A multi-path transmission method and system based on multi-objective optimization

Abstract

The application provides a multi-path transmission method and system based on multi-objective optimization, and belongs to the technical field of transmission path optimization.The system is composed of a data acquisition module, a path planning module, a multi-objective optimization module and a multi-path calculation module.The data acquisition module acquires the topological structure and node link information of a network, the path planning module calculates a feasible path scheme according to the source and destination nodes set by a transmission task, the multi-objective optimization module calculates a Pareto optimal solution satisfying multiple objectives by using an NAGA2 algorithm according to an optimization target, and the solution is used as the first path.Then, the multi-path calculation module reconstructs a topological graph according to the first path, and calculates a second path by using the path planning module and the multi-objective optimization module.The multi-path transmission method and system based on multi-objective optimization have the advantages of providing redundant transmission satisfying multiple target optimizations, and effectively improving the reliability of information transmission.

Description

Technical Field

[0001] This invention belongs to the field of path transmission optimization technology, and in particular relates to a multi-path transmission method and system based on multi-objective optimization. Background Technology

[0002] Heterogeneous networks, as an important research direction in the field of networking, have shown vigorous development momentum. Heterogeneous networks are networks composed of different types of devices, communication technologies, or network structures. Their research scope is typically very broad, involving satellite networks, unmanned aerial vehicle (UAV) ad hoc networks, mobile communication networks, and wired internet, among others. The goal of building heterogeneous networks is to provide users with more flexible, efficient, and reliable communication services by integrating various network resources, thereby adapting to increasingly diverse and complex communication needs.

[0003] However, heterogeneous networks currently have poor resilience to risks, and various attack methods from the information domain and physical domain are increasing. These threats can damage nodes and links in heterogeneous networks and interrupt transmission. After critical nodes are damaged, the remaining network resources are fragmented and cannot independently form a complete network connection. The single-link transmission communication method of conventional networks cannot ensure the reliable transmission of information.

[0004] In heterogeneous network environments such as satellite networks, wired internet, and mobile communication networks, multipath transmission protocols utilize multiple available paths to strive to meet strict QoS constraints in order to achieve higher data rates, lower network latency, and higher reliability. To better apply multipath transmission to heterogeneous network transmission, researchers have proposed different methods to improve transmission reliability.

[0005] However, the traditional Transmission Control Protocol (TCP) does not provide multihoming. Applications can only attach a single IP address to a specific TCP connection with another host. If the TCP socket-based application programming interface (API) connected to that IP address fails, the TCP connection must be re-established because it has been lost. To address this, the Internet Engineering Task Force (IETF) standardized the Flow Control Transport Protocol (SCTP), integrating multihoming functionality into its specification. Then, Iyengar et al. designed the Concurrent Multipath Transfer (CMT) method for multi-network interface devices to leverage this feature of SCTP, improving throughput, resource utilization, latency, and network reliability. However, CMT-SCTP has serious drawbacks: inappropriate packet scheduling, unnecessary packet retransmissions, unnecessary congestion window reduction (CWND), and receiver buffer blocking (RBB). Therefore, the IETF's Multipath TCP Working Group introduced Multipath (MPTCP), which allows TCP connections to use multiple paths to utilize resources and enhance redundancy. MPTCP allows for the dynamic establishment of multipath connections containing multiple streams, enabling simultaneous data transmission through different accessible substreams (including IPv4 and IPv6).

[0006] In Westwood SCTP (W-SCTP), SCTP enhancements were proposed to achieve load balancing across multiple interfaces using bandwidth-aware scheduling. Subsequently, Load Sharing SCTP (LS-SCTP) was proposed to aggregate bandwidth and maintain load balancing across multiple active transmission paths. Lu proposed another extension of SCTP to mitigate packet loss in lossy environments; this algorithm also limits redundant data transmission on different paths to minimize network congestion. The authors proposed a Selective Redundancy Multipath Transmission (SRMT) strategy, which uses the primary path to transmit data and the secondary path to transmit redundant data. Huang, C et al. proposed a Wireless Multipath SCTP (WiMP-SCTP), which aggregates bandwidth in a wireless environment through data striping and aggregation and uses it to transmit duplicated data. Shailendra et al. proposed an Efficient SCTP Multipath Scheme (MPSCTP) that can transmit packets simultaneously through multiple paths. MPSCTP is a solution for packet reordering and invalid CWND growth. Shailendra et al. proposed a Resource Pooling (BERP) CC algorithm based on bandwidth estimation, which enhances the throughput and performance of multipath communication. Furthermore, MPSCTP was later improved based on full path latency, including the data transmission rate on each channel. Therefore, this approach minimizes block latency across different channels, but due to its uniform bandwidth-sharing strategy, it still falls short in terms of available bandwidth utilization.

[0007] Based on the current research status, it is found that although researchers have discussed and analyzed many problems in multipath algorithms, these studies are mostly from the perspective of protocols, studying issues such as congestion and scheduling during transmission. However, in the design, multipath transmission algorithms often avoid the inherent characteristics of heterogeneous networks and focus more on optimizing and improving protocols at the transport layer. If a multi-level collaborative optimization transmission algorithm can be implemented based on the characteristics of heterogeneous networks, it will be more adaptable to the complexity of heterogeneous network environments and closer to the needs of real-world scenarios. Summary of the Invention

[0008] This invention proposes a multipath transmission scheme based on multi-objective optimization.

[0009] The technical problem this invention aims to solve is that it studies the problems of blocking and scheduling in the current multi-path algorithm's multi-pair transmission process, but does not propose different path algorithms at the network layer to realize a multi-level collaborative multi-path transmission algorithm.

[0010] The first aspect of this invention discloses a multi-path transmission method based on multi-objective optimization, the method comprising:

[0011] Step S1: Call the data acquisition module to obtain network traffic data and topology data from the real heterogeneous network in order to extract network topology information and node link status.

[0012] Step S2: Invoke the path planning module. Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set.

[0013] Step S3: Call the multi-objective optimization module to optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, so as to determine the optimal path as the first optimal transmission path.

[0014] Step S4: Call the multipath calculation module, based on the first optimal transmission path and the network topology information, to reduce the network topology and obtain a simplified network topology.

[0015] Step S5: Call the path planning module again to form a new set of feasible paths based on the simplified network topology; call the multi-objective optimization module again to determine a new optimal path using the new set of feasible paths, which will serve as the second optimal transmission path.

[0016] The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability.

[0017] According to the method of the first aspect of the present invention, in step S1: the data acquisition module is invoked, and the request tool is used to acquire traffic data and topology data from the network every 5 seconds, and further acquire the link status and link connectivity status of each node, thereby extracting the network topology information and the node link status.

[0018] According to the method of the first aspect of the present invention, in step S2:

[0019] The path planning module is invoked, which utilizes a variant of the depth-first search algorithm to recursively probe all paths in the network based on the source and destination nodes to form a set of feasible paths, while ensuring that each path in the set of feasible paths has no duplicate vertices.

[0020] Specifically, during the exploration process, a variant algorithm based on depth-first search is used to record the length of the current path, and the search stops when the path length exceeds the known maximum path length, in order to reduce unnecessary computation.

[0021] According to the method of the first aspect of the present invention, in step S3, a multi-objective optimization module is invoked to calculate the optimal path using the NSGA-II algorithm, wherein:

[0022] The NSGA-II algorithm is an algorithm for multi-objective optimization problems based on genetic algorithms and the concept of Pareto optimality. It utilizes the global search capability of the NSGA-II algorithm to avoid getting trapped in local optima during the iteration process.

[0023] If the optimization objectives are to maximize path reliability and minimize transmission latency, then the objective function is expressed as:

[0024] minF=min{Rreliability max delay min}

[0025] minF = w1D + w2(-R)

[0026] stD=∑t i ·e i

[0027] R = Πr i ·v i

[0028] R≥0.998

[0029] D≤500

[0030] Where: R is the reliability of the transmission path, calculated by multiplying the reliability of each node; D is the delay of the transmission path, calculated by summing the delays of each link.

[0031] After determining the objective function, the initial population obtained by the path planning module is subjected to non-dominated sorting and selection, crossover, mutation, constraint checking and repair to generate the first generation of offspring population and update the evolutionary generation.

[0032] For each generation, fast non-dominated sorting, crowding calculation, and elite strategy operations are performed until the maximum number of generations. The NSGA-II algorithm outputs the Pareto front solution, which is the first optimal transmission path in the current situation.

[0033] According to the method of the first aspect of the present invention, in step S4, a first type of deletion is performed on the network topology; specifically including:

[0034] Find the sub-networks where the source node and destination node are located, and find the sub-networks where each node other than the source node and destination node is located on the first optimal transmission path;

[0035] Determine whether the subnet where the node is located is the same as the subnet where the source node is located or the subnet where the destination node is located;

[0036] If they are the same, delete the node and the links associated with it; otherwise, delete all nodes in the subnetwork where the node is located and the links associated with each node in the subnetwork.

[0037] In this process, after performing the first type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the second type of deletion is performed.

[0038] According to the method of the first aspect of the present invention, in step S4, a second type of deletion is performed on the network topology; specifically, it includes: deleting each link on the first optimal transmission path in the network topology; wherein:

[0039] After performing the second type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the third type of deletion is performed.

[0040] According to the method of the first aspect of the present invention, in step S4, a third type of deletion is performed on the network topology; specifically including:

[0041] Based on the links on the first optimal transmission path, find the nodes on the path that have only 2 links connected to each other, and identify them as the links with bottlenecks.

[0042] Based on the links with bottlenecks, the first optimal transmission path is divided into two parts, and the second type of elimination is performed again to find the second optimal transmission path.

[0043] A second aspect of this invention discloses a multi-path transmission system based on multi-objective optimization, the system comprising: a data acquisition module, a path planning module, a multi-objective optimization module, and a multi-path calculation module; wherein:

[0044] Call the data acquisition module: Obtain network traffic data and topology data from real heterogeneous networks in order to extract network topology information and node link status;

[0045] Call the path planning module: Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set;

[0046] Call the multi-objective optimization module: optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, and determine the optimal path as the first optimal transmission path;

[0047] Call the multipath calculation module: Based on the first optimal transmission path and the network topology information, reduce the network topology to obtain a simplified network topology;

[0048] The path planning module is invoked again: a new set of feasible paths is formed based on the simplified network topology; the multi-objective optimization module is invoked again: a new optimal path is determined using the new set of feasible paths, which serves as the second optimal transmission path;

[0049] The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability.

[0050] A third aspect of this invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the multi-path transmission method based on multi-objective optimization described in the first aspect of this disclosure.

[0051] A fourth aspect of this invention discloses a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the multi-path transmission method based on multi-objective optimization described in the first aspect of this disclosure.

[0052] In summary, this invention enables reliable multipath transmission in heterogeneous networks. The method selects the desired path by balancing reliability and latency, and proposes a topology pruning method to ensure that the new path and the original path traverse different links as much as possible, thereby maximizing transmission reliability. Attached Figure Description

[0053] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0054] Figure 1 This is a schematic diagram of a multi-path transmission framework based on multi-objective optimization according to an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Definitions of abbreviations and key terms:

[0057] (1) NAGA-II: Non-dominated Sorting Genetic Algorithm II;

[0058] (2) TCP: Transmission Control Protocol;

[0059] (3)SCTP: Stream Control Transmission Protocol.

[0060] The first aspect of this invention discloses a multi-path transmission method based on multi-objective optimization, the method comprising (e.g.) Figure 1 As shown):

[0061] Step S1: Call the data acquisition module to obtain network traffic data and topology data from the real heterogeneous network in order to extract network topology information and node link status.

[0062] Step S2: Invoke the path planning module. Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set.

[0063] Step S3: Call the multi-objective optimization module to optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, so as to determine the optimal path as the first optimal transmission path.

[0064] Step S4: Call the multipath calculation module, based on the first optimal transmission path and the network topology information, to reduce the network topology and obtain a simplified network topology.

[0065] Step S5: Call the path planning module again to form a new set of feasible paths based on the simplified network topology; call the multi-objective optimization module again to determine a new optimal path using the new set of feasible paths, which will serve as the second optimal transmission path.

[0066] The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability.

[0067] According to the method of the first aspect of the present invention, in step S1: the data acquisition module is invoked, and the request tool is used to acquire traffic data and topology data from the network every 5 seconds, and further acquire the link status and link connectivity status of each node, thereby extracting the network topology information and the node link status.

[0068] According to the method of the first aspect of the present invention, in step S2:

[0069] The path planning module is invoked, which utilizes a variant of the depth-first search algorithm to recursively probe all paths in the network based on the source and destination nodes to form a set of feasible paths, while ensuring that each path in the set of feasible paths has no duplicate vertices.

[0070] Specifically, during the exploration process, a variant algorithm based on depth-first search is used to record the length of the current path, and the search stops when the path length exceeds the known maximum path length, in order to reduce unnecessary computation.

[0071] According to the method of the first aspect of the present invention, in step S3, a multi-objective optimization module is invoked to calculate the optimal path using the NSGA-II algorithm, wherein:

[0072] The NSGA-II algorithm is an algorithm for multi-objective optimization problems based on genetic algorithms and the concept of Pareto optimality. It utilizes the global search capability of the NSGA-II algorithm to avoid getting trapped in local optima during the iteration process.

[0073] If the optimization objectives are to maximize path reliability and minimize transmission latency, then the objective function is expressed as:

[0074] minF=min{Rreliability max delay min}

[0075] minF = w1D + w2(-R)

[0076] stD=∑t i ·e i

[0077] R = Πr i ·v i

[0078] R≥0.998

[0079] D≤500

[0080] Where: R is the reliability of the transmission path, calculated by multiplying the reliability of each node; D is the delay of the transmission path, calculated by summing the delays of each link.

[0081] After determining the objective function, the initial population obtained by the path planning module is subjected to non-dominated sorting and selection, crossover, mutation, constraint checking and repair to generate the first generation of offspring population and update the evolutionary generation.

[0082] For each generation, fast non-dominated sorting, crowding calculation, and elite strategy operations are performed until the maximum number of generations. The NSGA-II algorithm outputs the Pareto front solution, which is the first optimal transmission path in the current situation.

[0083] According to the method of the first aspect of the present invention, in step S4, a first type of deletion is performed on the network topology; specifically including:

[0084] Find the sub-networks where the source node and destination node are located, and find the sub-networks where each node other than the source node and destination node is located on the first optimal transmission path;

[0085] Determine whether the subnet where the node is located is the same as the subnet where the source node is located or the subnet where the destination node is located;

[0086] If they are the same, delete the node and the links associated with it; otherwise, delete all nodes in the subnetwork where the node is located and the links associated with each node in the subnetwork.

[0087] In this process, after performing the first type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the second type of deletion is performed.

[0088] According to the method of the first aspect of the present invention, in step S4, a second type of deletion is performed on the network topology; specifically, it includes: deleting each link on the first optimal transmission path in the network topology; wherein:

[0089] After performing the second type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the third type of deletion is performed.

[0090] According to the method of the first aspect of the present invention, in step S4, a third type of deletion is performed on the network topology; specifically including:

[0091] Based on the links on the first optimal transmission path, find the nodes on the path that have only 2 links connected to each other, and identify them as the links with bottlenecks.

[0092] Based on the links with bottlenecks, the first optimal transmission path is divided into two parts, and the second type of elimination is performed again to find the second optimal transmission path.

[0093] First embodiment (e.g.) Figure 1 (As shown)

[0094] The first step is to acquire network traffic and topology data. The data acquisition module uses the `request` tool to retrieve the link status and connectivity of each node in the network every 5 seconds. The main data acquisition module obtains network topology information and node link status.

[0095] The second step utilizes a variant of the depth-first search algorithm to recursively explore all paths in the network based on the source and destination nodes, while ensuring that each path in the path set has no duplicate vertices. During the search process, the algorithm records the length of the current path, and stops further searching when the path length exceeds the known maximum path length to reduce unnecessary computation.

[0096] Thirdly, this invention utilizes a multi-objective optimization module to actively select suitable paths to maximize path reliability and minimize transmission delay. This invention employs the NSGA-II algorithm to calculate the optimal path. The NSGA-II algorithm is a widely used algorithm for multi-objective optimization problems, based on genetic algorithms and the Pareto optimality concept. Utilizing the global search capability of the NSGA-II algorithm, it avoids getting trapped in local optima during iteration. The optimization objectives of this invention are multiple: maximizing reliability and minimizing delay. Therefore, the objective function can be defined as:

[0097] minF=min{Rreliability max delay min}

[0098] minF = w1D + w2(-R)

[0099] stD=∑t i ·e i

[0100] R = Πr i ·v i

[0101] R≥0.998

[0102] D≤500

[0103] Here, R represents the reliability of the transmission path, calculated as the cumulative product of the reliability of each node. D represents the delay of the transmission path, calculated as the sum of the delays of each link. After determining the objective function, the initial population obtained from the path planning module undergoes non-dominated sorting, selection, crossover, mutation, constraint checking, and repair to generate the first generation of offspring and update the generation number. Then, for each generation, fast non-dominated sorting, congestion calculation, elitist strategies, and other operations are performed until the maximum number of generations is reached. The algorithm outputs the Pareto front solution, which is the first optimal transmission path in the current situation.

[0104] The fourth step involves the multi-path calculation module reducing the topology based on the first optimal transmission path. The core operation involves three different types of simplification of the network topology graph.

[0105] After obtaining the first path that satisfies the multiple constraints using the NSGA-II-based multi-constraint optimal path algorithm, the first type of simplification operation is first performed.

[0106] The core idea of ​​the first type of simplification operation is to first find the sub-networks containing the source and destination nodes, then find the sub-networks containing each node (excluding the source and destination nodes) on the first path that satisfies the multiple constraints. Finally, it is determined whether the sub-network containing the node is the same as the sub-network containing the source or destination node. If they are the same, the node and its associated links are deleted; otherwise, all nodes in the sub-network containing the node and their associated connections are deleted. After performing the first type of simplification operation, a new network topology is obtained. The selected NSGA-II-based multi-constraint optimal path algorithm is then used again to obtain the second path that satisfies the multiple constraints. If the second path exists, the search is successful; otherwise, the second type of simplification operation is used.

[0107] The core idea of ​​the second type of simplification operation is to remove every link on the first path that satisfies the multiple constraints in the network topology. After performing the second type of simplification operation, a new network topology is obtained. The behavior tree-based multi-constraint optimal path algorithm is then used again to obtain the second path that satisfies the multiple constraints. If the second path exists, the search is successful; otherwise, the third type of simplification operation is used.

[0108] The core idea of ​​the third type of simplification operation is to find the nodes on the path that are connected by only two links, which may be the "link bottlenecks". Then, based on the unique link, the first path is divided into two parts. The second type of simplification method is used again to find the second path that meets the multiple constraints. This way, the search will definitely be successful.

[0109] The purpose of the three types of simplification operations is primarily to significantly improve the overall reliability of the two paths. The first type of simplification operation places nodes on the second path and nodes on the first path in different subnetworks, so that when the subnetwork of a node on the first path fails, nodes on the second path can still transmit data normally. The main purpose of the second type of simplification operation is to ensure that the two paths do not share a common link, so that when a link on the first path fails, data on the second path can still be transmitted normally. The third type of simplification operation is a special case of the second type, namely, how to calculate a second path that satisfies multiple constraints when data can only be transmitted through a certain segment of the link on the first path.

[0110] The fifth step involves calculating two path instructions that satisfy multiple constraints and transmitting them to the sending end. The sending end then performs redundant transmission using this method, which can maximize the reliability of the transmission.

[0111] A second aspect of this invention discloses a multi-path transmission system based on multi-objective optimization, the system comprising: a data acquisition module, a path planning module, a multi-objective optimization module, and a multi-path calculation module; wherein:

[0112] Call the data acquisition module: Obtain network traffic data and topology data from real heterogeneous networks in order to extract network topology information and node link status;

[0113] Call the path planning module: Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set;

[0114] Call the multi-objective optimization module: optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, and determine the optimal path as the first optimal transmission path;

[0115] Call the multipath calculation module: Based on the first optimal transmission path and the network topology information, reduce the network topology to obtain a simplified network topology;

[0116] The path planning module is invoked again: a new set of feasible paths is formed based on the simplified network topology; the multi-objective optimization module is invoked again: a new optimal path is determined using the new set of feasible paths, which serves as the second optimal transmission path;

[0117] The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability.

[0118] A third aspect of this invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the multi-path transmission method based on multi-objective optimization described in the first aspect of this disclosure.

[0119] A fourth aspect of this invention discloses a computer-readable storage medium. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the multi-path transmission method based on multi-objective optimization described in the first aspect of this disclosure.

[0120] In summary, this invention enables reliable multipath transmission in heterogeneous networks. The method selects the desired path by balancing reliability and latency, and proposes a topology pruning method to ensure that the new path and the original path traverse different links as much as possible, thereby maximizing transmission reliability.

[0121] Please note that the technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments have been described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification. The above embodiments only illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A multi-path transmission method based on multi-objective optimization, characterized in that, The method includes: Step S1: Call the data acquisition module to obtain network traffic data and topology data from the real heterogeneous network in order to extract network topology information and node link status. Step S2: Invoke the path planning module. Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set. Step S3: Call the multi-objective optimization module to optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, so as to determine the optimal path as the first optimal transmission path. Step S4: Call the multipath calculation module, based on the first optimal transmission path and the network topology information, to reduce the network topology and obtain a simplified network topology. Step S5: Call the path planning module again to form a new set of feasible paths based on the simplified network topology; call the multi-objective optimization module again to determine a new optimal path using the new set of feasible paths, which will serve as the second optimal transmission path. The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability. In step S4, the network topology is reduced in the first type; specifically, this includes: Find the sub-networks where the source node and destination node are located, and find the sub-networks where each node other than the source node and destination node is located on the first optimal transmission path; Determine whether the subnet where the node is located is the same as the subnet where the source node is located or the subnet where the destination node is located; If they are the same, delete the node and the links associated with it; otherwise, delete all nodes in the subnetwork where the node is located and the links associated with each node in the subnetwork. In this process, after performing the first type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the second type of deletion is performed. In step S4, a second type of deletion is performed on the network topology; specifically, this includes deleting each link on the first optimal transmission path from the network topology; wherein: After performing the second type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the third type of deletion is performed. In step S4, a third type of deletion is performed on the network topology; specifically, it includes: Based on the links on the first optimal transmission path, find the nodes on the path that have only 2 links connected to each other, and identify them as the links with bottlenecks. Based on the links with bottlenecks, the first optimal transmission path is divided into two parts, and the second type of elimination is performed again to find the second optimal transmission path.

2. The multi-path transmission method based on multi-objective optimization according to claim 1, characterized in that, In step S1: the data acquisition module is invoked, and the request tool is used to obtain traffic data and topology data from the network every 5 seconds, and further obtain the link status and link connectivity status of each node, thereby extracting the network topology information and the node link status.

3. The multi-path transmission method based on multi-objective optimization according to claim 2, characterized in that, In step S2: The path planning module is invoked, which utilizes a variant of the depth-first search algorithm to recursively probe all paths in the network based on the source and destination nodes to form a set of feasible paths, while ensuring that each path in the set of feasible paths has no duplicate vertices. Specifically, during the exploration process, a variant algorithm based on depth-first search is used to record the length of the current path, and the search stops when the path length exceeds the known maximum path length, in order to reduce unnecessary computation.

4. The multi-path transmission method based on multi-objective optimization according to claim 3, characterized in that, In step S3, the multi-objective optimization module is invoked to calculate the optimal path using the NSGA-II algorithm, wherein: The NSGA-II algorithm is an algorithm for multi-objective optimization problems based on genetic algorithms and the concept of Pareto optimality. It utilizes the global search capability of the NSGA-II algorithm to avoid getting trapped in local optima during the iteration process. If the optimization objectives are to maximize path reliability and minimize transmission latency, then the objective function is expressed as: Where: R is the reliability of the transmission path, calculated by multiplying the reliability of each node; D is the delay of the transmission path, calculated by summing the delays of each link. After determining the objective function, the initial population obtained by the path planning module is subjected to non-dominated sorting and selection, crossover, mutation, constraint checking and repair to generate the first generation of offspring population and update the evolutionary generation. For each generation, fast non-dominated sorting, crowding calculation, and elite strategy operations are performed until the maximum number of generations. The NSGA-II algorithm outputs the Pareto front solution, which is the first optimal transmission path in the current situation.

5. A multi-path transmission system based on multi-objective optimization, characterized in that, The system includes: a data acquisition module, a path planning module, a multi-objective optimization module, and a multi-path calculation module; wherein: Call the data acquisition module: Obtain network traffic data and topology data from real heterogeneous networks in order to extract network topology information and node link status; Call the path planning module: Based on the network topology information and the node link status, use a variant algorithm based on depth-first search to detect all paths in the network to form a feasible path solution set; Call the multi-objective optimization module: optimize the paths in the feasible path solution set to maximize path reliability and minimize transmission delay, and determine the optimal path as the first optimal transmission path; The multi-path calculation module is invoked: based on the first optimal transmission path and the network topology information, the network topology is reduced to obtain a simplified network topology; The path planning module is invoked again: a new set of feasible paths is formed based on the simplified network topology; the multi-objective optimization module is invoked again: a new optimal path is determined using the new set of feasible paths, which serves as the second optimal transmission path; The first optimal transmission path and the second optimal transmission path are transmitted to the sending end, and the sending end performs redundant transmission based on the first optimal transmission path and the second optimal transmission path to improve transmission reliability. The first type of reduction is performed on the network topology; specifically, it includes: Find the sub-networks where the source node and destination node are located, and find the sub-networks where each node other than the source node and destination node is located on the first optimal transmission path; Determine whether the subnet where the node is located is the same as the subnet where the source node is located or the subnet where the destination node is located; If they are the same, delete the node and the links associated with it; otherwise, delete all nodes in the subnetwork where the node is located and the links associated with each node in the subnetwork. In this process, after performing the first type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the second type of deletion is performed. The network topology is subject to a second type of reduction; specifically, this includes removing every link on the first optimal transmission path from the network topology; where: After performing the second type of deletion, a new network topology is obtained. The path planning module is called again to form a new set of feasible paths based on the simplified network topology. The multi-objective optimization module is called again to determine a new optimal path using the new set of feasible paths, which is then used as the second optimal transmission path. If the second optimal transmission path does not exist, the third type of deletion is performed. The third type of reduction in network topology includes: Based on the links on the first optimal transmission path, find the nodes on the path that have only 2 links connected to each other, and identify them as the links with bottlenecks. Based on the links with bottlenecks, the first optimal transmission path is divided into two parts, and the second type of elimination is performed again to find the second optimal transmission path.

6. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the multi-path transmission method based on multi-objective optimization as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements a multi-path transmission method based on multi-objective optimization as described in any one of claims 1-4.

Images