Intelligent multi-path dynamic routing method and system based on SD-WAN

Abstract

The application discloses an intelligent multi-path dynamic routing system based on SD-WAN, which comprises a logical network layer, a routing node control unit, a routing node monitoring unit and a routing node intelligent adjustment unit.The logical network layer is connected with a physical network and is used for constructing a communication link; the routing node control unit is used for dynamically setting a routing node in the logical network layer; the routing node monitoring unit is used for monitoring a communication state of the routing node in real time; and the routing node intelligent adjustment unit is used for dynamically adjusting the routing node according to a monitoring result of the routing node monitoring unit.The application can improve the defects of the prior art, successfully introduces an ant colony algorithm into a software defined wide area network dynamic routing process, and improves the accuracy of the intelligent routing system.

Description

Technical Field

[0001] This invention relates to the field of Internet communication technology, and in particular to an intelligent multi-path dynamic routing method and system based on SD-WAN. Background Technology

[0002] Software-defined wide area network (SD-WAN) is a solution that dynamically manages wide area networks using software technology, enabling high-speed network connections across geographical regions using virtual links. Through dynamic routing, it balances performance, cost, and reliability requirements. Traditional SD-WAN dynamic routing methods largely rely on analysis systems, similar to expert decision-making systems, to analyze numerous network parameters and then provide routing results. This process involves a significant amount of experience-based judgment, and relying solely on human analysis can lead to substantial biases in the results. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide an intelligent multi-path dynamic routing method and system based on SD-WAN, which can overcome the shortcomings of the existing technology, successfully introduce the ant colony algorithm into the dynamic routing process of software-defined wide area network, and improve the accuracy of the intelligent routing system.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows.

[0005] An intelligent multi-path dynamic routing system based on SD-WAN includes, The logical network layer communicates with the physical network and is used to build communication links; The routing node control unit is used to dynamically configure routing nodes in the logical network layer. The routing node monitoring unit is used to monitor the communication status of routing nodes in real time. The intelligent adjustment unit for routing nodes is used to dynamically adjust routing nodes based on the monitoring results from the routing node monitoring unit.

[0006] A routing method for the above-mentioned intelligent multi-path dynamic routing system based on SD-WAN includes the following steps: A communication connection is established between the logical network layer and the physical network. The routing node control unit sets up routing nodes in the logical network layer according to a preset scheme. The routing node monitoring unit monitors the communication status of the routing nodes in real time and sends the monitoring data to the routing node intelligent adjustment unit. After analyzing the monitoring data from the routing node monitoring unit, the routing node intelligent adjustment unit dynamically adjusts the setting method of the routing nodes in the logical network layer and sends the setting method to the routing node control unit. The routing node control unit reconfigures the routing nodes according to the latest setting method of the routing nodes in the logical network layer received.

[0007] Preferably, the monitoring data includes packet loss rate, latency, jitter, and port throughput.

[0008] Preferably, the intelligent adjustment unit for routing nodes incorporates an ant colony optimization model. This model comprises two ant colonies of equal size, with each ant forming a one-to-one pair. Ants within the same pair maintain real-time communication. The two ant colonies perform optimization operations alternately, with the operation timings of the two colonies differing by one clock cycle. Each ant starts from an input port on the logical network layer and moves towards an output port, encountering a new routing node with each step. The probability of selecting a different routing node during each move is proportional to the product of the pheromone concentration along the path and the fitness of the target routing node. Once an ant reaches the target routing node, it uses... The ant's own movement rate is used to weight and correct the fitness of the target routing node. The weighting correction coefficient is proportional to the ant's own movement rate. When another ant in the same ant pair selects a routing node in the next clock cycle, the corrected fitness is used as the fitness of the target routing node. The fitness of other target routing nodes is the uncorrected fitness. The above optimization process is repeated. When it is found that the average fitness of all routing nodes on a certain path is less than a set threshold or the variance of the fitness of all routing nodes is greater than a set threshold, the communication path is deleted, and a new communication path is established based on the current highest fitness. Then, the communication path change information is sent to the routing node control unit.

[0009] Preferably, the number of individuals in each ant colony is equal to the number of input ports on the logical network layer.

[0010] As a preferred option, the fitness of a routing node is f = α(T) / (k1L + k2D + k3J), where α is the weighted correction coefficient, T is the throughput, L is the packet loss rate, D is the delay, J is the jitter, and k1 to k3 are the proportional coefficients of the corresponding parameters. When using the uncorrected fitness of the routing node, α = 1.

[0011] Preferably, when the monitoring data collected by the routing node monitoring unit exceeds the normal threshold, the routing node corresponding to the abnormal monitoring data is deleted, and the routing node control unit randomly sets the routing node again.

[0012] The beneficial effects of adopting the above technical solution are as follows: This invention solves the problem of slow optimization speed of existing ant colony algorithms by improving the ant colony algorithm, making it adaptable to the real-time requirements of dynamic routing in software-defined wide area networks. Utilizing the ant colony algorithm to optimize the best configuration scheme for routing nodes can accurately determine the optimal configuration scheme, reducing the workload of manual intervention. Attached Figure Description

[0013] Figure 1 This is a flowchart of a specific embodiment of the present invention. Detailed Implementation

[0014] Reference Figure 1 The intelligent multi-path dynamic routing system based on SD-WAN disclosed in this invention includes, The logical network layer communicates with the physical network and is used to build communication links; The routing node control unit is used to dynamically configure routing nodes in the logical network layer. The routing node monitoring unit is used to monitor the communication status of routing nodes in real time. The intelligent adjustment unit for routing nodes is used to dynamically adjust routing nodes based on the monitoring results from the routing node monitoring unit.

[0015] The routing method of the above-mentioned routing system includes the following steps: A communication connection is established between the logical network layer and the physical network. The routing node control unit sets up routing nodes in the logical network layer according to a preset scheme. The routing node monitoring unit monitors the communication status of the routing nodes in real time and sends the monitoring data to the routing node intelligent adjustment unit. After analyzing the monitoring data from the routing node monitoring unit, the routing node intelligent adjustment unit dynamically adjusts the setting method of the routing nodes in the logical network layer and sends the setting method to the routing node control unit. The routing node control unit reconfigures the routing nodes according to the latest setting method of the routing nodes in the logical network layer received.

[0016] This invention introduces the ant colony algorithm into the intelligent routing process, successfully solving the problem of excessive delay in the optimization process of the traditional ant colony algorithm, thus meeting the real-time requirements of the intelligent routing system.

[0017] This invention transforms the original single ant colony into two ant colonies, establishing a one-to-one communication relationship between individuals within each colony. These two colonies then perform a serial, alternating optimization process. By utilizing real-time monitoring of link communication efficiency between individuals, the fitness of routing nodes is adjusted, significantly accelerating the convergence speed of the optimization process. Simultaneously, it changes the prediction method for the final optimal solution output in the original ant colony algorithm, instead synchronously eliminating abnormal links during the optimization process, thus ensuring the entire network is always in an optimal link configuration. The specific optimization process is described below.

[0018] The ant colony optimization model consists of two ant colonies of equal size. Ants within each colony form one-to-one pairs, and the two ants within the same pair maintain real-time communication. The number of ants in each colony equals the number of input ports on the logical network layer. The optimization operation is performed by alternating sequences of the two ant colonies, with the operation timings of the two colonies differing by one clock cycle. Each ant starts from an input port on the logical network layer and moves towards an output port, encountering a new routing node with each step. The probability of choosing a different routing node at each move is proportional to the product of the pheromone concentration along the path and the fitness of the target routing node. When an ant reaches a target routing node, it uses its movement speed to weighted adjust the fitness of the target routing node, with the weighting adjustment coefficient proportional to its movement speed. When another ant in the same ant pair selects a routing node in the next clock cycle, it uses the adjusted fitness as the fitness of this target routing node; the fitness of other target routing nodes uses the unadjusted fitness. The fitness of a routing node is f = α(T) / (k1L + k2D + k3J), where α is the weighted correction coefficient, T is the throughput, L is the packet loss rate, D is the latency, J is the jitter, and k1 to k3 are the proportional coefficients of the corresponding parameters. When using the uncorrected fitness of the routing node, α = 1. The optimization process is repeated. When the average fitness of all routing nodes on a certain path is found to be less than a set threshold or the variance of the fitness of all routing nodes is found to be greater than a set threshold, this communication path is deleted, and a new communication path is established based on the highest current fitness. Then, the communication path change information is sent to the routing node control unit.

[0019] In addition, when the monitoring data collected by the routing node monitoring unit exceeds the normal threshold, the routing node corresponding to the abnormal monitoring data is directly deleted without the need for processing by the routing node intelligent adjustment unit, and the routing node control unit randomly sets the routing node again.

[0020] This invention achieves efficient and accurate dynamic routing in software-defined wide area networks by optimizing the optimization process of the intelligent routing algorithm.

[0021] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent multi-path dynamic routing system based on SD-WAN, characterized in that: include, The logical network layer communicates with the physical network and is used to build communication links; The routing node control unit is used to dynamically configure routing nodes in the logical network layer. The routing node monitoring unit is used to monitor the communication status of routing nodes in real time. The intelligent adjustment unit for routing nodes is used to dynamically adjust routing nodes based on the monitoring results from the routing node monitoring unit.

2. A routing method for an intelligent multi-path dynamic routing system based on SD-WAN as described in claim 1, characterized in that... Includes the following steps: A communication connection is established between the logical network layer and the physical network. The routing node control unit sets up routing nodes in the logical network layer according to a preset scheme. The routing node monitoring unit monitors the communication status of the routing nodes in real time and sends the monitoring data to the routing node intelligent adjustment unit. After analyzing the monitoring data from the routing node monitoring unit, the routing node intelligent adjustment unit dynamically adjusts the setting method of the routing nodes in the logical network layer and sends the setting method to the routing node control unit. The routing node control unit reconfigures the routing nodes according to the latest setting method of the routing nodes in the logical network layer received.

3. The routing method of the intelligent multi-path dynamic routing system based on SD-WAN according to claim 2, characterized in that: The monitoring data includes packet loss rate, latency, jitter, and port throughput.

4. The routing method of the intelligent multi-path dynamic routing system based on SD-WAN according to claim 2, characterized in that: The intelligent adjustment unit for routing nodes incorporates an ant colony optimization model. This model consists of two ant colonies of equal size, with each ant forming a one-to-one pair. Ants within the same pair maintain real-time communication and perform optimization operations alternately using the two ant colonies. The operations of the two ant colonies are timed one clock cycle apart. Each ant starts from an input port on the logical network layer and moves towards an output port, encountering a new routing node with each step. The probability of selecting a different routing node during each move is proportional to the product of the pheromone concentration along the path and the fitness of the target routing node. Once an ant reaches the target routing node, it uses its own... The speed of movement is used to weight and correct the fitness of the target routing node. The weighting correction coefficient is proportional to the speed of movement. When another ant in the same ant pair selects a routing node in the next clock cycle, the corrected fitness is used as the fitness of the target routing node. The fitness of other target routing nodes is the uncorrected fitness. The above optimization process is repeated. When it is found that the average fitness of all routing nodes on a certain path is less than a set threshold or the variance of the fitness of all routing nodes is greater than a set threshold, the communication path is deleted, and a new communication path is established based on the highest current fitness. Then, the communication path change information is sent to the routing node control unit.

5. The routing method of the intelligent multi-path dynamic routing system based on SD-WAN according to claim 4, characterized in that: The number of individuals in each ant colony is equal to the number of input ports on the logical network layer.

6. The routing method of the intelligent multi-path dynamic routing system based on SD-WAN according to claim 5, characterized in that: The fitness of a routing node is f = α(T) / (k1L + k2D + k3J), where α is the weighted correction coefficient, T is the throughput, L is the packet loss rate, D is the delay, J is the jitter, and k1 to k3 are the proportional coefficients of the corresponding parameters. When using the uncorrected fitness of the routing node, α = 1.

7. The routing method of the intelligent multi-path dynamic routing system based on SD-WAN according to claim 2, characterized in that: If the monitoring data collected by the routing node monitoring unit exceeds the normal threshold, the routing node corresponding to the abnormal monitoring data will be deleted, and the routing node control unit will randomly reconfigure the routing node.

Images