Design method of satellite network routing protocol based on numerical control separation

Abstract

The invention provides a satellite network routing protocol design method based on numerical control separation, which comprises the following steps: defining a satellite network space architecture based on numerical control separation, the control plane and the data plane are provided with a control signaling optimization design module, a terminal capturing and registering module, a real-time information issuing and collecting module and a transmission path establishing and releasing module according to functions. The control signaling optimization design module is adopted to optimize routing signaling and process design, generation of redundant signaling can be reduced, the signaling propagation range can be shortened, and therefore the signaling overhead can be reduced; a real-time information issuing and collecting module is adopted, change information of all links in a network can be collected, sent and processed in real time, a link state database is formed, and the link state database is dynamically updated according to changes of link states; and a transmission path establishment and release module is adopted, so that high-weight feature optimization can be quickly selected, and the routing convergence speed can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite network data transmission, and in particular to a design method for a satellite network routing protocol based on the separation of number control. Background Art

[0002] The function of a single satellite is relatively single. However, when multiple satellites cooperate to form a satellite network, not only can more abundant functions be achieved, but also the anti-destruction ability of the satellite system will be greatly improved. It will be an important trend in the development of satellite networks to form a satellite system by networking multiple independent satellites to cooperate with each other to complete complex space exploration tasks.

[0003] The satellite network topology is characterized by frequent changes and periodicity. In the classical networking method, the constellation mode is a method with good prospects. It mainly relies on the existing mature satellite network space architecture in space. In these existing space network architectures, each satellite orbit forms a stable space structure, the relative position relationship between satellites changes periodically, and similar control mechanisms are adopted between satellites. The networking in the constellation can achieve complete coverage of a specific area, which is conducive to the smooth completion of space missions. The multi-layer heterogeneous hybrid satellite network plays a very important role and significance. However, the multi-layer satellite network has high complexity, a large number of nodes, a large number of links, and due to the continuous movement of satellite nodes, the entire network topology is also in a rapid dynamic change. Therefore, the routing technology is more complex.

[0004] Foreign research on satellite network routing technology started earlier, and many of them have become typical reference standards. However, many schemes are only simulated and verified and have not been actually verified in orbit. Moreover, there are now many types of satellite services and different information service requirements, and new networking schemes and dynamic priority routing and other technologies need to be explored to meet future development needs. Although domestic research on satellite network routing technology started slightly later, there are achievements in different directions, and some research has applied the latest technical achievements. Most of these researches are still in the initial stage and have not formed an effective solution. With the rapid development of satellite networks and the increasing requirements for information services, a new satellite network routing protocol is needed, which can reduce signaling overhead and improve data transfer efficiency to meet the needs of future large-scale satellite networking development. Summary of the Invention

[0005] In order to achieve the above object, the present invention provides a design method for a satellite network routing protocol based on the separation of number control, including the following steps:

[0006] A. Define the satellite network space architecture based on the separation of number control

[0007] The satellite network space architecture includes a ground network management center and a multi-layer satellite network. Each layer of satellite cluster is composed of satellite nodes, and the satellite nodes are configured as control nodes, switching nodes, and forwarding nodes. The ground network management center and the control nodes form a control plane, and the switching nodes and the forwarding nodes form a data plane. The control nodes of the satellite clusters in the same layer form a controller cluster, and a primary controller is set in each layer's controller cluster, and the others are secondary controllers;

[0008] The control plane and the data plane are configured with a control signaling optimization design module, a terminal capture and registration module, a real-time information publishing and collection module, and a transmission path establishment and release module according to their functions;

[0009] The control signaling optimization design module is used to maintain the network topology. When the link state changes, it triggers control signaling packets to update the link state information, so that the link state database can reflect the current latest link state and can keep the link state databases on each node synchronized and consistent;

[0010] The terminal capture and registration module is responsible for establishing a communication link between the satellite node and the terminal. This process can be initiated by the terminal to make a registration call or accept an incoming call;

[0011] The real-time information publishing and collection module can collect, send, and process the change information of each link in the network in real time, form a link state database, and dynamically update the link state database according to the change of the link state;

[0012] The transmission path establishment and release module calculates the transmission path according to the dynamic weight routing algorithm to construct a virtual path when transmitting data, and distributes the routing table to each node passing along the way, and releases the virtual path after the data transmission is completed;

[0013] B. Optimize the control signaling packets

[0014] The optimized control signaling data includes Link State Advertisement (LSA), Keep_Alive packet, Ack_Alive packet, Inform_bandwidth packet, Link State Update (LSU) packet, and Link State Acknowledgment (LSAck) packet;

[0015] The Link State Advertisement (LSA) is generated by the control node and the switching node to maintain information about the routing information of neighbor nodes and the channel cost;

[0016] The Keep_Alive packet is a detection message for judging whether the link has changed and is sent between two adjacent satellite nodes;

[0017] The Ack_Alive packet is sent in response by the satellite node that has received the Keep_Alive packet within a certain period of time, and is used to establish and maintain the adjacency relationship of the link;

[0018] The Inform_bandwidth data packet is used to notify the control node of the remaining bandwidth of the current link and the port number of the data packet passing through the satellite node;

[0019] The Link State Update packet LSU is used to notify the control node to update the link state information;

[0020] The Link State Acknowledgment packet LSAck is sent by the control node after updating its link state information, and is used to inform the source end that it has updated its link state information, and the data packet also contains the specific updated link state information;

[0021] C. Design the operation mechanism of the terminal capture and registration module

[0022] 1) The satellite node sends a broadcast;

[0023] 2) After receiving the broadcast, the terminal sends an access request to the satellite node;

[0024] 3) After receiving the access request, the satellite node forwards the access request to the ground network management center;

[0025] 4) The ground network management center analyzes and processes the terminal access application information;

[0026] 5) The ground network management center sends the access success response information to the satellite node, and the satellite node forwards the access success response information to the terminal;

[0027] 6) After receiving the access success response information, the terminal confirms successful access, and at the same time sends terminal access confirmation information to the ground network management center;

[0028] 7) After receiving the terminal access confirmation information, the ground network management center confirms whether the terminal has successfully accessed and updates the local network status;

[0029] 8) Upload the network status to the satellite node;

[0030] 9) After completing the terminal capture and access, the new terminal has obtained a service channel and the identity of the control node. The new terminal sends an access confirmation message to the control node on the obtained service channel. After receiving this message, the control node sends a terminal access confirmation response message to the new terminal to confirm that the identity of the new terminal is validly authorized, and the registration process is completed;

[0031] D. Design the operation mechanism of the real-time information publishing and collection module

[0032] The master and slave controllers of each layer of the satellite network share the information of each layer by configuring the following functions:

[0033] 1) Install control logic with atomicity to ensure the consistency of control logic;

[0034] 2) The slave controller monitors the status of satellites in the domain in real time;

[0035] 3) The master controller is responsible for collecting the global status of the on-board subnetworks to ensure that the master controller has the global network topology and status;

[0036] 4) The master controller provides fast routing guarantee based on SRv6;

[0037] 5) Guarantee for fast fault reporting and event transmission based on priority;

[0038] The propagation of events between the master and slave controllers adopts the publish / subscribe message mode. Each controller subscribes to the data channel, control channel, and its own channel in the network and is granted the permission to publish and subscribe to information. The controller can publish the local network and applications to the data channel, and publish a specific controller event and command in the corresponding control channel. Each controller periodically announces local messages in the control channel to enable the discovery and detection of faults in the controller;

[0039] The publish / subscribe message mode responsible for the controller communication service provides the following services:

[0040] 1) Provide persistent storage for published events and maintain the order of events published by the same controller;

[0041] 2) Immune to network partitioning, and each partition must operate independently and achieve synchronization when reconnected;

[0042] 3) Reduce the cross-site traffic required to propagate events, that is, the controller of one site should obtain updates of most other sites from nearby controllers to avoid congestion in cross-region connections;

[0043] 4) Strengthen access control to ensure access ownership;

[0044] E. Design the path selection algorithm

[0045] Suppose the topological structure diagram of the satellite network in a certain time period is G(V, E, C), where V is the set of nodes, the element v ∈ V is a node of the graph G, n is the number of nodes in the network (n = |V|), E is the finite edge set, and the element (u, v) ∈ E is denoted as (u, v) = u → v and is called an edge of the graph G. m is the number of links in the network (m = |E|), and C represents the cost function defined for each edge between node i and node j in the time period. The initial value is determined by factors such as link available bandwidth and transmission delay. The set of link remaining capacities is the remaining capacity on link (u, v). Element (S, D) is the source and destination node pair of the traffic flow. The LSP request is (S, D, bw), where bw is the requested bandwidth.

[0046] E1. Statistically analyze the network resource information based on the control nodes in the satellite network;

[0047] E2. When sending request information between source / destination node pairs, first delete the links in the satellite network whose remaining bandwidth is less than the requested bandwidth or the propagation delay is greater than the requested delay, and streamline the network structure to reduce the computational complexity of the algorithm;

[0048] E3. When a traffic transmission request between source / destination nodes appears, query the remaining bandwidth, link transmission delay, error rate, and network topology resource information of each link in the satellite network;

[0049] E4. Calculate the K shortest paths for each source / destination node pair according to the remaining bandwidth and transmission delay of the link. The steps are as follows:

[0050] (1) Set the initial weight of each link according to the remaining capacity of the link as:

[0051]

[0052] In the formula, is the remaining capacity of link (u, v), is the maximum value of the remaining capacity of the links in the current network link. The weight of the link with the largest remaining capacity in the network is 1, and the smaller the remaining capacity of the link, the larger the corresponding weight. D(u, v) is the transmission delay of link (u, v), D min is the minimum delay of all links, R(u, v) is the bit error rate of link (u, v), R min is the minimum bit error rate of all links, lg(R(u, v) / R min ) is the error rate of link (u, v), α 1 , λ 1 and β 1 are link weight adjustment factors;

[0053] (2) Using the link weights obtained by formula (1), find a shortest path between the current source / destination node pair as the critical path through the Dijkstra algorithm. Before finding the next critical path, delete the link with the largest weight in the current path from the network, so as to ensure that the next critical path will not repeat the previous path;

[0054] (3) Repeat step (2) until there is no reachable path between the current source / destination node pair or the number of obtained paths is K. Use K′ to represent the actual number of critical paths between the source / destination node pair;

[0055] Steps (1)-(3) are executed for all source / destination node pairs to obtain the critical path sets of all nodes;

[0056] E5. Dynamically and adaptively adjust the link weights

[0057] After determining the critical path, it is necessary to reset the weights of the critical paths in the network. The weights of the critical paths are defined as two parts: the initial weight and the conflict degree. The initial weight is formula (1), and the conflict degree is defined as:

[0058]

[0059] In the formula, (s′, d′) is a certain source / destination node pair, (s, d) is the source / destination node pair of the current service flow request, P is the set of all source / destination node pairs, and p = |P| is the number of elements in the set P. is the i-th critical path of (s′, d′), represents the importance of, λ 2 is a symbol identifier. When (s′, d′) is other node pairs external to the current request node, λ 2 = 1. When (s′, d′) is the current request node pair, λ 2 = -1;

[0060] Set the weight of the critical path to:

[0061]

[0062] In the formula, C 0 (u, v) is formula (1), which is the initial value of the link weight, representing the contributions of the remaining link capacity, transmission delay, and link error rate to the link weight. δ(u, v) is formula (2), representing the impact of conflicts on the link weight. α 2 is a scaling factor used to adjust the contribution of conflicts to the link weight. then represents the increment of the link weight on the i-th path of (s′, d′). For non-critical paths, C(u, v) is equal to C 0 (u, v);

[0063] E6. Calculate the route according to the adjusted weights, record the path and its occupied bandwidth, and establish a connection to transmit data.

[0064] After adopting the above technical solution, the present invention has the following beneficial effects:

[0065] (1) Optimize the routing signaling and process design, which can reduce the generation of redundant signaling, narrow the signaling propagation range, and thus reduce the signaling overhead;

[0066] (2) The designed real-time information release and collection module can collect, send, and process the change information of each link in the network in real time, form a link state database, and dynamically update the link state database according to the change of the link state;

[0067] (3) Adopting a dynamic weight routing algorithm can quickly select high-weight features for optimization, thereby improving the routing convergence speed. Brief Description of the Drawings

[0068] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0069] Figure 1 It is the format of the link state advertisement LSA;

[0070] Figure 2 It is the format of the Keep_Alive packet;

[0071] Figure 3 It is the format of the Ack_Alive packet;

[0072] Figure 4 It is the format of the Inform_bandwidth data packet;

[0073] Figure 5 It is the format of the link state update packet LSU;

[0074] Figure 6 It is the format of the link state acknowledgment packet LSAck;

[0075] Figure 7 It is the header content of the LSA in the LSAck;

[0076] Figure 8 It is the schematic diagram of the controller event propagation;

[0077] Figure 9 It is the schematic diagram of the service request and path planning;

[0078] Figure 10 It is the schematic diagram of the user terminal A sending a Path message;

[0079] Figure 11 It is the schematic diagram of the user terminal B sending a Resv message;

[0080] Figure 12 It is the schematic diagram of forwarding data packets according to the virtual path label;

[0081] Figure 13 Schematic diagram for demolishing virtual path tunnel Specific implementation manners

[0082] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0083] The embodiments of the present disclosure provide a design method for a satellite network routing protocol based on numerical control separation, including the following steps:

[0084] A. Define the satellite network space architecture based on numerical control separation

[0085] The satellite network space architecture includes a ground network management center and multiple layers of satellite networks. Each layer of satellite cluster is composed of satellite nodes. The satellite nodes are configured as control nodes, switching nodes, and forwarding nodes. The ground network management center and the control nodes form a control plane, and the switching nodes and the forwarding nodes form a data plane. The control nodes of the satellite clusters in the same layer form a controller cluster. A primary controller is set in each layer of the controller cluster, and the others are slave controllers;

[0086] The control plane and the data plane are correspondingly configured with a control signaling optimization design module, a terminal capture and registration module, a real-time information publishing and collecting module, and a transmission path establishment and release module;

[0087] The control signaling optimization design module is used to maintain the network topology structure. When the link state changes, it triggers a control signaling packet to update the link state information, so that the link state database can reflect the current latest link state and can keep the link state databases on each node synchronized and consistent;

[0088] The terminal capture and registration module is responsible for establishing a communication link between the satellite node and the terminal. This process can be initiated by the terminal to initiate a registration call or accept an incoming call;

[0089] The real-time information publishing and collecting module can collect, send, and process the change information of each link in the network in real time, form a link state database, and dynamically update the link state database according to the change of the link state;

[0090] The transmission path establishment and release module calculates the transmission path according to the dynamic weight routing algorithm to construct a virtual path when transmitting data, and distributes the routing table to each node passing along the way, and releases the virtual path after the data transmission is completed;

[0091] B. Optimize the control signaling packet

[0092] All kinds of control signaling data packets in the protocol are mainly used to maintain the satellite network topology. Once the link between nodes fails or is interrupted, control signaling data packets are needed to update the link state information, so that the link state database can reflect the current latest link state, and can keep the link state databases on each node synchronized and consistent. The optimized control signaling data includes Link State Advertisement (LSA), Keep_Alive packet, Ack_Alive packet, Inform_bandwidth packet, Link State Update (LSU) packet, and Link State Acknowledgment (LSAck) packet, as shown in Table 3.

[0093] Table 3 describes the data packets of router link state information

[0094]

[0095] Link State Advertisement (LSA) is generated by the control node and the switching node to maintain information about neighbor node routing information and channel costs. The set of LSAs constitutes the link state database. Each different type of LSA has different functions. Each LSA starts with a 20-byte standard header. The LSA header contains the link state type / LS type, link state identifier / LS identifier, and advertising router. The combination of these three fields uniquely identifies the LSA, as Figure 1 shown.

[0096] At the same time, there may be multiple instances of LSAs in the AS. It is necessary to determine which instance is newer, which is achieved by checking the LS sequence number, LS checksum, and LS age. These fields are all included in the 20-byte LSA header. Some protocol packets need to list the LSAs. When the instance version is not required, the LSA is referenced by the LS type, LS identifier, and advertising router. Otherwise, the LS sequence number, LS age, and LS checksum fields must be added.

[0097] The Keep_Alive packet is a probe message for determining whether the link has changed. It is sent between two adjacent satellite nodes. Its protocol value in the IP packet header is 193, and the type field of the data packet is assigned the value 1.

[0098] When the satellite network topology changes, that is, when the network enters a new time segment, it can be considered that the network topology is fixed within this time period. However, in a satellite network, link congestion and interruption may also occur. According to the protocol regulations, a node will send a probe packet to its neighbors at regular intervals to learn about the link connection status of its neighbors and report it to the controller. Before sending the probe packet, the node will assign corresponding values to the Keep_Alive packet, filling in the corresponding time slice and sequence number, as well as the source address and destination address. Once the current node does not receive a reply from its neighbor (within a certain time range), it can be considered that this link is interrupted. Finally, it is necessary to update the link connection status data of the node in a timely manner. The format of the Keep_Alive packet is as Figure 2 shown.

[0099] The Ack_Alive packet is sent in response by a satellite node that has received a Keep_Alive packet within a certain time, and is used to establish and maintain the adjacency relationship of the link. When a node receives a Keep_Alive packet sent by an adjacent node, if they are neighbors, the current node will send an acknowledgement packet Ack_Alive to the node that sent the Keep_Alive packet. In this way, the adjacency relationship between the two can be established and maintained within a certain period of time. The format of the Ack_Alive packet is as Figure 3 shown.

[0100] The Inform_bandwidth data packet is used to notify the control node of the remaining bandwidth of the current link and the port number through which the data packet passes through the satellite node. In a satellite network, when a data packet is sent from one node to another and the bandwidth of the link through which the data packet passes changes, the data packet will send an Inform_bandwidth data packet to the control node when it reaches the next hop, used to notify the remaining bandwidth of the current link and the port number through which the data packet passes through the terminal node.

[0101] When the Inform_bandwidth passes through each node and each link, this field is used to indicate and describe the corresponding bandwidth. Each time it passes through a routing node, the protocol will update this field accordingly, thus obtaining a real-time data, that is, the consumed bandwidth. The Inform_bandwidth data packet is used to notify the control node to change the weight value of the link bandwidth after occupying the bandwidth, and is sent through status triggering. When the occupied bandwidth value exceeds a threshold, an Inform_bandwidth data packet will be sent to the control node.

[0102] Status trigger case 1: If the occupied bandwidth (remaining bandwidth) of a certain node exceeds a threshold, the current router will send an Inform_bandwidth data packet to the control node through status triggering.

[0103] Status Trigger Condition 2: If the occupied bandwidth of a link or a node changes significantly within a certain period of time, an Inform_bandwidth data packet is sent to the control node through status trigger.

[0104] Status Trigger Condition 3: If the occupied bandwidth of each link in the entire network reaches the threshold value, each router node also sends an Inform_bandwidth data packet to the control node through status trigger. The format of the Inform_bandwidth packet is as Figure 4 shown.

[0105] The Link State Update packet LSU is used to notify the control node to update the corresponding link state information. The protocol value of LSU in the IP header is 89, and the LSU packet type is 4. It sends the required Link State Acknowledgment packet LSA to its neighbors. Each LSU packet forwards the LSAs it contains to the next hop further from its origin. Multiple LSAs may be included in one packet. The beginning of each LSA is a 20-byte common header. The format of the LSU data packet is as Figure 5 shown.

[0106] The Link State Acknowledgment packet LSAck is sent by the control node after it updates its link state information, used to inform the source end that it has updated its link state information, and the data packet also contains the specific updated link state information.

[0107] The protocol value of LSAck in the IP header is 89, and the packet type is 5. The format of the LSAck packet is as Figure 6 shown.

[0108] When the control node receives the LSU data packet sent by the source end, it will update the corresponding link state information. After the control node updates its link state advertisement, it will send an LSAck packet to the source end. This data packet is used to inform the source end node that it has updated its link state information, and the LSAck packet also contains the specific updated link state information.

[0109] According to the status of the current sending interface and the sending router of the LSU packet, the LSAck packet may be sent to the multicast address AllSPFRouters or AllDRouters or use unicast. Each LSA is confirmed through the description in the LSA header. These information uniquely identify an LSA and its instance, as Figure 7 shown.

[0110] C. Design the operating mechanism of the terminal capture and registration module

[0111] Capture and registration are for establishing a communication link between space and ground or between satellites. The process includes the following steps:

[0112] 1) The satellite node sends a broadcast;

[0113] 2) After the terminal receives the broadcast, it sends an access request to the satellite node;

[0114] 3) After the satellite node receives the access request, it forwards the access request to the ground network management center;

[0115] 4) The ground network management center analyzes and processes the terminal access application information;

[0116] 5) The ground network management center sends the access success response message to the satellite node, and the satellite node forwards the access success response message to the terminal;

[0117] 6) After the terminal receives the access success response message, it confirms the successful access and simultaneously sends a terminal access confirmation message to the ground network management center;

[0118] 7) After the ground network management center receives the terminal access confirmation message, it confirms whether the terminal has successfully accessed the network and updates the local network status;

[0119] 8) Upload the network status to the satellite node;

[0120] 9) After the terminal capture and access are completed, the new terminal has obtained a service channel and the identity of the control node. The new terminal sends an access confirmation message to the control node on the obtained service channel. After receiving this message, the control node sends a terminal access confirmation response message to the new terminal to confirm that the identity of the new terminal is validly authorized, and the registration process is completed;

[0121] D. Design the operation mechanism of the real-time information publishing and collection module

[0122] The information publishing and collection module is mainly responsible for collecting, sending, and processing the change information of each link in the network, and finally forming a link state database. As the network scale expands, a single control node is difficult to meet the computing management requirements. The controllers of the satellite clusters at the same layer need to form a cluster to share the topology change information within the satellite layer. At this time, the transmission delay requirement for each node to its directly connected node is relatively high. A master controller Master_Ctrl needs to be set in each layer of controller cluster, as Figure 8 shown. The master and slave controllers of each layer of the satellite network share the information of each layer by configuring the following functions:

[0123] 1) Install control logic with atomicity to ensure the consistency of control logic;

[0124] 2) The slave controller monitors the satellite status within the domain in real time;

[0125] 3) The master controller is responsible for collecting the global status of the on-board subnet to ensure that the master controller has the global network topology and status;

[0126] 4) The master controller provides fast routing guarantee based on SRv6;

[0127] 5) Fast fault reporting and event transmission guarantee based on priority;

[0128] The propagation of events between the master and slave controllers adopts the publish / subscribe message mode. Each controller subscribes to the data channel, control channel, and its own channel in the network and is granted the permission to publish and subscribe to information. The controller can publish local networks and applications to the data channel, and publish specific controller events and commands in the corresponding control channel. Each controller periodically announces local messages in the control channel to enable the discovery and detection of faults by the controller;

[0129] The publish / subscribe message mode responsible for the controller communication service provides the following services:

[0130] 1) Provide persistent storage for published events and maintain the order of events published by the same controller;

[0131] 2) Immune to network partitioning, and each partition must operate independently and achieve synchronization when reconnected;

[0132] 3) Reduce the cross-site traffic required to propagate events, that is, the controller of one site should obtain updates from nearby controllers for most other sites, avoiding congestion in cross-regional connections;

[0133] 4) Strengthen access control to ensure access ownership;

[0134] E. Design a path selection algorithm

[0135] The basic idea of the algorithm is: Define the initial weight of the satellite network link as a comprehensive function of the remaining bandwidth of the link, the link transmission delay, and the error rate (bit error rate). When selecting an LSP for the current node pair's path request, a cooperation strategy is adopted, fully considering the future requests of other node pairs for the link, and trying to avoid links with high criticality (conflict degree) in the network. A complete cooperation strategy can dynamically adjust the weight of the link in the satellite network by adaptively changing the weight of the link, thereby delaying the selection of links with high criticality and ultimately achieving the goal of minimizing interference. Therefore, the algorithm design mainly includes two parts: the determination of the criticality of the satellite link and the adaptive dynamic adjustment of the weight of the satellite link. On the basis of this part of the work, the single-source shortest path Dijkstra algorithm is used to obtain the actual label switching path for the current node pair's request, ultimately realizing the overall efficient utilization of satellite network resources.

[0136] Let the topological structure diagram of the satellite network within a certain time period be \(G(V, E, C)\), where \(V\) is the set of nodes, the element \(v\in V\) is a node of the graph \(G\), \(n\) is the number of nodes in the network (\(n = |V|\)), \(E\) is the finite edge set, and the element \((u, v)\in E\) is denoted as \((u, v)=u\rightarrow v\) and is called an edge of the graph \(G\), \(m\) is the number of links in the network (\(m = |E|\)), \(C\) represents the cost function defined on each edge between node \(i\) and node \(j\) in the time period, and the initial value is determined by factors such as link available bandwidth and transmission delay. The link remaining capacity set is the remaining capacity on the link \((u, v)\), the element \((S, D)\) is the source and destination node pair of the traffic flow, and the LSP request is \((S, D, bw)\), where \(bw\) is the requested bandwidth.

[0137] E1. Statistically analyze the network resource information according to the control nodes in the satellite network;

[0138] E2. When sending request information between source / destination node pairs, first delete the links in the satellite network where the remaining bandwidth is less than the requested bandwidth or the propagation delay is greater than the requested delay, and streamline the network structure to reduce the computational complexity of the algorithm;

[0139] E3. When a traffic transmission request between source / destination nodes appears, query the remaining bandwidth, link transmission delay, error rate, and network topology resource information of each link in the satellite network;

[0140] E4. Obtain the \(K\) shortest paths, also called critical paths, for each source / destination node pair according to the link remaining bandwidth and transmission delay. The principle for selecting the critical paths is that for each source / destination node pair, the link weights selected are the shortest (minimum weight) paths calculated from the remaining capacity and transmission delay. The link weight is related to the link remaining capacity, which is beneficial to keeping the network links unobstructed (the link remaining capacity is non-zero), beneficial to achieving network load balancing, and also related to the transmission delay, which is beneficial to ensuring the transmission delay of real-time data.

[0141] The actual LSP path requested for each source / destination node pair needs to comprehensively consider the demand status of all source / destination node pair links. The steps to obtain the \(K\) critical paths for a certain source / destination node pair are as follows:

[0142] (1) Set the initial weight of each link according to the link remaining capacity as:

[0143]

[0144] In the formula, is the remaining capacity of the link \((u, v)\), is the maximum value of the link remaining capacity in the current network links. The weight of the link with the largest remaining capacity in the network is 1, and the smaller the link remaining capacity, the larger the corresponding weight. \(D(u, v)\) is the transmission delay of the link \((u, v)\), \(Dmin is the minimum delay of all links, R(u, v) is the bit error rate of link (u, v), and R min is the minimum bit error rate of all links, and lg(R(u, v) / R min ) is the error rate of link (u, v), and α 1 , λ 1 and β 1 are link weight adjustment factors;

[0145] (2) Using the link weights obtained by formula (1), find a shortest path between the current source / destination node pair as the critical path through the Dijkstra algorithm. Before finding the next critical path, delete the link with the largest weight in the current path from the network, so as to ensure that the next critical path will not repeat the previous path;

[0146] (3) Repeat step (2) until there is no reachable path between the current source / destination node pair or the number of obtained paths is K (usually K is taken as 3 - 4), and use K′ to represent the actual number of critical paths between the source / destination node pair;

[0147] Execute steps (1)-(3) for all source / destination node pairs to obtain the critical path set of all nodes;

[0148] E5. Dynamically and adaptively adjust the link weights

[0149] After determining the critical path, it is necessary to reset the weights of the critical paths in the network. Define the weights of the critical paths as two parts, the initial weight and the conflict degree. The initial weight is formula (1), and the conflict degree is defined as:

[0150]

[0151] In the formula, (s′, d′) is a certain source / destination node pair, (s, d) is the source / destination node pair of the current traffic flow request, P is the set of all source / destination node pairs, and p = |P| is the number of elements in set P. is the i-th critical path of (s′, d′), represents the importance degree of, and λ 2 is a symbol identifier. When (s′, d′) is a node pair other than the current request node to the outside, λ 2 = 1. When (s′, d′) is the current request node pair, λ 2 = -1;

[0152] Set the weight of the critical path as:

[0153]

[0154] In the formula, C0 (u, v), namely formula (1), is the initial value of the link weight, representing the contributions of link remaining capacity, transmission delay, and link error rate to the link weight. δ(u, v), namely formula (2), represents the impact of conflicts on the link weight, and α 2 is a scaling factor used to adjust the contribution of conflicts to the link weight. represents the increment of the link weight on the i-th path of (s′, d′). For non-critical paths, C(u, v) is equal to C 0 (u, v);

[0155] As can be seen from formula (5), in the setting of link weights, the resource status of the link, i.e., the remaining capacity, and the conflicts and competitions of each node pair on the link are comprehensively considered. λ in formula (5) 2 reflects the combination of conflict avoidance and competition strategies. When reserving a link for a certain source / destination node pair, it should be ensured that when the actual request of this source / destination node pair arrives, other source / destination node pairs are preferably selected to reserve these links for it. Therefore, in formula (5), when a certain link belongs to the critical path of other node pairs outside the source / destination node pair, the conflict degree of this link increases which means that the weight of this link also needs to increase, and the current request should preferably avoid this link when selecting an LSP. If a certain link belongs to the critical path of the source / destination node pair itself, the conflict degree of this link increases which means that the weight of this link also needs to decrease. Therefore, when selecting an LSP for the current request, this link will be preferably selected to a certain extent. In this way, a competition strategy is introduced on the basis of conflict avoidance, so that when selecting an LSP path for the current request, on the one hand, the links on the critical paths of other node pairs are preferably avoided, and on the other hand, the links on the critical paths of the current source / destination node pair are preferably selected as much as possible, thereby ensuring the high efficiency of the algorithm performance.

[0156] E6. Calculate the route according to the adjusted weight, record the path and its occupied bandwidth, and establish a connection to transmit data.

[0157] The process of transmitting data using the satellite network protocol designed by the above method is as follows:

[0158] Step 1: Calculate the routing path based on the link state database to generate a routing table

[0159] Such as Figure 9As shown in the figure, when user terminal A sends a transmission request to user terminal B, the transmission request reaches the control node. The control node calculates the transmission path and distributes the routing table to the satellite nodes on the path. When the data packet reaches SAT1, it is matched with the routing table in SAT1 through the destination address of the data packet. By looking up all its next hops, the complete routing information of the data packet can be obtained, that is, a strict source route (in order to reach the destination address of the data packet, a complete and exact path is selected for it, which includes the address information of each necessary path). In addition to marking the destination network, interface, and hop count limit in the routing table entry, the delay information is also noted.

[0160] Step 2: Establish signaling nodes and tunnels according to the routing table

[0161] Figure 10 It shows the specific content of the Path message sent by user terminal A, as well as the data structure of the preserved link state block (PSB) when passing through the router node for each sent Path message. Among them, ERO (Explicit Route Object) specifies a predefined and clear path for the service data packet.

[0162] As Figure 11 shown, in the Resv message generated by user terminal B, the label is also recorded. RRO (Reserved Route Object) saves the IP addresses of the routers traversed by the virtual path tunnel. In the virtual path tunnel, it is possible to check whether there is a loop in the route through RRO and collect detailed path information during the virtual path establishment process hop by hop.

[0163] Step 3: Forward data packets according to the established tunnels

[0164] On the established path (LSP), the control node distributes the routing table to each satellite during link establishment and label distribution, and records the corresponding ingress label and egress label on each satellite node.

[0165] After allocating labels to each satellite using the label distribution protocol, each interface on the satellite is assigned a label number. The ingress label represents the IP address of the ingress port, and the egress label represents the IP address of the egress port.

[0166] After the label establishment and distribution are completed, the relevant information in the RSB (ingress label, egress label, and the corresponding interface IP address) is assigned to the corresponding entry in the routing table of the satellite node. In this way, each router on each satellite stores the label number corresponding to the next hop, and thus the service data packet can start to be forwarded.

[0167] As Figure 12As shown, when forwarding service data packets, corresponding ingress labels and egress labels can be used for forwarding, and it is no longer necessary to forward data based on IP addresses. Figure 12 Some entries in the new routing table on each satellite node are listed in Figure 12 .

[0168] Since user terminals A and B are registered on satellites SAT1 and SAT4 respectively, and the satellites report the registration information to the control node through LLDP, after the control node calculates the routing information, it sends the transmission path information to the nodes along the way.

[0169] Step 4: Tunnel demolition and signaling node deletion

[0170] As Figure 13 shown, the Pathtear message is used to demolish the tunnel. After this message is generated, it is immediately sent to the nodes on the virtual path, and the path status and related reservation status of the nodes along the way are immediately deleted. It is sent when it is necessary to actively demolish the created tunnel after the service transmission ends, or when the status on the virtual path times out.

[0171] Although the present invention has been disclosed as above with embodiments, it is not intended to limit the present invention. Any person skilled in the art within the technical field can make some changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to what is defined by the claims.

Claims

1. A design method for satellite network routing protocol based on numerical control separation, characterized in that: The steps include: A. Define the satellite network space architecture based on CNC separation The satellite network space architecture includes a ground network management center and a multi-layer satellite network. The satellite clusters at each layer are composed of satellite nodes. The satellite nodes are configured as control nodes, switching nodes and forwarding nodes. The ground network management center and the control nodes form the control plane, and the switching nodes and forwarding nodes form the data plane. The control nodes of the satellite clusters at the same layer form a controller cluster. A master controller is set in the controller cluster of each layer, and the others are slave controllers. The control plane and data plane are configured with control signaling optimization design module, terminal capture and registration module, real-time information publishing and collection module and transmission path establishment and release module according to their functions; The control signaling optimization design module is used to maintain the network topology. When the link status changes, the control signaling data packet is triggered to update the link status information, so that the link status database can reflect the latest link status and keep the link status databases on each node synchronized and consistent. The terminal capture and registration module is responsible for establishing a communication link between the satellite node and the terminal. This process can be started by the terminal initiating a registration call or accepting an external call; The real-time information publishing and collection module can collect, send and process the change information of each link in the network in real time to form a link state database, and dynamically update the link state database according to the change of the link state; The transmission path establishment and release module calculates the transmission path and constructs a virtual path according to the dynamic weight routing algorithm when transmitting data, and sends the routing table to each node along the way, and releases the virtual path after the data transmission is completed; B. Optimize control signaling packets The optimized control signaling data includes connection status announcement LSA, Keep_Alive packet, Ack_Alive packet, Inform_bandwidth packet, connection status update packet LSU, and connection status confirmation packet LSAck; The link state advertisement LSA is generated by the control node and the switching node to maintain information about neighbor node routing information and channel costs; The Keep_Alive packet is a detection message for determining whether a link has changed, and is sent between two adjacent satellite nodes; The Ack_Alive packet is sent by the satellite node that receives the Keep_Alive packet within a certain period of time, and is used to establish and maintain the adjacency relationship of the link; The Inform_bandwidth data packet is used to inform the control node of the remaining bandwidth of the current link and the port number of the satellite node through which the data packet passes; The link state update packet LSU is used to notify the control node to update the link state information; The link state confirmation packet LSAck is sent by the control node after updating its link state information, and is used to inform the source end that it has updated its link state information, and the data packet also contains the specific updated link state information; C. Design terminal capture and registration module operation mechanism 1) Satellite node sends broadcast; 2) After receiving the broadcast, the terminal sends a network access request to the satellite node; 3) After receiving the network access request, the satellite node forwards the network access request to the ground network management center; 4) The ground network management center analyzes and processes the terminal network access application information; 5) The ground network management center sends the network access success response information to the satellite node, and the satellite node forwards the network access success response information to the terminal; 6) The terminal receives the network access success response information, confirms the network access success, and sends the terminal network access confirmation information to the ground network management center; 7) After receiving the terminal network access confirmation information, the ground network management center confirms whether the terminal network access is successful or not and updates the local network status; 8) Upload network status to satellite nodes; 9) After completing the terminal capture and access, the new terminal has obtained a service channel and the identity of the control node. The new terminal sends a network access confirmation message to the control node on the obtained service channel. After receiving this message, the control node sends a terminal network access confirmation response message to the new terminal to confirm that the identity of the new terminal is valid and authorized, completing the registration process; D. Design the operation mechanism of real-time information release and collection module The master and slave controllers of each layer of the satellite network share information of each layer by configuring the following functions: 1) Atomic control logic installation to ensure the consistency of control logic; 2) Monitor the status of satellites in the domain in real time from the controller; 3) The main controller is responsible for collecting the global status of the on-board subnet to ensure that the main controller has the global network topology and status; 4) The main controller provides fast routing guarantee based on SRv6; 5) Priority-based rapid fault reporting and event transmission guarantee; The propagation of events between master and slave controllers adopts the publish / subscribe message mode. Each controller subscribes to the data channel, control channel and its own channel in the network and is granted the authority to publish and subscribe to information. The controller can publish the local network and application to the data channel, and publish a specific controller event and command in the corresponding control channel. Each controller periodically notifies local messages on the control channel to enable the controller to discover and detect faults. The publish / subscribe messaging model responsible for controller communication services provides the following services: 1) Provide persistent storage of published events to maintain the order of events published by the same controller; 2) immune network segmentation, each segment must run independently and synchronize when reconnected; 3) Reduce the cross-site traffic required to propagate events, i.e., the controller of one site should get updates from most other sites from nearby controllers to avoid congestion of cross-region connections; 4) Strengthen access control to ensure access ownership; E. Designing a path selection algorithm Suppose the topological structure of the satellite network in a certain time period is G(V,E,C), where V is a node set, element v∈V is a node of the graph G, n is the number of nodes in the network (n=|V|), E is a finite edge set, element (u,v)∈E is denoted as (u,v)=u→v is called an edge of the graph G, m is the number of links in the network (m=|E|), C represents the cost function between node i and node j defined on each edge in the time period, the initial value is determined by factors such as link available bandwidth and transmission delay, and the link residual capacity set is is the remaining capacity on the link (u,v), the element (S,D) is the source and destination node pair of the service flow, the LSP request is (S,D,bw), bw is the requested bandwidth; E1. Counting network resource information based on control nodes in the satellite network; E2. When requesting information to be sent between source / destination node pairs, first delete the links in the satellite network whose remaining bandwidth is less than the requested bandwidth or whose propagation delay is greater than the requested delay, and simplify the network structure to reduce the computational complexity of the algorithm; E3. When a service transmission request between source / destination nodes occurs, query the remaining bandwidth, link transmission delay, error rate and network topology resource information of each link in the satellite network; E4. Calculate the K shortest paths for each source / destination node pair based on the link remaining bandwidth and transmission delay. The steps are as follows: (1) According to the remaining capacity of the link, the initial weight of each link is set as: In the formula, is the remaining capacity of link (u,v), is the maximum value of the remaining capacity of the link in the current network link. The weight of the link with the largest remaining capacity in the network is 1, and the smaller the remaining capacity of the link, the larger the corresponding weight. D(u,v) is the transmission delay of the link (u,v), and D min is the minimum delay of all links, R(u,v) is the bit error rate of link (u,v), R min is the minimum bit error rate of all links, lg(R(u,v) / R min ) is the error rate of link (u,v), α1, λ1 and β1 are link weight adjustment factors; (2) Using the link weights obtained from formula (1), the Dijkstra algorithm is used to find the shortest path between the current source / destination node pair as the critical path. Before finding the next critical path, the link with the largest weight in the current path is deleted from the network to ensure that the next critical path will not be repeated with the previous path. (3) Repeat step (2) until there is no reachable path between the current source / destination node pair or the number of paths obtained is K, and K′ represents the actual number of critical paths between the source / destination node pair; Execute steps (1)-(3) for all source / destination node pairs to obtain the critical path set of all nodes; E5, Dynamically and adaptively adjust link weight After the critical path is determined, the weight of the critical path in the network needs to be reset. The weight of the critical path is defined as two parts: the initial weight and the conflict degree. The initial weight is formula (1), and the conflict degree is defined as: Where (s′, d′) is a source / destination node pair, (s, d) is the source / destination node pair requested by the current service flow, P is the set of all source / destination node pairs, p = |P| is the number of elements in the set P, is the i-th critical path of (s′,d′), express The importance of , λ2 is the symbol mark, when (s′, d′) is the other node pair of the current request node, λ2=1, when (s′, d′) is the current request node pair, λ2=-1; Set the weight of the critical path to: In the formula, C0(u,v), that is, formula (1), is the initial value of the link weight, which represents the contribution of the link remaining capacity, transmission delay and link error rate to the link weight. δ(u,v), that is, formula (2), represents the impact of the conflict on the link weight. α2 is the proportional factor used to adjust the contribution of the conflict to the link weight. Then it represents the increment of the link weight on the i-th path of (s′, d′). For non-critical paths, C(u, v) is equal to C0(u, v); E6. Calculate the route based on the adjusted weight, record the path and its occupied bandwidth, and establish a connection to transmit data.

Images