establishing a window with a length of 5 collection cycles, starting from the current time. The straight line fitting adopts the least squares method, which is a well-known technique, and the specific method will not be described here.

[0080] The specific method for obtaining the trend index is as follows: establish a two-dimensional rectangular coordinate system with time as the horizontal axis and health as the vertical axis, obtain a scatter plot of the health corresponding to each time within the preset recent window in the two-dimensional rectangular coordinate system, and use the least squares method to fit a straight line to the scatter plot, thereby using the slope of the fitted straight line as the trend index.

[0081] The physical meaning of the trend index is the average rate of change of link health within a preset recent window. When When the link health status is trending upward, it indicates that the link quality is improving; when When, it indicates that the link health remains stable; when When this occurs, it indicates that the link health is declining and the link quality is deteriorating. The larger the absolute value of health G, the faster the deterioration. Since the range of health G values ​​is... Furthermore, the preset recent window length is fixed at 5 collection cycles, and the value range of the trend index is limited to a reasonable range, thereby accurately quantifying the changing trend of link health.

[0082] Any point within the preset prediction window is designated as the target point. The time distance between the target point and the current point is obtained. The product of the target point distance and the trend index of the link is obtained. The sum of this product and the health of the link at the current point is recorded as the predicted local health of the link at the target point. The minimum value of the predicted local health of the link at all points within the preset prediction window is recorded as the predicted health of the link. The preset prediction window is obtained by establishing a window with a length of 10 seconds starting from the current point.

[0083] As an example, the predicted local health of this link at the target time is calculated as follows:

[0084]

[0085] In the formula, This is the predicted local health of the link at the target time. The health status of this link; This is the trend index for this link; This represents the time distance between the target time and the current time.

[0086] It should be noted that the predicted local health is used to characterize the expected health status of a link at a target time within a preset prediction window. If a link currently has a high health, but its trend index is negative and its rate of decline is rapid, then the minimum predicted local health of that link within the prediction window will decrease; if a link's health is stable or showing an upward trend, then its predicted health can remain at a high level. This method can reduce the impact of link mutations on continuous service transmission.

[0087] Step S3: Determine the routing metric value for each link to each service to be transmitted based on the importance level of the service to be transmitted and the predicted health; filter out the primary forwarding path and the backup forwarding path based on the routing metric values ​​corresponding to all links in the candidate path.

[0088] It should be noted that after obtaining the predicted health of each link, it is also necessary to match the link status with the service attributes of the service to be transmitted. Different services to be transmitted have different sensitivities to link quality. For example, control commands and critical business data with higher importance levels require more stable and low-risk links, while ordinary data with lower importance levels can appropriately bear higher link costs.

[0089] Specifically, for any link and any service to be transmitted, the difference between the preset health threshold and the predicted health of the link is obtained, and the maximum value between the difference and 0 is recorded as the health adaptation cost of the link.

[0090] The inverse proportional result of the importance level of the service to be transmitted is recorded as the important adaptation cost of the service to be transmitted.

[0091] The sum of the health adaptation cost of the link and the critical adaptation cost of the service to be transmitted is recorded as the routing metric of the link for the service to be transmitted.

[0092] As an example, the routing metric for the service to be transmitted on this link is calculated as follows:

[0093]

[0094] In the formula, This is the routing metric for the service to be transmitted on this link; To preset health thresholds; The predicted health of this link; It is a function for maximizing the value; The importance level of the service to be transmitted; The cost of health adaptation for this link; This represents a significant adaptation cost for the service to be transmitted.

[0095] The formula for calculating the routing metric includes a predicted health value, which is the minimum predicted local health value of the link at all times within a preset prediction window. This conservative strategy of taking the minimum value ensures that routing decisions can avoid the risk of link quality deterioration in the near future. A preset health threshold is also included. Used to determine whether the predicted health of the link meets business requirements, with a value range of [value range missing]. Specifically, based on the statistical characteristics of the health distribution of all links under normal satellite network operation (such as mean) and standard deviation ) Determine, for example ,in For a safety factor, a value between 1 and 3 is used, so that... It can dynamically adapt to the network environment. Importance level Characterizes the service's requirement for transmission stability, with a value range of... The higher the value, the higher the demand; in specific embodiments, under emergency services (such as control commands, voice calls)... You can take 0.8 for regular business (such as video). You can take 0.5 for general business operations (such as file transfers). It can be set to 0.2, and in practice, it can be determined through the preset mapping table of the business management system or user-defined configuration.

[0096] It should be noted that the routing metric is used to characterize the adaptation cost when a link carries a corresponding service to be transmitted. The smaller the value, the closer the predicted health of the link is to or exceeds the preset health requirements. At the same time, the higher the importance level of the service to be transmitted, the more the system tends to reserve high-quality link resources for it. The higher the value, the less healthy the predicted link is, or the lower the service's priority demand for high-quality link resources.

[0097] It should be noted that the routing metric of a single link can only reflect the local link's adaptability to the transmission service. However, actual satellite network forwarding usually requires passing through multiple inter-satellite links, so the link-level cost needs to be extended to the path-level cost.

[0098] Specifically, for any link, the average of the routing metrics of all services to be transmitted in the set of services to be transmitted on that link is denoted as the overall routing metric of that link.

[0099] For each candidate path, the total path cost is obtained by summing the overall routing metrics of all links in the candidate path.

[0100] The candidate paths in the candidate path set are sorted in ascending order of total path cost, and the candidate path with the lowest total path cost is determined as the primary forwarding path.

[0101] Among the candidate paths other than the primary forwarding path, the candidate path with the second smallest total cost and not exactly the same as the primary forwarding path is determined as the backup forwarding path.

[0102] If there are multiple candidate paths with the second lowest total cost and not exactly the same as the primary forwarding path, then the backup candidate path with the fewest overlapping links with the primary forwarding path will be determined as the backup forwarding path.

[0103] It should be noted that the total path cost characterizes the overall forwarding cost of a candidate path for the current set of services to be transmitted. A lower total path cost indicates a better overall health of the links traversed by the path, lower service adaptation costs, and greater suitability as the primary bearer path. Therefore, identifying the candidate path with the lowest total path cost as the primary forwarding path ensures that core services are preferentially transmitted via the path with the best overall quality. Identifying the candidate path with the second lowest total path cost, which is not entirely identical to the primary forwarding path, as a backup forwarding path provides a relatively high-quality alternative path. When multiple suboptimal backup candidate paths exist, prioritizing the path with the fewest overlapping links with the primary forwarding path reduces the probability of both primary and backup paths being affected by the same link failure, thus improving the forwarding continuity of the satellite network.

[0104] Step S4: Determine the backup activation tendency value for each service to be transmitted based on the data packet size and importance level of the service to be transmitted; determine the final forwarding path for each service to be transmitted in the primary forwarding path and the backup forwarding path based on the backup activation tendency value, and use the final forwarding path for data transmission.

[0105] It should be noted that after determining the primary and backup forwarding paths, it is not suitable to fix all services to be transmitted on the primary forwarding path, nor is it suitable to randomly assign services to backup forwarding paths. If all services use the primary forwarding path, it can easily lead to continuous occupation of primary path link resources, reducing network scheduling flexibility. If services with high importance or large data packet size are diverted to backup forwarding paths, the quality of backup paths may be inferior to that of the primary path, or additional instability may arise during path switching, affecting the continuous transmission of critical services or large data packet services. Therefore, this embodiment calculates a backup activation tendency value based on the data packet size and importance level of the services to be transmitted, making services with high importance and large data packet size more likely to remain on the primary forwarding path, and making services with low importance and small data packet size more suitable as backup path diversion targets.

[0106] Specifically, for any service to be transmitted, the inverse proportional normalization result of the data packet size of the service to be transmitted is recorded as the backup bearer adaptation index of the service to be transmitted.

[0107] The inverse proportional result of the importance level of the service to be transmitted is recorded as the low importance fit index of the service to be transmitted.

[0108] The product of the backup bearer adaptation index and the low importance adaptation index of the service to be transmitted is recorded as the backup activation tendency value of the service to be transmitted.

[0109] As an example, the backup activation tendency value for this service to be transmitted is calculated as follows:

[0110]

[0111] In the formula, The backup enable value for the service to be transmitted; The size of the data packet for the service to be transmitted; The importance level of the service to be transmitted; This is a linear normalization function, and the normalization object is the size of the data packets of all services to be transmitted.

[0112] The formula for calculating the standby activation tendency value includes the standby capacity adaptation index. This index reflects the impact of packet size on the adaptation of allocated paths. Smaller packets consume less bandwidth on backup paths, making it more suitable for traffic offloading; larger packets approach 0, avoiding load on backup paths. Low Importance Adaptation Index This index reflects the impact of importance level on the suitability of backup paths. The lower the importance, the closer the index is to 1, indicating lower requirements for transmission stability and suitability for traffic offloading; the higher the importance, the closer it is to 0, and it should be prioritized for the primary path. A product structure is used instead of a summation structure because if the suitability index of any dimension approaches 0 (i.e., that dimension is completely unsuitable for backup paths), the overall backup activation tendency value will inevitably approach 0. For example, even if the data packet is very small but extremely important, it should not be offloaded to a backup path. This product structure effectively achieves traffic offloading assessment across two dimensions: data packet size and importance level.

[0113] It should be noted that the standby activation tendency value is used to characterize the appropriateness of assigning the service to be transmitted to the standby forwarding path. The larger the value, the smaller the data packet of the service to be transmitted and the lower the importance level. Even if it is transmitted through the backup forwarding path, the impact on the continuity of core services and the bandwidth resources of the backup path is relatively small. Therefore, it is more suitable as a backup path service. The smaller the value, the larger or more important the data packet of the service to be transmitted is, and the stronger its dependence on stable links and continuous transmission conditions. It should be reserved in the main forwarding path first.

[0114] It should be noted that since multiple services may coexist in the set of services to be transmitted, applying a fixed threshold to each service individually could significantly affect the traffic splitting results due to the threshold setting, and would make it difficult to adaptively divide the primary and backup path services based on the overall distribution of the current service set. Therefore, this embodiment uses the backup activation tendency value of each service to be transmitted as the traffic splitting feature, performing binary clustering on the services to be transmitted, thus dividing the service set into two categories based on its own traffic splitting tendency. Services with a lower backup activation tendency value typically correspond to services with larger data packets or higher importance levels and should be retained in the primary forwarding path; services with a higher backup activation tendency value typically correspond to services with smaller data packets and lower importance levels, and can be switched to the backup forwarding path when it becomes available. This ensures stable transmission of core services while improving the utilization rate of backup path resources.

[0115] Specifically, based on the backup activation tendency value of each service to be transmitted, a traffic splitting characteristic sequence of the services to be transmitted is constructed;

[0116] The traffic diversion feature sequences are subjected to binary clustering to obtain a first business cluster and a second business cluster. The clustering method used is K-means clustering with K=2. K-means clustering is a well-known technique, and the specific method will not be described here.

[0117] Calculate the mean standby activation tendency value of the services to be transmitted in the first service cluster and the second service cluster respectively;

[0118] The cluster of services with the smallest average standby activation tendency value is designated as the primary path service cluster, and the cluster of services with the largest average standby activation tendency value is designated as the standby path service cluster.

[0119] The primary forwarding path is determined as the final forwarding path for each service to be transmitted within the primary path service cluster.

[0120] Determine whether the backup forwarding path is available based on the availability of the links in the backup forwarding path;

[0121] If the backup forwarding path is available, then the backup forwarding path will be determined as the final forwarding path for each service to be transmitted in the backup path service cluster.

[0122] If the backup forwarding path is unavailable, the primary forwarding path will be determined as the final forwarding path for each service to be transmitted in the backup path service cluster.

[0123] A service forwarding table is generated based on the final forwarding path corresponding to each service to be transmitted, and the data packets of each service to be transmitted are forwarded according to the service forwarding table.

[0124] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0125] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A dynamic routing optimization method for satellite networks based on link state awareness, characterized in that, The method includes: Acquire real-time status information for each link in the satellite network; the real-time status information includes the available bandwidth, total bandwidth, one-way latency, bit error rate, and availability of each link; acquire the set of services to be transmitted and its detailed information; the detailed information includes a set of candidate paths composed of links and the data packet size and importance level of each service to be transmitted; Based on the available bandwidth, total bandwidth, and one-way latency of each link, determine the overall transmission capacity of each link; combine the overall transmission capacity with the bit error rate and availability of the link to determine the health of each link; and determine the predicted health of each link based on the trend of the health changes. Based on the importance level of the service to be transmitted and the predicted health, determine the routing metric value for each link to each service to be transmitted; based on the routing metric values ​​corresponding to all links in the candidate path, filter to obtain the primary forwarding path and the backup forwarding path; Based on the data packet size and importance level of the service to be transmitted, determine the backup activation tendency value for each service to be transmitted; based on the backup activation tendency value, determine the final forwarding path for each service to be transmitted from the primary forwarding path and the backup forwarding path, and use the final forwarding path for data transmission. The method for obtaining the standby activation tendency value for each service to be transmitted includes: For any service to be transmitted, the inverse proportional normalization result of the data packet size of the service to be transmitted is denoted as the backup bearer adaptation index of the service to be transmitted. The inverse proportional result of the importance level of the service to be transmitted is recorded as the low importance fit index of the service to be transmitted. The product of the backup bearer adaptation index and the low importance adaptation index of the service to be transmitted is recorded as the backup activation tendency value of the service to be transmitted. The specific method for determining the final forwarding path for each service to be transmitted in the primary forwarding path and the backup forwarding path based on the backup activation tendency value, and using the final forwarding path for data transmission, includes: Based on the backup activation tendency value of each service to be transmitted, construct the traffic splitting characteristic sequence of the services to be transmitted; The traffic splitting feature sequences are subjected to binary clustering to obtain a first service cluster and a second service cluster; Calculate the mean standby activation tendency value of the services to be transmitted in the first service cluster and the second service cluster respectively; The cluster of services with the smallest average backup activation tendency value is designated as the primary path service cluster, and the cluster of services with the largest average backup activation tendency value is designated as the backup path service cluster. The primary forwarding path is determined as the final forwarding path for each service to be transmitted in the primary path service cluster; The final forwarding path for each service to be transmitted in the backup path service cluster is determined based on the availability of the links in the backup forwarding path. A service forwarding table is generated based on the final forwarding path corresponding to each service to be transmitted, and the data packets of each service to be transmitted are forwarded according to the service forwarding table.

2. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The method for obtaining the overall transmission capability of each link includes: For any given link, the ratio of the available bandwidth of that link to the total bandwidth is denoted as the available bandwidth ratio of that link. Obtain the inversely proportional normalized result of the one-way delay of the link, and multiply the inversely proportional normalized result by the bandwidth availability ratio of the link, which is recorded as the comprehensive transmission capacity of the link.

3. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The method for obtaining the health status of each link includes: For any link, obtain the product of the inverse proportional result of the link's bit error rate and its availability, and then take the linearly normalized result of the product of the link's overall transmission capacity and denote it as the health of the link.

4. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The method for obtaining the predicted health status includes: For any link, perform a linear fit on the health of the link within a preset recent window, and record the slope of the fitted line as the trend index of the link. For any given moment within the preset prediction window, obtain the time distance between that moment and the current moment, obtain the product of that time distance and the trend index of the link, and sum the product with the health of the link at the current moment, which is recorded as the predicted local health of the link at that moment; the minimum value of the predicted local health of the link across all moments within the preset prediction window is recorded as the predicted health of the link.

5. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The method for obtaining the routing metric value for each link for each service to be transmitted includes: For any link and any service to be transmitted, obtain the difference between the preset health threshold and the predicted health of the link, and record the maximum value between the difference and 0 as the health adaptation cost of the link. The inverse proportional result of the importance level of the service to be transmitted is recorded as the important adaptation cost of the service to be transmitted. The sum of the health adaptation cost of the link and the critical adaptation cost of the service to be transmitted is recorded as the routing metric of the link for the service to be transmitted.

6. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The methods for obtaining the primary forwarding path and the backup forwarding path include: For any link, the average of the routing metrics of all services to be transmitted in the service set to be transmitted on that link is denoted as the overall routing metric of that link. For each candidate path, the total path cost is obtained by summing the overall routing metrics of all links in the candidate path. The primary forwarding path and backup forwarding paths are determined based on the total path cost of the candidate paths.

7. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 6, characterized in that, The specific method for determining the primary forwarding path and backup forwarding paths based on the total path cost of candidate paths includes: The candidate paths in the candidate path set are sorted in ascending order of total path cost, and the candidate path with the lowest total path cost is determined as the primary forwarding path. Among the candidate paths other than the primary forwarding path, the candidate path with the second smallest total cost and not exactly the same as the primary forwarding path is determined as the backup forwarding path.

8. The dynamic routing optimization method for satellite networks based on link state awareness according to claim 1, characterized in that, The specific method for determining the final forwarding path of each service to be transmitted in the backup path service cluster based on the availability of links in the backup forwarding path is as follows: The availability of the backup forwarding path is determined based on the availability of the links in the backup forwarding path. If the backup forwarding path is available, then the backup forwarding path is determined as the final forwarding path for each service to be transmitted in the backup path service cluster. If the backup forwarding path is unavailable, the primary forwarding path will be determined as the final forwarding path for each service to be transmitted in the backup path service cluster.

Images