Space-space heterogeneous network cross-domain link selection and message distribution method

Abstract

The invention provides an air-space heterogeneous network cross-domain link selection and message distribution method, and belongs to the field of communication ad hoc networks. According to the method disclosed by the invention, the cross-domain path capability and the flow condition of each subnet are summarized through a heterogeneous network situation awareness technology, and the obtained network situation information provides a basis for subsequent link planning; carrying out capability evaluation on each path through a heterogeneous network cross-domain link selection technology, and selecting an optimal cross-domain path by comprehensively considering link time delay, packet loss and network load so as to meet the communication requirements of services and realize network load balancing; through a segment routing technology, protocol format conversion of cross-domain data packets is realized by routing nodes, and cross-domain message distribution is realized. According to the invention, interconnection and intercommunication of space and air heterogeneous networks can be realized, network overhead of message forwarding among the heterogeneous networks is reduced, meanwhile, limited network resources are fully utilized, load balancing is achieved, and an effective networking technical solution is provided for the fields of emergency rescue and the like.

Description

Technical Field

[0001] The present invention belongs to the field of communication ad hoc networks, and particularly relates to a method for cross-domain link selection and message distribution in an air-space heterogeneous network. Background Art

[0002] After a natural disaster occurs, basic communication facilities such as base stations or access points are extremely vulnerable to damage. The withdrawal of basic communication facilities will lead to the paralysis of the communication network in the affected area, causing great resistance to disaster relief and rescue work. It is difficult to repair the above basic communication facilities in a short time. Therefore, in case of an emergency, quickly and accurately deploying communication relays to form a temporary rescue communication network is a feasible and efficient method. By equipping handheld devices for rescue personnel and deploying aerial units to cover communication blind spots, a mobile ad hoc network can be quickly formed to provide communication guarantee for post-disaster reconstruction.

[0003] However, the construction of the above temporary communication network requires each communication unit to be equipped with the same communication basic equipment. In emergency scenarios such as disaster relief, communication units may come from all walks of life, and the communication equipment they are equipped with varies. It is not realistic to uniformly equip and deploy communication basic equipment for them in large-scale emergency scenarios. Therefore, communication units from different equipment manufacturers and service providers will build their own subnets. Communication within the subnets can be carried out without obstacles, while due to differences in basic configurations such as communication systems and routing protocols between subnets, it will be difficult to carry out efficient information interaction. The traditional method uses route redistribution for cross-subnet communication, but this method brings a large network overhead. At the same time, due to limited communication resources in emergency scenarios, after solving the basic problem of interoperability between heterogeneous networks, how to efficiently and reasonably utilize limited communication resources to achieve load balancing is also an important issue that needs to be considered.

[0004] In summary, developing an application technology that is easy to deploy and has scalability to achieve cross-subnet message circulation, realize the integration of heterogeneous networks, and adaptively adjust the forwarding path according to the network state to achieve load balancing has high practical value and is an important research task at present. Summary of the Invention

[0005] During the process of heterogeneous network integration, traditional methods for realizing interoperability between heterogeneous subnets under different communication systems have problems such as large network overhead and inability of the cross-domain forwarding path of data packets to adaptively change according to network situations and transmission requirements. To address these problems, the present invention proposes a method for cross-domain link selection and message distribution in an air-space heterogeneous network. By constructing a virtual backbone network to detect the situation of each subnet and plan the cross-domain link path, and realizing cross-domain message forwarding through segment routing technology, the integration of heterogeneous networks is achieved.

[0006] A method for cross - domain link selection and message distribution in an air - space heterogeneous network provided by the present invention is applied to a network scenario composed of multiple heterogeneous sub - networks. The method of the present invention first selects gateway nodes in the heterogeneous network to construct a virtual backbone network, equips the protocol stacks of each subnet where the gateway nodes are located, and then deploys a functional module for implementing the method of the present invention between the application layer and the transport layer in the TCP / IP protocol stack of the gateway nodes. The method of the present invention includes the following steps:

[0007] Step 1: Perform subnet situation awareness, including traffic detection, delay detection, and packet loss rate detection. Obtain the traffic of data packets sent within each subnet through traffic detection, and obtain the link - level delay and packet loss rate between adjacent sub - networks through delay detection and packet loss rate detection; perform subnet traffic prediction, including dividing each subnet into blocks, and using a neural network to predict the traffic of each subnet. The input of the neural network is the historical traffic matrix of the subnet, and the output is the predicted future traffic matrix, and the traffic between each block of the subnet is recorded in the matrix;

[0008] Step 2: Divide the cross - domain forwarding of data packets into the selection of gateway nodes to pass through and the forwarding of nodes within the sub - network to pass through; the gateway node of the subnet where the source node of the data packet is located plans a cross - domain link for the data packet according to the maintained virtual backbone network topology, determines the next intermediate gateway node, and the data packet is forwarded after format conversion according to the routing protocol of the subnet to be transferred when it is transferred at each gateway node;

[0009] Step 3: When there are more than 1 cross - domain forwarding paths for the cross - domain forwarding of data packets, score each cross - domain forwarding path through the detected link - level delay and packet loss rate between sub - networks and the predicted traffic matrix of the sub - network, and select the path with the highest score as the cross - domain forwarding path of the data packet.

[0010] The subnet situation awareness in step 1 includes: (1) Traffic detection: Nodes in each subnet regularly obtain the data packet traffic sent by their own application layer and the target node addresses of the data packets, and report them to the gateway node of the subnet. (2) Delay detection: Detect the link-level delay between logically adjacent gateway nodes; each gateway node maintains a detection table to store the latest time for sending data packets to the target gateway nodes that are logically adjacent; when the sending time of a certain target gateway node in the table exceeds the set duration, the gateway node actively sends a delay maintenance data packet to the target gateway node to measure the link-level delay; when a data packet arrives at a certain target gateway node at the gateway node, a frame is added to the packet header of the data packet to record the sending time of the packet. When the data packet is forwarded to the target gateway node, the target gateway node calculates the link-level delay from the previous-hop gateway node to this node; Logically adjacent gateway nodes refer to two nodes that only span 1 subnet from each other. (3) Packet loss rate detection: Each gateway node counts the data packets generated and arriving at each logically adjacent gateway node, and counts the data packets received from each logically adjacent gateway node. Each gateway node regularly sends an ACK message to each logically adjacent gateway node. The message records the total number of data packets received from the other party by this node in the previous cycle, and at the same time restarts the data packet counting process; after each gateway node receives the ACK message, it calculates the packet loss rate.

[0011] The gateway node of the method of the present invention plans and selects the cross-domain forwarding paths for different target subnets in advance in each cycle through a periodic update method, generates a routing table for storage, and directly queries the routing table to obtain the cross-domain forwarding path whenever a data packet is sent; when the cycle is updated, the cross-domain forwarding paths for different target subnets are re-planned and selected, and the routing table is updated.

[0012] Compared with the prior art, the advantages and positive effects of the method of the present invention are as follows:

[0013] (1) The method of the present invention aggregates the cross-domain path capabilities and traffic conditions of each subnet through the virtual backbone network composed of gateway nodes, detects the network operation situation, and provides a data basis for subsequent link planning; through the segment routing technology, the cross-domain forwarding of data packets is divided into the selection of gateway nodes to pass through and the forwarding within several subnets, realizing the interconnection and interoperability between heterogeneous subnets operating under different protocol stacks with low network overhead; when the network situation changes, a path adaptive adjustment mechanism is introduced to select the optimal path in the current situation to ensure the cross-domain circulation of service messages.

[0014] (2) The method of the present invention realizes the protocol format conversion of cross - domain data packets through the gateway node, significantly reducing the network overhead of cross - domain message distribution; through the cross - domain link selection technology, according to the service requirements and the actual network situation, the cross - domain forwarding link is flexibly selected, and the cross - domain path configuration of the data packet is optimized pertinently; through the joint optimization of network situation awareness and cross - domain link selection for path configuration, when network congestion or anomalies occur, the path configuration can be changed in a timely manner according to the network situation, realizing the load balancing of limited network resources.

[0015] (3) The method of the present invention significantly reduces the network overhead of cross - domain message distribution and realizes the load balancing of limited network resources, having important theoretical significance and wide application value. The present invention is applicable to the cross - domain message forwarding task of the space - air heterogeneous mobile ad - hoc network, providing an effective networking technology solution for fields such as emergency rescue. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is the implementation architecture diagram of the cross - domain link selection and message distribution method for the space - air heterogeneous network of the present invention;

[0017] Figure 2 is the schematic diagram of deploying the cross - domain link selection and message distribution method for the space - air heterogeneous network of the present invention in the scenario for application;

[0018] Figure 3 is the specific scenario illustration diagram of measuring the cross - domain link delay of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In order to make the purpose, technical solutions and advantages of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0020] As Figure 1 shown, the implementation of the cross - domain link selection and message distribution method for the space - air heterogeneous network of the present invention mainly includes two parts, namely the heterogeneous network situation awareness technology and the heterogeneous network cross - domain link selection and message distribution technology. Among them, the heterogeneous network situation awareness technology will build a virtual backbone network to undertake the situation awareness task, providing support for cross - domain link selection and message distribution; the heterogeneous network cross - domain link selection and message distribution technology will also be deployed in the backbone network nodes, which select the forwarding link for the cross - domain data packet according to the situation awareness information and build a cross - domain path through the segment routing technology. The method of the present invention realizes the interconnection and intercommunication of the space - air heterogeneous network, reduces the network overhead of message forwarding between heterogeneous networks, and at the same time realizes the full utilization of limited network resources, achieving load balancing.

[0021] As Figure 2 shown, a deployment scenario of an embodiment of the present invention is jointly composed of five heterogeneous sub - networks A, B, C, D, and E. The gateway node selected in the cross - sub - network node is asFigure 2 The red nodes form a virtual backbone network on the Internet. The method of the present invention is implemented as application software or a device and embedded between the application layer and the transport layer in the TCP / IP protocol stack of the virtual backbone network nodes, undertaking the situation awareness of each subnet and the cross-domain link planning task to achieve the management of the entire network.

[0022] The heterogeneous network situation awareness technology of the method of the present invention mainly includes three parts: cross-domain virtual backbone network construction, subnet situation detection, and subnet traffic prediction. The virtual backbone network, as the main body, undertakes the situation awareness task of the entire network and is responsible for subsequent link planning; the subnet situation awareness statistically analyzes the cross-domain link performance of each subnet; the subnet traffic prediction statistically analyzes the traffic level within the subnet.

[0023] Step 1.1: Construct a virtual backbone network. To achieve heterogeneous network situation awareness, several nodes in the network need to be selected to undertake the awareness task. The gateway nodes are selected from each subnet and are interconnected through multi-hop links within the subnet to jointly form a virtual backbone network to undertake responsibilities. The selection and maintenance rules for gateway nodes are as follows: Select the node with the highest connectivity in the node intersection of adjacent subnets as the channel between two adjacent subnets, that is, the gateway node. The gateway node is located in multiple subnets at the same time, equipped with the protocol stacks of each subnet, capable of communicating with each subnet where it is located, and undertaking the responsibility of cross-subnet communication. The gateway node periodically notifies the other nodes in each subnet where it is located of its gateway identity.

[0024] Step 1.2: Subnet situation awareness. Heterogeneous subnets operate in a complex environment, and their overall situation is constantly changing. To achieve network load balancing during high communication traffic and improve the overall network throughput, it is first necessary to collect information on the operating conditions of the network. For heterogeneous subnet information, it is mainly divided into two parts: node attributes and edge attributes for collection. Node attributes are the traffic generation situation of the node in the previous period; edge attributes are the edge delay and packet loss rate. Among them, the node traffic situation is used as the original traffic data for subsequent traffic prediction; the multi-hop link delay between gateway nodes is used as the subsequent speed index for evaluating cross-domain subnets; the multi-hop link packet loss rate between gateway nodes is used as the subsequent reliability index for evaluating cross-domain subnets. In the embodiment of the present invention, the subnet situation awareness mainly includes the following three aspects.

[0025] (1) Traffic Detection: Due to limited network resources, in order to allocate communication resources to service traffic as much as possible, the principle of minimizing the generated measurement traffic should be adhered to when detecting network information. Since the module designed in the present invention is in the middle layer of the communication protocol stack of the gateway node, when a data packet is generated at the application layer, the module can obtain the target node of the data packet. In the embodiment of the present invention, the period is set to 1 s. When a data packet is sent from the application layer, the functional module of the present invention records all the data packets sent by the application layer within 1 s from the current moment and their target node addresses. Nodes within the subnet report the currently recorded traffic and target addresses to the gateway node within the subnet by reporting periodically to the gateway node within the subnet, so as to provide raw data for the subsequent generation of the traffic matrix.

[0026] (2) Delay Detection: In order to minimize the generation of measurement traffic as much as possible, service traffic should be fully utilized. When a service data packet generated by a node in the subnet arrives at another gateway node, a frame is added to the packet header to record the sending time of the packet. When the data packet is forwarded to the target gateway node, the target gateway node can calculate the link-level delay from the previous-hop gateway node to this node by obtaining the time stamp. However, due to the randomness of service packet generation and the diversity of sending paths, there may be no service packet exchanges between two gateway nodes for a long time. To solve the link delay detection between gateway nodes with no service exchanges for a long time, each gateway node within the subnet should maintain a data packet sending list, storing the data packet sending situations between itself and the other gateway nodes in each subnet it belongs to. The table entry includes the addresses of the other gateway nodes and the latest time of sending data packets to the corresponding gateway nodes. For example, Figure 3 the gateway node 1 in the intersection of Network A and Network B stores the sending situations of the two gateway nodes that are logically adjacent to it. Nodes are logically adjacent means that nodes only cross 1 subnet with each other. Taking Figure 3 the scenario as an example, gateway node 1 is logically adjacent to gateway nodes 2 and 3 respectively, but gateway nodes 2 and 3 are not logically adjacent. When there is no communication between gateway node 1 and its adjacent gateway nodes for more than the set duration PacketTIME, gateway node 1 will actively initiate a delay maintenance data packet to measure the link delay.

[0027] (3) Packet loss rate detection: Introduce a timing feedback mechanism. When a data packet is generated at a gateway node and reaches other gateway nodes within the subnet (i.e., logically adjacent gateway nodes), the node starts counting, and the count is incremented by 1 for each generated data packet. At the same time, as a gateway node, when receiving data packets from the remaining gateway nodes within the subnet, it should also perform counting. Whenever the time reaches the set duration PacketTIME, all gateway nodes send an ACK (acknowledgment character) message to the gateway nodes logically adjacent to themselves. The message records the total number of data packets received from the other party by this node during the previous period, and at the same time restarts the data packet counting process. When a gateway node receives an ACK message from its adjacent gateway node, it can learn about the reception status of the data packets it has sent and calculate the packet loss rate. Thus, the transmission success rate statistics of the link across subnets, that is, the link between gateway nodes, are achieved.

[0028] Step 1.3: Subnet traffic prediction. Through the subnet detection method, the situation information of the entire network has been successfully perceived. When the heterogeneous network operates in a high-load environment, the method of constructing a static cross-domain transfer path will cause all cross-domain traffic to be poured onto the pre-set cross-domain path, resulting in congestion and even packet loss. In order to plan the path of cross-domain data packets, it is necessary to predict the future traffic situation in each subnet based on the detected traffic information to plan the cross-domain forwarding path, thereby achieving load balancing and improving network throughput. For this reason, the present invention designs a subnet traffic prediction algorithm: First, divide each subnet into blocks through a graph partitioning algorithm. Second, use the detected subnet traffic information to generate a traffic matrix. Finally, deploy a neural network within each subnet, using the historical traffic matrix as input to predict the future traffic matrix. This process mainly includes the following steps:

[0029] Step 1.3.1. Graph partitioning of the entire network topology: Use the LDG (Linear Deterministic Greedy) algorithm to partition adjacent nodes in each subnet, and divide a single subnet into several blocks. The LDG algorithm considers placing neighbor nodes together as much as possible during partitioning to reduce the cutting of edges in the network topology. It uses a greedy algorithm to place a node in the subgraph with the most neighbors, while ensuring the node load balance of each subgraph.

[0030] Step 1.3.2. Traffic matrix generation: Through the graph partitioning algorithm, assume that the topology within the subnet is divided into k blocks, then the traffic matrix is a k-dimensional square matrix. Let the traffic matrix be TM, and its composition is:

[0031]

[0032] where throughput ijDenote the total traffic from the $i$-th block to the $j$-th block, and its calculation formula is:

[0033]

[0034] where $N$ i and $N$ j are the numbers of nodes in blocks $i$ and $j$ respectively, denotes the traffic from the $n$-th i node in block $i$ to the $n$-th j node in block $j$.

[0035] Step 1.3.3. Network traffic prediction: Use a GRU (Gated Recurrent Unit) neural network for time series prediction. Train the prediction model with historical measurement data, take the measurement traffic matrix of the previous second as the input, and predict the traffic matrix of the next second.

[0036] After the above steps, the trained subnet traffic prediction model can be used to predict the future network traffic situation, and it can be continuously corrected during the prediction process, which helps with cross-domain link planning and achieving network-wide load balancing.

[0037] In the heterogeneous network cross-domain link selection and message distribution technology, it mainly includes cross-domain segmented routing technology and cross-domain path selection technology. The cross-domain segmented routing algorithm of the present invention runs between the application layer and the network layer, avoiding operations on network layer protocols and adapting to various heterogeneous networks. The cross-domain path selection technology targets the differences in cross-domain service communication requirements, differentially adapts to the requirements of service data based on the aforementioned subnet situation awareness and traffic prediction results, and at the same time introduces an adaptive dynamic adjustment mechanism to avoid network congestion and achieve load balancing.

[0038] In heterogeneous subnets, each heterogeneous subnet runs under different routing protocols. For example, the ground backbone network uses a wired network routing protocol, and the air-based data link network uses an ad hoc network routing protocol, etc. It is difficult for heterogeneous subnets under different communication systems to communicate with each other across networks. The method of the present invention designs cross-domain segmented routing technology, which divides the cross-domain forwarding of data packets into the selection of gateway nodes to pass through and several segments of in-subnet forwarding, achieving interconnection and interoperability between heterogeneous subnets with low network overhead. The cross-domain segmented routing process of data packets in the method of the present invention includes the following steps 2.1 to 2.5.

[0039] Step 2.1. Send a data packet to the gateway: The node in the source subnet sends the data packet to the gateway node in this subnet.

[0040] Step 2.2. The gateway node processes the data packet: After receiving the data packet, the gateway node unpacks the data packet to the upper layer, and finds that its real target node is in other subnets. According to the backbone network topology it maintains, it plans a cross-domain link for it, determines the next intermediate gateway node, converts the format of the data packet according to the new subnet routing protocol, and sends the data packet to the next intermediate gateway node through the new subnet.

[0041] Step 2.3. Multi-hop forwarding of data packets within the subnet: After being sent by the gateway node, it is relayed within the subnet and the data packet is sent to the next planned gateway node.

[0042] Step 2.4. Repeat steps 2.2 and 2.3 and forward to the gateway node of the subnet where the target node is located.

[0043] Step 2.5. Subnet processing of data packets within the target subnet: The gateway node of the target subnet forwards the data packet to the target node to complete the segmented routing of the data packet.

[0044] By introducing the segmented routing operation above the network layer, the routing process of data packets across subnets is converted into the selection of gateway nodes along the way and the routing process within each subnet along the way, which can avoid the flooding of a large number of non-service data packets caused by route redistribution, and only need to synchronize the necessary cross-domain transfer information between the gateway nodes of each subnet.

[0045] When there are multiple alternative cross-domain paths for cross-domain data packets, the method of the present invention evaluates the capabilities of each path through the performance and traffic conditions of each subnet cross-domain path obtained by subnet situation detection, and selects the optimal cross-domain path by comprehensively considering link delay, packet loss and network load to meet the communication requirements of services and achieve network load balancing. When the network situation changes, a path adaptive adjustment mechanism is introduced to select the optimal path under the current situation to ensure the cross-domain circulation of service messages.

[0046] Based on the above process, the link-level delay and packet loss rate of each cross-domain multi-hop link within each subnet have been measured, and the traffic matrix of the subnet in the next time period has been predicted using an intelligent algorithm. Let the average cross-domain link delay and packet loss rate of the jth subnet obtained from subnet situation awareness and traffic prediction be used as indicators of subnet transmission rate and reliability respectively. At the same time, the elements of the predicted traffic matrix are summed up and normalized using the nominal capacity W of the subnet itself to obtain the subnet load evaluation index W j , and its calculation formula is as follows:

[0047]

[0048] where W j is the load evaluation index of the jth subnet, k is the total number of blocks in the jth subnet, throughput xyis the element located in the x-th row and y-th column of the traffic matrix of the j-th subnet.

[0049] When there are multiple cross-domain paths from the source subnet to the destination subnet of a cross-domain data packet, each path should be scored separately, and the path with the highest score should be selected as the cross-domain transfer path of the message. For the i-th path, obtain the delay and packet loss rate of each segment of the path. Each segment of the path corresponds to two network gateway nodes: the sending gateway and the receiving gateway; the delay of the segment is obtained by the sending gateway detecting the link-level delay to the receiving gateway, and the packet loss rate of the segment is obtained by the receiving gateway detecting the packet loss rate from the sending gateway to the receiving gateway; calculate the load evaluation index of the subnet where each segment of the path is located. In an embodiment of the present invention, the average link-level delay and average packet loss rate of multiple recent periods can also be obtained as the delay and packet loss rate of the corresponding segment of the path.

[0050] The score calculation formula for the i-th path is as follows:

[0051]

[0052] where delay i,j represents the delay of the j-th segment of the i-th path, represents the total delay of the cross-domain path i; P i,j represents the packet loss rate of the j-th segment of the cross-domain path i, represents the total packet loss rate of the cross-domain path i; W i,j represents the traffic of the subnet passed by the j-th segment of the cross-domain path i; ω1 and ω2 are set weights, and the value range is 0 to 1. In an embodiment of the present invention, taking the path "network gateway node 1 - network gateway node 2 - network gateway node 3" as an example to illustrate path segmentation, the first segment is "network gateway node 1 - network gateway node 2", and the second segment is "network gateway node 2 - network gateway node 3". In an embodiment of the present invention, there may be multiple possible transfer paths for each path, that is, the source network gateway node and the destination network gateway node are the same, but the specific transfer network gateway nodes are not exactly the same, but there are some identical transfer network gateway nodes. Such two paths are considered to belong to one path. Therefore, for each path, respectively calculate the minimum delay and minimum packet loss rate among all possible transfer paths of path i as the delay and packet loss rate of path i. is the maximum value of the subnet load evaluation index in the i-th path.

[0053] The method of the present invention updates the cross-domain path selection in a periodic manner. At each cycle, the gateway node can plan the cross-domain path and generate a routing table entry for storage. Whenever a data packet needs to be sent to a specified target subnet, it can be directly queried from the routing table. When the timing reaches the update cycle, the scores of each path are recalculated to update the routing table, thereby realizing the adaptive adjustment of the cross-domain path. In addition to the periodic correction, when network traffic prediction senses that a certain subnet is about to experience congestion, it can also trigger path correction to avoid network congestion.

[0054] In summary, the present invention successfully addresses the problem of heterogeneous network networking in various situations, and proposes and implements the cross-domain link selection and message distribution technology for heterogeneous networks. This technology can effectively achieve information circulation between heterogeneous subnets, and through subnet situation detection for assistance, plan paths for cross-domain message forwarding. Using the method of the present invention does not require synchronization of routing tables between subnets, and the size of network maintenance messages is smaller, which can reduce the cross-domain network communication overhead. At the same time, it can adaptively adjust the path according to the actual needs of the task and the actual situation of the current network to achieve load balancing of the whole network resources.

[0055] The method of the present invention aims to solve the problems of difficult communication between heterogeneous data links with different routing protocols and network congestion caused by traditional pre-configured fixed cross-domain chain routing. Through subnet situation awareness technology, cross-domain segment routing technology, cross-domain link planning and adaptive adjustment technology, it realizes heterogeneous network fusion and load balancing. The method of the present invention realizes a heterogeneous network platform fusion communication that constructs an online backbone network, plans cross-domain links through backbone network nodes, and relays data packets for cross-chain transmission, thereby realizing load balancing. It can effectively solve the problem of difficult communication between different subnets in the process of space-air heterogeneous networking. At the same time, it can achieve load balancing of the whole network resources through adaptive adjustment of cross-domain forwarding paths, which has important practical application value. This method can reduce the routing maintenance overhead while ensuring the interconnection and interoperability of heterogeneous networks, and improve the utilization efficiency of network resources. In addition, the modular design and scalability of this method enable it to be customized and extended according to different application scenarios and requirements, and have broad application potential.

[0056] Except for the technical features described in the specification, they are all known technologies to those skilled in the art. The present invention omits the description of well-known components and well-known technologies to avoid redundancy and unnecessary limitation of the present invention. The implementation manners described in the above embodiments do not represent all implementation manners consistent with the present application. On the basis of the technical solution of the present invention, various modifications or deformations that can be made by those skilled in the art without creative labor are still within the protection scope of the present invention.

Claims

1. A cross-domain link selection and message distribution method for an air-space heterogeneous network, which is applied to a network composed of more than one heterogeneous subnet, and is characterized in that This method first selects gateway nodes in the heterogeneous network to construct a virtual backbone network, equips the protocol stacks of each subnet where the gateway nodes are located, and then deploys a functional module that implements the following steps between the application layer and the transport layer in the TCP / IP protocol stack of the gateway nodes: Step 1: Perform subnet situation awareness, including traffic detection, delay detection, and packet loss rate detection. Obtain the traffic of the data packets sent within each subnet through traffic detection, and obtain the link-level delay and packet loss rate between adjacent subnets through delay detection and packet loss rate detection; perform subnet traffic prediction, including dividing each subnet into blocks, and using a neural network to perform traffic prediction on each subnet. The input of the neural network is the historical traffic matrix of the subnet, and the output is the predicted future traffic matrix, and the traffic between each block of the subnet is recorded in the matrix; Step 2: Divide the cross-domain forwarding of data packets into the selection of the gateway nodes to pass through and the forwarding of the nodes within the subnets to pass through; the gateway node of the subnet where the source node of the data packet is located plans the cross-domain link for the data packet according to the virtual backbone network topology maintained, determines the next intermediate gateway node, and the data packet is forwarded after format conversion according to the routing protocol of the subnet to be transferred when it is transferred at each gateway node; Step 3: When there are more than 1 cross-domain forwarding paths for the cross-domain forwarding of data packets, score each cross-domain forwarding path through the detected link-level delay and packet loss rate between subnets and the predicted traffic matrix of the subnets, and select the path with the highest score as the cross-domain forwarding path of the data packet.

2. The method according to claim 1, wherein In the method described above, the rules for selecting and maintaining gateway nodes are: select the node with the highest connectivity in the intersection of the communication nodes of adjacent subnets as the gateway node. The gateway node is located in multiple subnets at the same time, is equipped with the protocol stacks of each subnet, and can communicate with each subnet where it is located; the gateway node periodically notifies the other nodes in each subnet where it is located of its gateway identity.

3. The method according to claim 1, wherein In the described Step 1 for performing subnet situation awareness, it includes: Traffic detection: The nodes within each subnet regularly obtain the traffic of the data packets sent by their own application layer and the destination node address of the data packets, and report them to the gateway node of the subnet; Delay detection: Detect the link-level delay between logically adjacent gateway nodes; each gateway node maintains a detection table to store the latest time for sending data packets to the target gateway node that is logically adjacent; if the sending time of a certain target gateway node in the table exceeds the set duration, the gateway node actively sends a delay maintenance data packet to the target gateway node to measure the link-level delay; when the gateway node generates a data packet reaching a certain target gateway node in the table, add a frame to record the sending time of the packet to the packet header, and when the data packet is forwarded to the target gateway node, the target gateway node calculates the link-level delay from the previous-hop gateway node to this node; Logically adjacent gateway nodes refer to two nodes that only cross 1 subnet with each other; Packet loss rate detection: Each gateway node counts the data packets generated to each logically adjacent gateway node and counts the data packets received from each logically adjacent gateway node. Each gateway node periodically sends an ACK message to each logically adjacent gateway node. This message records the total number of data packets received from the other party at this node in the previous cycle, and at the same time restarts the data packet counting process; after each gateway node receives the ACK message, it calculates the packet loss rate.

4. The method according to claim 1 or 3, characterized in that, The subnet traffic prediction in step 1 includes: First, divide each subnet into blocks using the LDG algorithm; Second, represent the traffic between each subnet block as a traffic matrix. Suppose the topology within the subnet is divided into k blocks, then the traffic matrix TM of this subnet is a k-dimensional square matrix, and the matrix is represented as follows: Among them, throughput ij represents the total traffic from the i-th block to the j-th block, and the calculation formula is: N i and N j are the number of nodes in the i-th block and the j-th block respectively, is the traffic from the n i -th node in the i-th block to the n j -th node in the j-th block; Then, use the GRU neural network as the network traffic prediction model, input the historical traffic matrix of the subnet into the GRU neural network in chronological order, and continuously predict and output the traffic matrix of the subnet at the next moment; GRU represents the gated recurrent unit.

5. The method according to claim 1, wherein In step 2, the cross-domain forwarding of data packets includes: Step 2.1: The source node of the data packet sends the data packet to the gateway node of the subnet where it is located; Step 2.2: The gateway node unpacks the received data packet to obtain the target node of the data packet. According to the subnet where the target node is located and the virtual backbone network topology maintained, plan a cross-domain link for the forwarding of the data packet, determine the next intermediate gateway node, convert the format of the data packet according to the subnet routing protocol of the next transfer, and then send it to the next intermediate gateway node; Step 2.3: The data packet is relayed and forwarded within the subnet and sent to the next intermediate gateway node determined in step 2.2; Step 2.4: Repeat steps 2.2 and 2.3 to forward the data packet to the gateway node of the subnet where the target node is located; Step 2.5: The gateway node of the subnet where the target node is located forwards the data packet to the target node, completing the segmented routing of the data packet.

6. The method according to claim 1, characterized in that In step 3, scoring each cross-domain forwarding path includes: For the i-th path, obtain the delay and packet loss rate of each segment of the path. Each segment of the path corresponds to two gateway nodes: the sending gateway and the receiving gateway; the delay of this segment of the path is detected by the sending gateway to the link-level delay of the receiving gateway, and the packet loss rate of this segment of the path is detected by the receiving gateway from the sending gateway to the receiving gateway; calculate the load evaluation index of each subnet where each segment of the path is located: by adding up the elements in the predicted traffic matrix and then normalizing with the nominal capacity of the subnet itself, this index is obtained; Then calculate the score of the i-th path where delay i,j represents the delay of the j-th segment of the i-th path, represents the total delay of the i-th path; P i,j represents the packet loss rate of the j-th segment of the i-th path, represents the total packet loss rate of the i-th path; W i,j represents the load evaluation index value of the subnet passed by the j-th segment of the i-th path; ω1 and ω2 are the set weights, and the value range is 0 to 1.

7. The method according to claim 1 or 2, characterized in that In the method described above, the gateway node plans and selects the cross-domain forwarding paths for different target subnets in advance in each cycle through a periodic update method, generates a routing table for storage, and directly queries the routing table to obtain the cross-domain forwarding path whenever a data packet is sent; when the cycle is updated, re-plan and select the cross-domain forwarding paths for different target subnets and update the routing table.

Images