An inter-satellite link load balancing method, system, device, and storage medium

Abstract

The application discloses an inter-satellite link load balancing method, system and device and a storage medium, relates to the field of satellite communication, and aims at the congestion problem of a user plane network of a core network. In the application, an SMF network element receives inter-satellite link state information from each UPF network element; under the limitation of inter-satellite link delay and bandwidth, a routing path from a source satellite node to a target satellite node is calculated according to the inter-satellite link state information; and the routing path decision is distributed to the UPF network element of the corresponding satellite node. The application can realize load balancing of the inter-satellite link, avoid the congestion of the user plane network of the satellite-borne core network, and improve the service quality of the satellite-borne core network.

Description

Technical Field

[0001] This invention relates to the field of satellite communication technology, and in particular to an inter-satellite link load balancing method, system, apparatus, and computer-readable storage medium. Background Technology

[0002] Satellite internet, as a necessary extension of terrestrial networks, can effectively solve the problem of insufficient coverage of terrestrial networks in remote and maritime areas. Compared to geostationary orbit satellite networks, low-Earth orbit (LEO) satellite networks offer lower communication latency and signal transmission loss.

[0003] Due to the uneven distribution of global traffic and the high dynamics of satellite network topology, the load on various nodes in satellite networks is uneven. In traditional networks, the Shortest Path First (SPF) algorithm is typically used for traffic scheduling, with typical algorithms including Dijkstra's and Floyd's algorithms. The principle of this algorithm is to obtain the shortest route with the fewest nodes. The Equal-Cost Multi-Path (ECMP) algorithm is a commonly used load balancing algorithm in data center network traffic scheduling. Its principle is to distribute flows with different QoS (Quality of Service) values ​​to multiple paths with equal path costs by using hash value calculation, round-robin, or weighted strategies to achieve load balancing.

[0004] The SPF and ECMP algorithms mentioned above are both static traffic scheduling algorithms, which can easily lead to network congestion. Summary of the Invention

[0005] The purpose of this invention is to provide an inter-satellite link load balancing method, system, device, and storage medium to address all or part of the aforementioned problems, thereby resolving the core network user plane network congestion problem.

[0006] The technical solution adopted in this invention is as follows:

[0007] An inter-satellite link load balancing method is applied to a Session Management Function (SMF) network element in a satellite network. The satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node. Each satellite node deploys a User Plane Function (UPF) network element. The method includes:

[0008] Receive inter-satellite link status information from each UPF network element;

[0009] Under the constraints of inter-satellite link latency and bandwidth, based on the inter-satellite link status information, the routing path from the source satellite node to the target satellite node is calculated by optimizing the path from the source satellite node to the target satellite node. In the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of the preceding satellite node to the following satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the preceding satellite node and the following satellite node. The expected heuristic factor is obtained by weighted summation of the bandwidth utilization and transmission latency tolerance of the inter-satellite link.

[0010] The routing path decision is distributed to the UPF network element of the corresponding satellite node.

[0011] This application also provides an inter-satellite link load balancing system applied to an SMF network element in a satellite network. The satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, with a UPF network element deployed on each satellite node. The system includes:

[0012] Measurement task distribution interface;

[0013] Measurement report receiving interface;

[0014] The routing analysis module distributes measurement tasks to each UPF network element through the measurement task distribution interface, enabling each UPF network element to collect inter-satellite link status information; it receives the inter-satellite link status information from each UPF network element through the measurement report receiving interface; under the constraints of inter-satellite link latency and bandwidth, it calculates the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information by optimizing the path from the source satellite node to the target satellite node; it distributes the routing path decision to the UPF network element corresponding to the satellite node; during the optimization of the path from the source satellite node to the target satellite node, the transition probability of the previous satellite node to the next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization and transmission latency tolerance of the inter-satellite link.

[0015] This application also provides an inter-satellite link load balancing device applied to a satellite network, the satellite network including: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, each satellite node deploying a User Plane Function (UPF) network element; the device includes a processor and a storage medium, the storage medium storing computer instructions, and when the processor runs the computer instructions, it executes the above-described inter-satellite link load balancing method.

[0016] This application also provides a computer-readable storage medium for use in an SMF network element of a satellite network, the satellite network comprising: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, wherein a User Plane Function (UPF) network element is deployed on each satellite node; the storage medium stores a computer program, which, when executed by a processor, performs the aforementioned inter-satellite link load balancing method.

[0017] On the other hand, this application provides an inter-satellite link load balancing method applied to UPF network elements in a satellite network, wherein the satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, with UPF network elements deployed on each satellite node; the method includes:

[0018] Receive measurement tasks sent by SMF network elements;

[0019] In response to the measurement task, inter-satellite link status information is collected and reported to the SMF network element. This allows the SMF network element to calculate the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information, under the constraints of inter-satellite link latency and bandwidth. During the path optimization process, the transition probability from one satellite node to the next is calculated based on the expected heuristic factor of the inter-satellite link between the former and latter satellite nodes. The expected heuristic factor is obtained by weighted summation of the bandwidth utilization and transmission latency tolerance of the inter-satellite link.

[0020] Receive and execute routing path decisions issued by the SMF network element.

[0021] This application also provides an inter-satellite link load balancing system applied to a UPF network element in a satellite network. The satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, with a UPF network element deployed on each satellite node. The system includes:

[0022] Measurement task receiving interface;

[0023] Measurement report submission interface;

[0024] The measurement task execution module receives measurement tasks sent by the SMF network element; in response to the measurement task, it collects inter-satellite link status information and reports it to the SMF network element, so that the SMF network element, under the constraints of inter-satellite link latency and bandwidth, calculates the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information by optimizing the path from the source satellite node to the target satellite node; in the process of optimizing the path from the source satellite node to the target satellite node, the transition probability of the previous satellite node to the next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization and transmission latency tolerance of the inter-satellite link;

[0025] The policy execution module receives and executes the routing path decisions issued by the SMF network element.

[0026] This application also provides an inter-satellite link load balancing device applied to a satellite network, the satellite network including: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, each satellite node deploying a User Plane Function (UPF) network element; the device includes a processor and a storage medium, the storage medium storing computer instructions, and when the processor runs the computer instructions, it executes the above-described inter-satellite link load balancing method.

[0027] This application also provides a computer-readable storage medium for use in a UPF network element of a satellite network, the satellite network including: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, each satellite node having a user plane function UPF network element deployed on it; the storage medium stores a computer program, which, when executed by a processor, performs the above-described inter-satellite link load balancing method.

[0028] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0029] The inter-satellite link load balancing method of this application optimizes path routing based on inter-satellite link state information, fully considering network transmission latency and bandwidth. Thus, for each PDU (Protocol Data Unit) session request, the best routing path can be planned for it under load balancing, achieving the effect of overall load balancing and improving the overall network service quality. Attached Figure Description

[0030] The present invention will be described by way of example and with reference to the accompanying drawings, wherein:

[0031] Figure 1 This is a diagram of the spaceborne core network architecture.

[0032] Figure 2 This is a flowchart of a PDU session establishment in one embodiment.

[0033] Figure 3 This is a flowchart of an inter-satellite link load balancing method applied to an SMF network element embodiment.

[0034] Figure 4 This is a satellite network layout diagram.

[0035] Figure 5 This is a network topology diagram of the inter-satellite link load balancing method in an embodiment.

[0036] Figure 6 This is a flowchart illustrating the implementation of route path optimization in one embodiment.

[0037] Figure 7 This is an iterative flowchart of route path optimization in one embodiment.

[0038] Figure 8 This is a data flow diagram of an embodiment of the inter-satellite link load balancing method.

[0039] Figure 9 This is a flowchart of an inter-satellite link load balancing method applied to a UPF network element embodiment.

[0040] Figure 10 This is a construction diagram of an SMF (Session Management Function) network element and a UPF (User Plane Function) network element in one embodiment.

[0041] Figure 11 This is a measurement network topology diagram in a test inter-satellite link state information sensing embodiment.

[0042] Figure 12 Is Figure 5 Comparison of end-to-end average transmission delay under different algorithms during implementation.

[0043] Figure 13 Is Figure 5 A comparison chart of average throughput under different algorithms during implementation.

[0044] Figure 14 Is Figure 5 A comparison chart of average packet loss rates under different algorithms during implementation. Detailed Implementation

[0045] All features disclosed in this specification, or all steps in all disclosed methods or processes, may be combined in any way, except for mutually exclusive features and / or steps.

[0046] Any feature disclosed in this specification (including any appended claims and abstract) may be replaced by other equivalent or similar features, unless specifically stated otherwise. That is, unless specifically stated otherwise, each feature is merely one example of a series of equivalent or similar features.

[0047] To address the problem that traditional satellite routing path planning algorithms can easily cause congestion in the core network user plane, this application provides an inter-satellite link load balancing method, system, device, and storage medium, aiming to achieve overall load balancing and improve the overall network service quality.

[0048] The 5G core network adopts a service-oriented design concept, realizing the separation of the control plane and the user plane and the modular design of network functions. It can flexibly choose the deployment method according to business needs, thus it has the ability to provide services for different business scenarios.

[0049] Research has shown that deploying core network functions on satellites via onboard virtual machines or containerized deployment enables satellite networks to possess core network capabilities. Onboard core network deployment schemes can be broadly categorized into three types: full onboard deployment, control plane network elements onboard, and user plane (UPF) network elements onboard. Onboarding user plane network elements significantly reduces inter-satellite user data transmission latency and allows for the design of appropriate routing strategies to reduce satellite-to-ground data transmission latency and balance inter-satellite link load. To balance inter-satellite link load, both control plane and user plane network elements can be deployed on satellites; alternatively, considering the limitations of onboard computing and storage resources, the onboard core network elements can be somewhat curated, employing only UPF network elements onboard, while deploying the core network control plane network elements on the ground. In this application, one UPF network element is deployed on each satellite; therefore, through reasonable routing path planning, the UPF network element can effectively reduce data transmission latency and balance link load. For ease of explanation, this application only describes the implementation scheme of UPF network element satellite deployment. If SMF network element is also deployed on satellite, the same principle applies.

[0050] like Figure 1 The diagram shows the architecture of the spaceborne core network. (Based on the attached...) Figure 1 As shown, the core network control plane includes the following core network elements:

[0051] The AMF (Access and Mobility Management Function) network element is primarily responsible for authenticating and managing the mobility of the UE. The AMF authenticates the UE based on the subscription information in the authentication service function. Simultaneously, it establishes a connection with the base station through the N2 interface to perform context management, reachability management, and mobility restriction management for the UE.

[0052] The SMF network element is responsible for functions such as session management, UE IP address allocation, and routing path management. The SMF network element establishes a connection with the UPF network element through the N4 interface, and when the UE initiates a PDU session request, it sends corresponding rules to the UPF network element to indicate the data routing path of the session and the data packet processing logic, ensuring that the user's data is forwarded correctly and guaranteeing the corresponding quality of service.

[0053] For a user to access the Data Network (DN), the User Equipment (UE) initiates a PDU session request to the core network. The SMF (Service Provider Function) then passively performs path routing planning for this PDU session request. The UPF (User Provider Function) acts as a gateway between the UE and the DN, providing data access functionality between them. It receives control and management information from the SMF via the N4 interface and processes user plane service flows based on the routing path decisions issued by the SMF, ensuring the quality of service for these flows. When a PDU session is established, the Radio Access Network (RAN) and the UPF on the session path both establish corresponding contexts. For the UPF, the established context is sent via the N4 interface using the Packet Forwarding Control Protocol (PFCP). The PDU session establishment process is as follows: Figure 2 As shown, the UE indirectly sends a PDU session request to the SMF network element through the AMF. After the SMF network element completes operations such as IP allocation and session routing path decision, it issues a path routing decision to the corresponding UPF network element (all UPF network elements) through the N4 interface. This routing path decision includes PDR (Packet Detection Rule), URR (Usage Reporting Rule), FAR (Forwarding Action Rule), BAR (Buffering Action Rule), QER (QoS Enforcement Rule), usage reporting rule, etc., which are used to instruct the UPF network element how to handle the UE's uplink and downlink data packets.

[0054] When core network control plane elements process each UE's session establishment request, they allocate a routing path for the corresponding data flow. However, as the number of users increases, the number of established PDU sessions also gradually increases, increasing link load. Established PDU sessions may fail to maintain normal operation, resulting in a sharp increase in transmission latency and a decrease in throughput. This is because the total traffic of established PDU sessions exceeds the link's load capacity, leading to network congestion and a decline in session service quality. According to traditional path routing planning methods, if the source and destination satellite nodes are the same, the core network will select the same routing path for the UE. This strategy does not consider link load capacity. When the link load is too high, still selecting that link to establish a PDU session and transmit data will cause network congestion and affect the service quality of other PDU sessions currently in use.

[0055] To address the aforementioned issues, this application proposes an inter-satellite link load balancing method, applied to the Session Management Function (SMF) network element in a satellite network. This satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node. For example... Figure 3 As shown, this load balancing method includes the following process:

[0056] S1. Receive inter-satellite link status information from each UPF network element.

[0057] like Figure 4 The diagram shows the satellite network layout, illustrating the static topology of each satellite node in the constellation using a star-like constellation model. This embodiment uses this two-dimensional model as the satellite network layout topology for illustrative purposes.

[0058] like Figure 4 As shown, the orbital numbers of constellations are denoted as... The number of satellites in any orbit is denoted as The total number of satellites is then expressed as . Indicates satellite n The logical location of the satellite, its orbital number is i Phase number is j The set of all satellites within this constellation can be represented as The subscript here i Indicates satellite node i edge set ,side Indicates satellite node i and satellite nodes j Inter-satellite link (ISL) between them, subscript here i andj Representing satellite nodes i and satellite nodes j .

[0059] For example, such as Figure 5 The diagram shown is a constellation consisting of 9 satellite nodes and 12 inter-satellite links. . Figure 5 In this problem, UPF1 is located on the source satellite node, and UPF9 is located on the target satellite node. The routing path planning problem is to find the inter-satellite links between UPF1 and UPF9. Traditional methods usually use the principle of minimizing the number of satellite nodes or the shortest transmission delay, but they do not consider the load of each inter-satellite link. When the number of PDU sessions surges, some satellite nodes often become overloaded, causing network congestion. The solution provided in this application incorporates inter-satellite link status information into the routing path planning, thereby balancing the load.

[0060] In some optional implementations, the method for receiving inter-satellite link status information from each UPF network element includes:

[0061] The (SMF network element) generates measurement tasks and sends them to the UPF network elements deployed on each satellite, so that each UPF network element responds to the measurement tasks, collects inter-satellite link status information, and feeds it back to the SMF network element.

[0062] Specifically, from the perspective of SMF network elements, the process of performing inter-satellite link state information measurement tasks includes:

[0063] S11. Generate measurement tasks and distribute them to each UPF network element.

[0064] SMF network elements generate and publish measurement tasks, which are published to all UPF network elements.

[0065] S12. Receive the measurement results of the inter-satellite link status information.

[0066] After the measurement task is issued, measurement reports carrying inter-satellite link status information are received from each UPF network element.

[0067] Furthermore, obtaining inter-satellite link status information is done before making routing decisions, and the reception time of PDU session requests is uncertain. When optimizing routing, it is preferable to obtain the latest inter-satellite link status information. Therefore, as an optional implementation, SMF network elements periodically initiate measurement tasks to periodically trigger each UPF network element to collect and report inter-satellite link status information.

[0068] S2. Under the constraints of inter-satellite link latency and bandwidth, calculate the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information.

[0069] The vias take into account the latency and bandwidth limitations of inter-satellite links for routing path planning, and achieve load balancing of inter-satellite links for each PDU session request, thereby avoiding network congestion.

[0070] As an optional implementation, under the constraints of inter-satellite link latency and bandwidth, the pheromone concentration of each inter-satellite link is initialized according to the inter-satellite link state information of each group, and the ant colony algorithm is used to optimize the path from the source satellite node to the target satellite node until convergence, so as to obtain the required path route.

[0071] The inter-satellite link latency limit is defined as the maximum transmission latency of all inter-satellite links. The inter-satellite link bandwidth is limited to the maximum available bandwidth of the inter-satellite link, based on any two adjacent satellite nodes. i and satellite nodes j For example, inter-satellite links Maximum available bandwidth is defined as Each set of inter-satellite link status information includes the measured transmission bandwidth. and transmission delay ,in and They represent the satellite nodes measured at time t. i and satellite nodes j inter-satellite links The transmission bandwidth and transmission latency, among which, By statistically analyzing network cards over a period of time T Internal satellite nodes i To satellite node j Number of bytes sent (i.e., transmission rate) is obtained by conversion (i.e., differentiation with respect to time); By statistically analyzing the network interface card from the satellite node i To satellite node j The time when the data frame was completely sent is obtained.

[0072] In this application, the ant colony algorithm is used to iteratively optimize the path from the source satellite node to the target satellite node, and the path at convergence is taken as the final optimal routing path.

[0073] like Figure 6 As shown, the optimization process of the ant colony algorithm includes:

[0074] S21. Initialize pheromone concentration.

[0075] The initial value of pheromone concentration directly determines the convergence speed of the algorithm. In this application, the primary objective is to achieve load balancing. Therefore, inter-satellite link state information is used as the input parameter of the ant colony algorithm. Under the constraints of inter-satellite link latency and bandwidth, the pheromone concentration of each inter-satellite link is initialized according to the inter-satellite link state information to accelerate the convergence speed of the ant colony algorithm, thereby improving the routing path decision speed.

[0076] As an optional implementation, the pheromone concentration of each inter-satellite link is initialized based on the remaining bandwidth utilization of that link. In this implementation, the initial value of the pheromone concentration is set as follows:

[0077] Formula (1): ;

[0078] In the formula, Indicates inter-satellite link The initial value of pheromone concentration.

[0079] S22, Ant Transfer.

[0080] Since the transmission costs from the UE to the source satellite node and from the destination satellite node to the DN are fixed, optimizing the routing path from the UE to the DN only requires considering the path between the source satellite node and the destination satellite node. The source and destination satellite nodes can be determined based on the UE's PDU session request.

[0081] Each ant represents a path from the source satellite node to the destination satellite node. Each ant maintains a taboo list during iteration to avoid forming a loop. The path passes through multiple inter-satellite links. When an ant reaches a satellite node (other satellite nodes between the source and destination satellite nodes are collectively referred to as intermediate satellite nodes), the ant calculates the transition probability to the next satellite node and selects the satellite node with the highest transition probability as the next hop satellite node.

[0082] As an optional implementation, in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of the preceding satellite node to the following satellite node is based on the expected heuristic factor of the inter-satellite link between the preceding and following satellite nodes. and pheromone content The calculated heuristic factor is... It is obtained by weighted summation of the bandwidth utilization and transmission delay tolerance of the inter-satellite link.

[0083] Specifically, the transition probability is calculated as follows:

[0084] Formula (2): ;

[0085] in, Ants In time t By satellite node i Transfer to satellite node j The transition probability; Indicates being in a satellite node i ants The set of all reachable nodes; Indicates time t Inter-satellite links The pheromone content on the surface Indicates time t Inter-satellite links The pheromone content on the body; for of Power of 1 for of Power; Indicates time t Inter-satellite links The expected inspirational information on the above Indicates time t Inter-satellite links The expected inspirational information; for of Power of 1 for of Power; Refers to a set Any reachable node in; These represent the information heuristic factor and the equilibrium heuristic factor, respectively, with their values ​​set in advance.

[0086] pass The calculation yielded:

[0087] Formula (3): ;

[0088] in, It is obtained by weighted summation of inter-satellite link bandwidth utilization and transmission delay tolerance:

[0089] Formula (4): ;

[0090] In the formula, This represents the weighting factor, whose value is pre-set based on the importance of transmission delay and transmission bandwidth. In the formula, This means inter-satellite links. bandwidth utilization This means inter-satellite links. The tolerance for transmission delay.

[0091] During the ant migration process, before calculating the next-hop satellite node using a roulette wheel approach, it's first determined whether all adjacent satellite nodes are in the taboo list. If so, the path search fails. Otherwise, the next-hop satellite node is calculated based on the migration probability, and it's determined whether the next satellite node is the destination satellite node. If so, the search is successful. Figure 7 As shown.

[0092] S23, Pheromone concentration update.

[0093] As an optional implementation, in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, candidate paths are selected, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor. The pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the pheromone increment of each inter-satellite link.

[0094] After each iteration, the optimal k (pre-defined) paths are selected using the Top-k algorithm and added to the candidate path set P, based on the pheromone evaporation factor. (Pre-set) The pheromone concentration of each inter-satellite link is volatilized separately, and the calculation method is as follows:

[0095] Formula (5): ;

[0096] In the formula, Indicates inter-satellite link The concentration of pheromones after evaporation, of which Indicates time t The next moment.

[0097] In addition, to balance global search capability and algorithm convergence speed, a pheromone evaporation factor is set. With iteration rounds m They are inversely correlated. That is, in the early stages of iteration, Larger values ​​can enhance global search capabilities; in later iterations, A smaller value can improve the convergence speed and accelerate convergence to the optimal routing path. Specifically, in one optional implementation, The value of is designed as follows:

[0098] Formula (6): ;

[0099] In the formula, Indicates the maximum number of iterations of the algorithm; Indicates the maximum pheromone evaporation factor. This represents the minimum pheromone evaporation factor.

[0100] After the pheromone concentration of each inter-satellite link is evaporated, the inter-satellite link pheromone concentration of each candidate path involved in the candidate path set P is updated again:

[0101] Formula (7): ;

[0102] In the formula, Inter-satellite links after this iteration The pheromone increment on the inter-satellite link, which is generated by each ant in this iteration. It is obtained by summing the concentrations of pheromones released. It is represented as:

[0103] Formula (8): ;

[0104] In the formula, Indicates the first Only one ant (i.e., the first) n (path) in this iteration in inter-satellite links The concentration of pheromones released is calculated as follows:

[0105] Formula (9): ;

[0106] In the formula, This represents the increase in pheromone concentration value after each iteration of the ant's search, which is a set constant. This represents the total length of the path traveled by the nth ant in one iteration.

[0107] In formula (7), The calculation method is as follows:

[0108] Formula (10): ;

[0109] In the formula, Indicated by satellite node i and satellite nodes j Inter-satellite links with endpoints, The pheromone increment factor for candidate paths is represented, and its value is set in advance.

[0110] This clearly and completely explains the process executed in each iteration of the ant colony optimization algorithm for optimizing the path between the source and destination satellite nodes. By repeating this process and iterating until convergence (the candidate path no longer changes, or the maximum number of iterations M is reached), the optimal route is obtained. This optimal route can balance the link load of the satellite network, avoid link congestion, and improve the overall network service quality.

[0111] S3. Distribute the routing path decision to the UPF network element of the corresponding satellite node.

[0112] like Figure 8 The diagram illustrates the data flow of an inter-satellite link load balancing method in one embodiment. According to the diagram, the SMF network element cyclically executes the process of issuing measurement tasks and receiving measurement reports. Once it receives a PDU session request from the UE, it obtains the latest inter-satellite link status information from the latest measurement report to optimize the routing path. Then, the optimized routing path decision is distributed to the corresponding UPF network elements in the form of PDU session parameter configuration information (i.e., the various rules mentioned above). Based on the PDU session parameter configuration results fed back by each UPF network element, the PDU session establishment result is returned to the UE.

[0113] The above inter-satellite link load balancing methods are applied to SMF network elements. In this application embodiment, an inter-satellite link load balancing method applied to UPF network elements under the same satellite network structure is also described. For example... Figure 9 The method includes:

[0114] S4. Receive measurement tasks sent by SMF network elements.

[0115] As mentioned in the previous embodiments, in some embodiments, the SMF network element will periodically issue measurement tasks, and the UPF network element will periodically receive the measurement tasks sent by the SMF network element.

[0116] S5. In response to the measurement task, the inter-satellite link status information is collected and reported to the SMF network element. This allows the SMF network element to calculate the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information, under the constraints of inter-satellite link latency and bandwidth. The method by which the SMF network element calculates the routing path from the source satellite node to the target satellite node is exactly the same as the method in step S2 of the inter-satellite link load balancing method embodiment applied to the SMF network element described above, and therefore will not be repeated.

[0117] UPF network elements receive measurement tasks and, in response to these tasks, measure inter-satellite link status information such as real-time inter-satellite transmission delay and transmission rate.

[0118] As an optional implementation method, the process of UPF network elements collecting inter-satellite link status information includes:

[0119] S51. The measurement task is parsed into a sub-task of measuring the inter-satellite link status information between neighboring satellite nodes.

[0120] Each satellite occupies a different logical position, thus requiring different inter-satellite link status information for measurement. Therefore, the measurement task is first resolved into a sub-task of measuring inter-satellite link status information with neighboring satellite nodes. For example, for... Figure 5In the UPF1 network element, its inter-satellite link status information sensing subsystem needs to specifically parse the measurement task into inter-satellite link status information with the UPF2 network element and inter-satellite link status information with the UPF4 network element.

[0121] S52. Measure the status information of each inter-satellite link according to each measurement subtask.

[0122] Based on the analyzed measurement subtasks, the corresponding inter-satellite link status information is measured. The inter-satellite link status information includes the transmission delay and transmission bandwidth of the inter-satellite link, with the transmission bandwidth characterized by the number of bytes sent by the port within a certain time period.

[0123] As an optional implementation, the UPF network element employs active measurement technology to measure inter-satellite link status information. Specifically, the UPF network element probes inter-satellite link transmission latency by injecting ICMP-based latency probe packets into the network and obtains inter-satellite link transmission bandwidth by reading the number of bytes sent by the network interface card over a period of time.

[0124] After the inter-satellite link status information measurement is completed, it is reported to the routing decision subsystem in the form of a measurement report.

[0125] S6. Receive and execute routing path decisions issued by SMF network elements.

[0126] As described in the previous embodiments, the routing path decisions received by the UPF network element include packet inspection rules, forwarding action rules, QoS enforcement rules, and usage report rules. The UPF network element executes the received routing path decisions and processes the UE's uplink data packets.

[0127] Based on the concept of this application, this application also provides an inter-satellite link load balancing system applied to SMF network elements and an inter-satellite link load balancing system applied to UFP network elements, respectively. For example... Figure 10 The diagram shows the construction of an SMF network element and a UPF network element (only one is shown, the others are similar) in one embodiment. According to the diagram, inter-satellite link load balancing systems are configured in both the SMF and UPF network elements. For ease of distinction, the inter-satellite link load balancing system configured in the SMF network element is called the routing decision subsystem, and the inter-satellite link load balancing system configured in the UPF network element is called the inter-satellite link state information awareness subsystem. The two inter-satellite link load balancing systems can be implemented independently.

[0128] like Figure 10As shown, the routing decision subsystem includes a measurement task distribution interface, a measurement report receiving interface, and a routing analysis module. The routing analysis module distributes measurement tasks to each UPF network element through the measurement task distribution interface, enabling each UPF network element to collect inter-satellite link status information; it receives the inter-satellite link status information from each UPF network element through the measurement report receiving interface; under the constraints of inter-satellite link latency and bandwidth, it calculates the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information; and it distributes the routing path decision to the corresponding UPF network element. The method used by the routing analysis module to calculate the routing path from the source satellite node to the target satellite node is exactly the same as the method used in step S2 of the inter-satellite link load balancing method embodiment applied to SMF network elements described above, and therefore will not be repeated.

[0129] The inter-satellite link status information sensing subsystem includes: a measurement task receiving interface, a measurement report reporting interface, a measurement task execution module, and a policy execution module. The measurement task execution module receives measurement tasks sent by the SMF network element; in response to the measurement task, it collects inter-satellite link status information and reports it to the SMF network element, enabling the SMF network element to calculate the routing path from the source satellite node to the target satellite node based on the inter-satellite link status information, under the constraints of inter-satellite link latency and bandwidth. The policy execution module receives and executes the routing path decisions issued by the SMF network element. The method by which the SMF network element calculates the routing path from the source satellite node to the target satellite node is exactly the same as the method used in step S2 of the inter-satellite link load balancing method embodiment applied to the SMF network element described above, and therefore will not be repeated.

[0130] Combining the routing decision subsystem and the inter-satellite link state information awareness subsystem, such as... Figure 10 As shown, the routing analysis module in the routing decision subsystem generates measurement tasks and distributes them to each UPF network element through the measurement task distribution interface. The measurement task execution module in the inter-satellite link status information perception subsystem receives the measurement tasks through the measurement task receiving interface, collects inter-satellite link status information in response to the measurement tasks, and transmits it to the routing decision subsystem in the form of a measurement report through the measurement report reporting interface. The routing analysis module receives the inter-satellite link status information through the measurement report receiving interface, and when path routing decisions are needed, optimizes the path routing based on the inter-satellite link status information, and distributes the path routing decisions to the UPF network elements through the N4 interface. For example, through the N4 interface, PFCP is used to distribute packet detection rules, forwarding action rules, QoS enforcement rules, and usage report rules indicating the routing path decision information to the corresponding UPF network elements. The policy execution module receives the above routing path decisions through the N4 interface and executes the routing path decisions to process the UE's uplink data accordingly.

[0131] Furthermore, based on the concept of this application, this application also proposes an inter-satellite link load balancing device, applied to the same satellite network, namely, the satellite network includes: a source satellite node connected to a terminal, a target satellite node connected to a data network, and at least one intermediate satellite node, with User Plane Function (UPF) network elements deployed on each satellite node. The device includes a processor and a storage medium, the storage medium storing computer instructions. When the processor executes the computer instructions, it performs the inter-satellite link load balancing method in any of the above embodiments. That is, when the processor executes the computer instructions, it can execute the above-described inter-satellite link load balancing method applied to SMF network elements, or it can execute the above-described inter-satellite link load balancing method applied to UPF network elements. For the former embodiment, the inter-satellite link load balancing device can act as an SMF network element; for the later embodiments, the inter-satellite link load balancing device can act as a UPF network element.

[0132] Similarly, this application also proposes a computer-readable storage medium that can be applied to SMF or UPF network elements in a satellite network, with the satellite network structure being the same as in the previous embodiments. This storage medium stores a computer program. When the program is executed by a processor in a scenario involving SMF network elements, it performs the aforementioned inter-satellite link load balancing method for SMF network elements. Similarly, when the program is executed by a processor in a scenario involving UPF network elements, it performs the same inter-satellite link load balancing method.

[0133] The inter-satellite link load balancing scheme provided in the above embodiments aims to balance the link load on the user plane of the satellite core network. By sensing the inter-satellite link status information and optimizing the inter-satellite paths accordingly, real-time planning of PDU session routing paths can be achieved.

[0134] In this embodiment, the feasibility of the proposed solution was also verified. The core network was deployed using Free5gc, and the core network control plane and user plane were separated. UE and RAN were simulated using UERANSIM, and the test environment was configured according to Table 1.

[0135] Table 1 Test Environment Configuration Table

[0136]

[0137] Measure network topology such as Figure 11 As shown, the network includes one UE, one satellite-based base station (RAN), three satellite-based UPF network elements (UPF1, UPF2, and UPF3), a core network control plane, and a DN. Based on this network topology, the real-time transmission bandwidth and transmission latency between the UPF network elements are measured.

[0138] The inter-satellite link state information sensing and measurement subsystem on UPF1 is responsible for measuring the transmission bandwidth and transmission delay between UPF1's SelfIP and UPF2's NeighborIP. According to the actual measurement results, the transmission delay is 10585. The transmission bandwidth is 0 because no data is being transmitted. The remaining UPF network elements measured the transmission bandwidth and latency between SelfIP and NeighborIP nodes using the same measurement procedure, and then uploaded the data to the SMF network elements. Based on the above experimental results, it is sufficient to prove the feasibility of deploying UPF network elements on satellites and core network control plane network elements on the ground for inter-satellite link state information sensing and acquisition.

[0139] In addition, the performance of the inter-satellite link load balancing method was tested and verified in this embodiment. The test network topology is as follows: Figure 5 As shown, the core network control plane elements are deployed on the ground, and UPF elements are deployed on nine satellites, namely UPF1, UPF2, ..., UPF9. The test network also includes UE, RAN, and DN. Figure 5 It can be seen that in this test network, the satellite where UPF1 is located is the source satellite node, and the satellite where UPF9 is located is the destination satellite node. That is, the UE accesses the satellite network through UPF1 and is routed to UPF9 to access the terrestrial data network.

[0140] Ten iperf3 servers were started in the DN to test the throughput and packet loss rate of different UEs. Every 10 seconds, a UE initiated a PDU session request until all ten users completed the PDU session establishment process. After the UEs established their PDU sessions, the corresponding iperf3 client program was started, setting the bandwidth of the data stream to 1 Mbps, the transmission time to 200 seconds, and the transport layer protocol to UDP. Simultaneously, the IP address assigned to the UE initiated a ping process to the IP address of the DN to test transmission latency. The final comparison results of the average transmission latency, average throughput, and average packet loss rate obtained from different routing path planning algorithms are as follows: Figure 12 , Figure 13 and Figure 14 As shown.

[0141] from Figures 12-14As can be seen, the load balancing method proposed in this application (represented by IAC-ITS) can effectively schedule sessions, and its performance parameters are superior to the shortest path algorithm and the ECMP algorithm. The shortest path algorithm performs the worst because it lacks load balancing capabilities; while the ECMP algorithm has some load balancing capabilities, its performance is inferior to that of this application under non-uniform inter-satellite link throughput. Compared to the shortest path algorithm and the ECMP algorithm, the method of this application can reduce the average transmission latency by 85.37% and 79.65%, respectively; and increase the end-to-end average throughput by 9.82% and 1.22%, respectively. Furthermore, due to the advantage of this method in inter-satellite link state information awareness, it can avoid establishing PDU sessions on congested inter-satellite links, thus its packet loss rate is 0. Experimental results show that the inter-satellite link load balancing method proposed in this application can effectively sense the inter-satellite link status information and make effective routing decisions through the IAC-ITS traffic scheduling algorithm, thereby avoiding network congestion in the user plane of the onboard core network, achieving the purpose of load balancing, and improving the service quality of the onboard core network.

[0142] This invention is not limited to the specific embodiments described above. The invention extends to any new feature or combination disclosed in this specification, as well as any new method or process step or combination disclosed herein.

Claims

1. An inter-satellite link load balancing method, applied to a session management function (SMF) network element in a satellite network, comprising: The satellite network comprises a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and each satellite node is respectively provided with a user plane function (UPF) network element; the method comprises: receiving inter-satellite link state information from each UPF network element; under the inter-satellite link delay and bandwidth limitation, calculating a routing path from the source satellite node to the target satellite node according to the inter-satellite link state information by optimizing the path from the source satellite node to the target satellite node, comprising: initializing the corresponding inter-satellite link pheromone concentration with the residual bandwidth utilization of each inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of a previous satellite node to a next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization and the transmission delay tolerance of the inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, a candidate path is screened out, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor, and the pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the pheromone increment of each inter-satellite link; distributing the routing path decision to the UPF network element of the corresponding satellite node.

2. The inter-satellite link load balancing method of claim 1, wherein, receiving inter-satellite link state information from each UPF network element, comprising: generating a measurement task and sending it to each UPF network element, so that each UPF network element collects the inter-satellite link state information in response to the measurement task and feeds back.

3. The inter-satellite link load balancing method of claim 2, wherein, generating a measurement task and sending it to each UPF network element, comprising: periodically initiating the measurement task, so that each UPF network element periodically collects and reports the inter-satellite link state information.

4. An inter-satellite link load balancing system, applied to an SMF network element in a satellite network, characterized in that, The satellite network comprises a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and each satellite node is respectively provided with a UPF network element; the system comprises: a measurement task issuing interface; a measurement report receiving interface; The route analysis module sends a measurement task to each UPF network element through the measurement task issuing interface, so that each UPF network element collects inter-satellite link state information; receives the inter-satellite link state information from each UPF network element through the measurement report receiving interface; under the inter-satellite link delay and bandwidth limitation, according to the inter-satellite link state information, the route path from the source satellite node to the target satellite node is calculated by optimizing the path from the source satellite node to the target satellite node; the route path decision is distributed to the UPF network element of the corresponding satellite node; the method for optimizing the path from the source satellite node to the target satellite node comprises: initializing the corresponding inter-satellite link pheromone concentration with the residual bandwidth utilization rate of each inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of the previous satellite node to the next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted sum of the bandwidth utilization rate and the transmission delay tolerance of the inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, the candidate path is screened out, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor, and the pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the pheromone increment of each inter-satellite link.

5. An inter-satellite link load balancing apparatus applied to a satellite network, characterized in that, The satellite network includes: a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and a user plane function (UPF) network element is arranged on each satellite node; the device includes a processor and a storage medium, the storage medium stores computer instructions, and the processor executes the computer instructions to perform the inter-satellite link load balancing method according to any one of claims 1-3. 6.A computer readable storage medium, applied to an SMF network element of a satellite network, characterized in that, The satellite network includes: a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and a user plane function (UPF) network element is arranged on each satellite node; the storage medium stores a computer program, and the computer program is executed by the processor to perform the inter-satellite link load balancing method according to any one of claims 1-3.

7. An inter-satellite link load balancing method, applied to a UPF network element in a satellite network, characterized in that, The satellite network includes: a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and a UPF network element is arranged on each satellite node; the method comprises: receiving a measurement task sent by an SMF network element; In response to the measurement task, inter-satellite link state information is collected and reported to the SMF network element, so that the SMF network element calculates a routing path from the source satellite node to the target satellite node by optimizing the path from the source satellite node to the target satellite node according to the inter-satellite link state information under the constraints of inter-satellite link delay and bandwidth, which includes: initializing the corresponding inter-satellite link pheromone concentration with the residual bandwidth utilization rate of each inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of a previous satellite node to a next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization rate and the transmission delay tolerance of the inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, candidate paths are screened out, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor, and the pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the inter-satellite link pheromone increment; Receiving and executing the routing path decision issued by the SMF network element.

8. The inter-satellite link load balancing method of claim 7, wherein, In response to the measurement task, inter-satellite link state information is collected and reported to the SMF network element, so that the SMF network element calculates a routing path from the source satellite node to the target satellite node by optimizing the path from the source satellite node to the target satellite node according to the inter-satellite link state information under the constraints of inter-satellite link delay and bandwidth, which includes: initializing the corresponding inter-satellite link pheromone concentration with the residual bandwidth utilization rate of each inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of a previous satellite node to a next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization rate and the transmission delay tolerance of the inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, candidate paths are screened out, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor, and the pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the inter-satellite link pheromone increment; The satellite network includes: a source satellite node connected with a terminal, a target satellite node connected with a data network, and at least one intermediate satellite node, and a UPF network element is arranged on each satellite node; the system includes: A measurement task receiving interface; 9.A system for inter-satellite link load balancing, applied to a UPF network element in a satellite network, characterized in that, A measurement report reporting interface; A measurement task execution module receives a measurement task sent by an SMF network element; in response to the measurement task, inter-satellite link state information is collected and reported to the SMF network element, so that the SMF network element calculates a routing path from the source satellite node to the target satellite node by optimizing the path from the source satellite node to the target satellite node according to the inter-satellite link state information under the constraints of inter-satellite link delay and bandwidth, which includes: initializing the corresponding inter-satellite link pheromone concentration with the residual bandwidth utilization rate of each inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, the transfer probability of a previous satellite node to a next satellite node is calculated based on the expected heuristic factor of the inter-satellite link between the previous satellite node and the next satellite node, and the expected heuristic factor is obtained by weighted summation of the bandwidth utilization rate and the transmission delay tolerance of the inter-satellite link; in the process of optimizing the path from the source satellite node to the target satellite node, after each iteration, candidate paths are screened out, and the pheromone concentration of each inter-satellite link is volatilized according to the pheromone volatilization factor, and the pheromone concentration of each inter-satellite link in the candidate path is updated according to the pheromone increment factor of the candidate path and the inter-satellite link pheromone increment; A policy execution module receives and executes the routing path decision issued by the SMF network element. ​ ​ 10. An inter-satellite link load balancing apparatus applied to a satellite network, characterized in that, The satellite network comprises a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and each satellite node is respectively deployed with a user plane function UPF network element; the device comprises a processor and a storage medium, the storage medium stores computer instructions, and the processor executes the computer instructions to perform the inter-satellite link load balancing method of claim 7 or 8. 11.A computer readable storage medium, applied to a UPF network element of a satellite network, characterized in that, The satellite network comprises a source satellite node connected with a terminal, a target satellite node connected with a data network and at least one intermediate satellite node, and each satellite node is respectively deployed with a user plane function UPF network element; the storage medium stores a computer program, and the computer program is executed by the processor to perform the inter-satellite link load balancing method of claim 7 or 8.

Images