Multi-topology routing dynamic management method and system based on flow demand perception

Abstract

The invention relates to a multi-topology routing dynamic management method and system based on flow demand perception, and the method comprises the following steps: collecting the real-time flow of each access point in a current network in real time, and obtaining the real-time flow data corresponding to each access point; judging whether a dynamic adjustment condition is met or not according to the real-time flow data; if yes, calculating an adjustment probability; dynamically adjusting the multi-topology routing according to the adjustment probability; and if not, continuing to monitor the real-time traffic of each access point. According to the invention, burst traffic can be quickly responded, resource allocation is optimized, and the service quality is effectively improved in a complex dynamic network environment. Meanwhile, the problems of overlarge switch flow table updating overhead, delay increase and high CPU and memory occupation caused by frequent adjustment are also avoided, so that the network stability and the resource utilization efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of network communications, and more particularly to a multi-topology routing dynamic management method and system based on traffic demand perception. Background Art

[0002] With the advancement of network technology, to ensure efficient and reliable data transmission, networks need to flexibly respond to changing traffic demands at different access points. Traditional single-path routing methods typically select a fixed path for each node pair. However, when this path fails to meet the Quality of Service (QoS) requirements (such as bandwidth and latency) of all data flows, it can cause congestion, increase latency and packet loss, and degrade user experience.

[0003] To address the limitations of single-path routing, multipath routing has been proposed. Multi-topology routing is one implementation of multipath routing. Multi-topology routing generates multiple logical topologies based on the physical topology, each with independent routing rules. Therefore, the data paths for the same node pair in different logical topologies may be different. By distributing data flows across multiple logical topologies, multipath transmission can be achieved. Software-Defined Networking (SDN) is a network architecture that separates the control plane from the data plane. The controller is responsible for centralized management and issuing routing decisions and traffic control, while switches forward data according to the controller's instructions. The controller and switches communicate using protocols such as OpenFlow and P4. To implement multi-topology routing, the SDN controller installs multiple flow tables in the relevant switches, defining forwarding rules across multiple topologies. In this way, SDN allows for the selection of one or more paths based on the QoS requirements of data flows in dynamic network environments. For example, high-bandwidth data flows can be prioritized for high-bandwidth paths, while latency-sensitive data flows can be allocated to lower-latency paths. Through offload processing, multi-topology routing can meet the differentiated QoS requirements of multiple data flows and improve the overall transmission efficiency and resource utilization efficiency of the network.

[0004] However, traditional multi-topology routing methods are typically based on static configurations, assuming a relatively stable network environment and balanced traffic demands across all access points. This makes them difficult to adapt to dynamically changing traffic demands. For example, when traffic demands surge at certain nodes, static multi-topology routing cannot allocate more paths to them, resulting in insufficient bandwidth and network congestion. For nodes with low traffic demands, idle link bandwidth wastes resources. Therefore, developing a multi-topology routing management method that can dynamically adjust to real-time traffic demands has become crucial for improving network performance and resource efficiency. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a multi-topology routing dynamic management method and system based on traffic demand perception in response to the problems existing in the prior art.

[0006] The technical solution adopted by the present invention to solve the technical problem is to construct a multi-topology routing dynamic management method based on traffic demand perception, including the following steps:

[0007] Collect the real-time traffic of each access point in the current network in real time to obtain the real-time traffic data corresponding to each access point;

[0008] Determining whether a dynamic adjustment condition is met based on the real-time traffic data;

[0009] If so, calculate the adjusted probability;

[0010] Dynamically adjusting the multi-topology routing according to the adjustment probability;

[0011] If not, continue to monitor the real-time traffic of each access point.

[0012] In the multi-topology routing dynamic management method based on traffic demand awareness according to the present invention, determining whether the dynamic adjustment condition is met according to the real-time traffic data includes:

[0013] comparing the real-time traffic data with a traffic threshold;

[0014] If the real-time traffic data is greater than the traffic threshold, it is determined that the dynamic adjustment condition is met;

[0015] If not, it is determined that the dynamic adjustment condition is not met.

[0016] In the multi-topology routing dynamic management method based on traffic demand awareness according to the present invention, the calculation of the adjustment probability includes:

[0017] Calculate burst traffic increment;

[0018] Determine the latency overhead incurred by the controller-switch interaction during the adjustment process;

[0019] Determine the constant coefficient corresponding to traffic changes and adjustment overhead;

[0020] The adjustment probability is obtained by performing calculation based on the burst traffic increment, the delay overhead generated by the interaction between the controller and the switch during the adjustment process, the traffic change, and a constant coefficient corresponding to the adjustment overhead.

[0021] In the multi-topology routing dynamic management method based on traffic demand awareness described in the present invention, the adjustment probability satisfies the following formula:

[0022]

[0023] Among them, P adjust represents the adjustment probability, Δtraffic represents the burst traffic increment, L CP represents the delay overhead caused by the interaction between the controller and the switch during the adjustment process. A, B, and C are constant coefficients related to traffic changes and adjustment overhead.

[0024] In the multi-topology routing dynamic management method based on traffic demand awareness described in the present invention, the delay overhead generated by the interaction between the controller and the switch includes:

[0025] Determine controller processing delays;

[0026] Determine the propagation delay between the controller and the switch;

[0027] Determine synchronization delays between multiple controllers;

[0028] The delay overhead generated by the interaction between the controller and the switch is obtained by calculation based on the controller processing delay, the propagation delay between the controller and the switch, and / or the synchronization delay between the multiple controllers.

[0029] In the multi-topology routing dynamic management method based on traffic demand awareness of the present invention, the delay overhead generated by the interaction between the controller and the switch satisfies the following formula:

[0030]

[0031] in, Indicates controller processing delay, represents the propagation delay between the controller and the switch, L syn Indicates the synchronization delay between multiple controllers.

[0032] In the multi-topology routing dynamic management method based on traffic demand awareness according to the present invention, after dynamically adjusting the multi-topology routing according to the adjustment probability, the following steps are further performed:

[0033] Determining whether the real-time traffic data is less than a traffic threshold;

[0034] If so, the initial multi-topology routing is restored.

[0035] In the multi-topology routing dynamic management method based on traffic demand awareness according to the present invention, the method further includes:

[0036] Collect historical traffic data;

[0037] Analyzing the historical traffic data and training the data using a long short-term memory network to obtain a traffic prediction model;

[0038] The multi-topology routing is dynamically adjusted based on the traffic prediction model.

[0039] The present invention also provides a multi-topology routing dynamic management system based on traffic demand perception, comprising:

[0040] A traffic collection unit is used to collect real-time traffic of each access point in the current network in real time and obtain real-time traffic data corresponding to each access point;

[0041] a condition judgment unit, configured to judge whether a dynamic adjustment condition is satisfied based on the real-time traffic data;

[0042] A probability calculation unit, configured to calculate the adjustment probability when the dynamic adjustment condition is met;

[0043] A dynamic adjustment unit is used to dynamically adjust the multi-topology routing according to the adjustment probability.

[0044] The multi-topology routing dynamic management system based on traffic demand awareness according to the present invention further includes:

[0045] The feedback optimization unit is used to monitor whether the real-time traffic data is less than the traffic threshold, and restore the initial multi-topology routing when the real-time traffic data is less than the traffic threshold.

[0046] The implementation of the present invention's multi-topology routing dynamic management method and system based on traffic demand awareness has the following beneficial effects: real-time collection of real-time traffic at each access point in the current network to obtain real-time traffic data corresponding to each access point; determining whether dynamic adjustment conditions are met based on the real-time traffic data; if so, calculating the adjustment probability; dynamically adjusting the multi-topology routing based on the adjustment probability; if not, continuing to monitor the real-time traffic at each access point. The present invention can quickly respond to traffic bursts, optimize resource allocation, and effectively improve service quality in complex dynamic network environments. At the same time, the present invention also avoids the problems of excessive switch flow table update overhead, increased latency, and high CPU and memory usage caused by frequent adjustments, thereby improving network stability and resource utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0048] Figure 1 This is a flow chart of an embodiment of a multi-topology routing dynamic management method based on traffic demand perception provided by the present invention;

[0049] Figure 2 This is a flow chart of another embodiment of the multi-topology routing dynamic management method based on traffic demand perception provided by the present invention;

[0050] Figure 3 This is the overall flow chart of the multi-topology routing dynamic management method based on traffic demand perception provided by the present invention;

[0051] Figure 4 This is a principle block diagram of the multi-topology routing dynamic management system based on traffic demand perception provided by the present invention;

[0052] Figure 5 This is a schematic diagram of the distribution of the multi-topology routing provided by the present invention in state 1;

[0053] Figure 6 This is a distribution diagram of the multi-topology routing provided by the present invention in state 2. DETAILED DESCRIPTION

[0054] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0055] In order to solve the problems existing in traditional solutions, the present invention provides a multi-topology routing dynamic management method based on traffic demand perception, which is applicable to SDN networks. It can monitor the traffic of each access point in the network in real time and dynamically adjust the multi-topology routing using a probabilistic decision-making mechanism when the real-time traffic exceeds the traffic threshold, thereby optimizing network resource utilization.

[0056] refer to Figure 1 , Figure 1 The present invention provides a flowchart of a preferred embodiment of the multi-topology routing dynamic management method based on traffic demand perception.

[0057] Specifically, such as Figure 1 As shown, the multi-topology routing dynamic management method based on traffic demand perception includes the following steps:

[0058] Step S11: collecting the real-time traffic of each access point in the current network in real time to obtain the real-time traffic data corresponding to each access point.

[0059] Step S12: Determine whether the dynamic adjustment conditions are met based on the real-time traffic data.

[0060] Optionally, in an embodiment of the present invention, determining whether the dynamic adjustment condition is satisfied based on the real-time traffic data includes: comparing the real-time traffic data with a traffic threshold; if the real-time traffic data is greater than the traffic threshold, determining that the dynamic adjustment condition is satisfied; if not, determining that the dynamic adjustment condition is not satisfied. The traffic threshold (denoted by θ) can be determined by static configuration or dynamic adjustment.

[0061] Step S13: If yes, calculate the adjustment probability; if no, continue to monitor the real-time traffic of each access point.

[0062] Optionally, in an embodiment of the present invention, calculating the adjustment probability includes: calculating a burst traffic increment; determining a delay overhead generated by the interaction between the controller and the switch during the adjustment process; determining a constant coefficient corresponding to the traffic change and the adjustment overhead; and calculating the adjustment probability based on the burst traffic increment, the delay overhead generated by the interaction between the controller and the switch during the adjustment process, the traffic change, and the constant coefficient corresponding to the adjustment overhead. The adjustment probability satisfies the following formula:

[0063]

[0064] Among them, P adjust represents the adjustment probability, Δtraffic represents the burst traffic increment, L CP represents the delay overhead generated by the interaction between the controller and the switch during the adjustment process. A, B, and C are constant coefficients related to traffic changes and adjustment overhead. Specifically, coefficient A is used to adjust the degree of influence of the burst traffic increment on the probability change, while coefficients B and C are used to adjust the degree of influence of the adjustment overhead on the probability change. It should be noted that the A coefficient, B coefficient, and C coefficient can be determined to reasonable values based on the network traffic characteristics, thereby accurately controlling the degree of influence of traffic increment and overhead on the probability change. Furthermore, in the embodiment of the present invention, Δtraffic is the difference between the real-time traffic and the traffic value predicted at that moment in the traffic prediction model.

[0065] Optionally, in an embodiment of the present invention, the delay overhead generated by the interaction between the controller and the switch includes: determining the controller processing delay; determining the propagation delay between the controller and the switch; determining the synchronization delay between multiple controllers; and calculating the delay overhead generated by the interaction between the controller and the switch based on the controller processing delay, the propagation delay between the controller and the switch, and / or the synchronization delay between the multiple controllers. The delay overhead generated by the interaction between the controller and the switch satisfies the following formula:

[0066]

[0067] in, Indicates controller processing delay, represents the propagation delay between the controller and the switch, L syn represents the synchronization delay between multiple controllers. μ is the service rate of the controller, For switch v i The control packet flow required for control (that is, the sum of the number of flow table entries that need to be deleted and the number of flow table entries that need to be added), is a Boolean variable representing the controller c and switch v i The second term represents the propagation delay between the controller and the switch, where Represents controller c and switch v i The distance between Indicates the signal propagation rate. The third term L syn It should be noted that in a small-scale network, if there is only one controller in the small-scale network, the synchronization delay between multiple controllers in (2) can be ignored. If there are multiple controllers in the network, then L syn The value of is determined by the connection relationship between multiple controllers, load and processing capacity, synchronization protocol efficiency, etc.

[0068] Step S14: Dynamically adjust the multi-topology routing according to the adjustment probability.

[0069] Specifically, in the embodiment of the present invention, after the adjustment probability is calculated in step S13, whether to adjust the multi-topology routing configuration can be determined based on the calculated adjustment probability. That is, when the traffic demand increment is large and the adjustment overhead is small, the probability of selecting the multi-topology routing adjustment will increase.

[0070] Furthermore, in an embodiment of the present invention, Figure 1 As shown, after the multi-topology routing is dynamically adjusted according to the adjustment probability, the following steps are further performed:

[0071] Step S15: Determine whether the real-time traffic data is less than the traffic threshold.

[0072] Step S16: If yes, then restore the initial multi-topology route, that is, restore the original multi-topology route. If no, then continue to monitor the real-time traffic.

[0073] After dynamically adjusting the multi-topology routing based on the adjustment probability, the present invention can ensure efficient allocation of network resources by continuing monitoring and restoring the original model when the burst traffic falls below the traffic threshold.

[0074] The present invention achieves comprehensive optimization between adjustment cost and network performance by adopting the method of "adjustment based on probability" rather than "adjustment when exceeding the threshold". On the one hand, traffic demand is random and sudden. If it is adjusted immediately every time it exceeds the threshold, it may not be able to adapt to the scenario of rapid decline, resulting in frequent cancellation of adjustments and increased repeated overhead of the system. On the other hand, frequent adjustments will increase the number of path switching, resulting in fluctuations in data packet delays and even packet loss, affecting user experience. Through the probabilistic adjustment mechanism, the present invention can effectively filter short-term fluctuations, avoid the problems of excessive switch flow table update overhead, increased delays, high CPU and internal occupancy caused by frequent adjustments, thereby improving network stability and resource utilization. At the same time, relying on the feedback mechanism, the present invention can adjust the topology configuration in time after traffic changes, further improving the adaptability and flexibility of the network.

[0075] refer to Figure 2 , Figure 2 This is a flow chart of another embodiment of the multi-topology routing dynamic management method based on traffic demand awareness provided by the present invention.

[0076] like Figure 2 As shown, in this embodiment, after executing step S10, the following steps are included:

[0077] Step S201: Collect historical traffic data.

[0078] Specifically, the historical traffic data may be the traffic of each access point within a period of time.

[0079] Step S202: Analyze historical traffic data and train with a long short-term memory network to obtain a traffic prediction model.

[0080] Step S203: Dynamically adjust the multi-topology routing based on the traffic prediction model.

[0081] Specifically, since the traffic changes of each access point in the actual network usually cannot be affected by the time tidal characteristics, the present invention can extract the periodic traffic change rules based on historical traffic data, and use the long short-term memory network (LSTM-RNN) for training to obtain a traffic prediction model, and use the traffic prediction model to realize dynamic adjustment of multi-topology routing within a certain period.

[0082] refer to Figure 4 , Figure 4 This is a principle block diagram of the multi-topology routing dynamic management system based on traffic demand perception provided by the present invention.

[0083] like Figure 4 As shown, the multi-topology routing dynamic management system based on traffic demand perception includes:

[0084] The traffic collection unit 401 is used to collect the real-time traffic of each access point in the current network in real time, and obtain the real-time traffic data corresponding to each access point.

[0085] The condition judgment unit 402 is used to judge whether the dynamic adjustment condition is met according to the real-time traffic data.

[0086] The probability calculation unit 403 is configured to calculate the adjustment probability when the dynamic adjustment condition is met.

[0087] The dynamic adjustment unit 404 is configured to dynamically adjust the multi-topology routing according to the adjustment probability.

[0088] Further, if Figure 4 As shown, the multi-topology routing dynamic management system based on traffic demand perception also includes: a feedback optimization unit 405, which is used to monitor whether the real-time traffic data is less than the traffic threshold, and restore the initial multi-topology routing when the real-time traffic data is less than the traffic threshold.

[0089] The following is an explanation using a specific application example.

[0090] refer to Figure 5 and Figure 6 , Figure 5 and Figure 6 The figure shows the traffic distribution of each access point on the same network at different times. In this embodiment, a mobile terminal accesses the network through an access layer switch and establishes a connection to the Internet via aggregation and core layer switches. When a mobile terminal requests Internet content, the data flow follows the following path: Internet → core layer switch → aggregation layer switch → access layer switch → mobile terminal. In the mesh topology shown in the figure, core layer switches are interconnected, and each aggregation layer switch is connected to multiple core layer switches, allowing for multiple optional paths for data transmission. Each path is configured using multi-topology routing. Multiple transmission paths can be assigned to an access layer switch.

[0091] Specifically, such as Figure 3 As shown, first, the system collects historical traffic data, and then uses the long short-term memory network to train the traffic prediction model of each access point with a cycle of one day, and dynamically adjusts the multi-topology routing configuration within one day based on the traffic prediction model.

[0092] Next, take access switch S1 and access switch S2 as an example: Figure 5 As shown in Figure 1, the number of mobile terminals accessing switch S1 is large and the traffic demand is high, so two blue transmission paths are allocated to it; while the number of mobile terminals accessing switch S2 is small and the traffic demand is low, so only one red transmission path is allocated. As time goes by and the location of mobile terminals changes, the network enters Figure 6At this time, the number of mobile terminals accessing switch S2 increases, and the traffic demand surges, but this change is not predicted by the traffic prediction model. After the real-time traffic monitoring system detects that the traffic of access switch S2 exceeds the set threshold, the system calculates the adjustment probability. At this time, in the adjustment probability formula In this example, Δtraffic is the increment of burst traffic, which is the difference between the real-time traffic and the traffic value predicted by the traffic prediction model at that moment. CP To adjust the controller latency overhead for multi-topology routing, in this example, the adjustment overhead requires deleting flow entries from three core switches (S4, S6, and S5) (i.e., deleting a path assigned to switch S1: Internet → S4 → S6 → S5 → S3 → S1) and adding flow entries from three core switches (S4, S6, and S5) (i.e., adding a path for switch S2: Internet → S4 → S6 → S5 → S7 → S2). In the formula, μ is the service rate of the controller in the network. For switch v i The control packet required for the control. Taking switch S4 as an example, the switch needs to delete and add an entry, so λ S4 =1+1=2. is a Boolean variable representing the controller c and switch v i In this example, all switches are connected to one controller, so this field is set to 1. Represents controller c and switch v i The distance between them depends on the network parameters. represents the signal propagation rate, which depends on the transmission medium of the wired network. In this example, a small network with a single controller is used, so the third term in the formula can be ignored. In a larger network with multiple controllers, the value depends on parameters such as the connectivity between controllers, the signal propagation rate, and the physical distance. A, B, and C in the probability adjustment formula are constants. Experimentation can be conducted based on network traffic characteristics to determine reasonable values, adjusting the impact of traffic growth and overhead on probability changes.

[0093] Then, the system decides whether to adjust the multi-topology routing configuration based on the calculated adjustment probability. When the traffic demand increment is large and the adjustment overhead is small, the probability of the system choosing adjustment will increase. Assuming that the system selects an adjustment plan based on the adjustment probability, the relevant table entries are deleted and the relevant table entries are added. The adjusted multi-topology routing configuration is as follows: one blue transmission path is reduced for access switch S1, leaving only one transmission path; a red transmission path is added for access switch S2, for a total of two transmission paths. Figure 6As shown in the figure, after the adjustment, network resource allocation better meets current traffic demands, ensuring efficient resource utilization and high service quality. However, due to the limited bandwidth of links between core switches, multi-topology routing configuration requires adding multiple relevant entries to the switches, but the switch flow table capacity is limited. Therefore, it is not possible to allocate all possible simple paths as transmission paths to each target access point. Routing configuration requires a balance between traffic demands and device resource constraints.

[0094] Compared to existing solutions, this invention introduces a probabilistic adjustment mechanism. By quantifying controller processing delays, controller-switch interaction delays, and inter-controller synchronization delays, it constructs a mathematical model for adjustment costs and designs an adjustment probability formula based on traffic increments. This approach enables rapid response and optimized resource allocation when access point traffic surges, avoiding the repetitive overhead associated with frequent adjustments. This balance between performance improvement and adjustment overhead is achieved, contributing to intelligent management of multi-topology routing and efficient utilization of network resources.

[0095] Specifically, the specific coordination operation process between the various units in the multi-topology routing dynamic management system based on traffic demand perception can refer to the above-mentioned multi-topology routing dynamic management method based on traffic demand perception, which will not be repeated here.

[0096] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0097] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0098] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0099] The above embodiments are intended only to illustrate the technical concepts and features of the present invention. Their purpose is to enable those skilled in the art to understand the present invention and implement it accordingly. They are not intended to limit the scope of protection of the present invention. All equivalent variations and modifications within the scope of the claims of the present invention are intended to be covered by the claims of the present invention.

Claims

1. A multi-topology routing dynamic management method based on traffic demand perception, characterized in that: The following steps are involved: Collect the real-time traffic of each access point in the current network in real time to obtain the real-time traffic data corresponding to each access point; Determining whether a dynamic adjustment condition is met based on the real-time traffic data; If so, calculate the adjusted probability; Dynamically adjusting the multi-topology routing according to the adjustment probability; If not, continue to monitor the real-time traffic of each access point.

2. The multi-topology routing dynamic management method based on traffic demand perception according to claim 1 is characterized in that: The determining whether the dynamic adjustment condition is met according to the real-time traffic data includes: comparing the real-time traffic data with a traffic threshold; If the real-time traffic data is greater than the traffic threshold, it is determined that the dynamic adjustment condition is met; If not, it is determined that the dynamic adjustment condition is not met.

3. The multi-topology routing dynamic management method based on traffic demand perception according to claim 1 is characterized in that: The calculating and adjusting probability comprises: Calculate burst traffic increment; Determine the latency overhead incurred by the controller-switch interaction during the adjustment process; Determine the constant coefficient corresponding to traffic changes and adjustment overhead; The adjustment probability is obtained by performing calculation based on the burst traffic increment, the delay overhead generated by the interaction between the controller and the switch during the adjustment process, the traffic change, and a constant coefficient corresponding to the adjustment overhead.

4. The multi-topology routing dynamic management method based on traffic demand perception according to claim 1 is characterized in that: The adjustment probability satisfies the following formula: Among them, P adjust represents the adjustment probability, Δtraffic represents the burst traffic increment, L CP represents the delay overhead caused by the interaction between the controller and the switch during the adjustment process. A, B, and C are constant coefficients related to traffic changes and adjustment overhead.

5. The multi-topology routing dynamic management method based on traffic demand perception according to claim 3 is characterized in that: The delay overhead caused by the interaction between the controller and the switch includes: Determine controller processing delays; Determine the propagation delay between the controller and the switch; Determine synchronization delays between multiple controllers; The delay overhead generated by the interaction between the controller and the switch is obtained by calculation based on the controller processing delay, the propagation delay between the controller and the switch, and / or the synchronization delay between the multiple controllers.

6. The multi-topology routing dynamic management method based on traffic demand perception according to claim 3 is characterized in that: The delay overhead generated by the interaction between the controller and the switch satisfies the following formula: in, Indicates controller processing delay, represents the propagation delay between the controller and the switch, L syn Indicates the synchronization delay between multiple controllers.

7. The multi-topology routing dynamic management method based on traffic demand perception according to any one of claims 1 to 6, characterized in that: After dynamically adjusting the multi-topology routing according to the adjustment probability, the following steps are further performed: Determining whether the real-time traffic data is less than a traffic threshold; If so, the initial multi-topology routing is restored.

8. The multi-topology routing dynamic management method based on traffic demand perception according to claim 1 is characterized in that: The method further comprises: Collect historical traffic data; Analyzing the historical traffic data and training the data using a long short-term memory network to obtain a traffic prediction model; The multi-topology routing is dynamically adjusted based on the traffic prediction model.

9. A multi-topology routing dynamic management system based on traffic demand perception, characterized in that: include: A traffic collection unit is used to collect real-time traffic of each access point in the current network in real time and obtain real-time traffic data corresponding to each access point; A condition judgment unit, configured to judge whether a dynamic adjustment condition is met based on the real-time traffic data; A probability calculation unit, configured to calculate an adjustment probability when the dynamic adjustment condition is met; A dynamic adjustment unit is used to dynamically adjust the multi-topology routing according to the adjustment probability.

10. The multi-topology routing dynamic management system based on traffic demand perception according to claim 9, characterized in that: Also includes: The feedback optimization unit is used to monitor whether the real-time traffic data is less than a traffic threshold, and restore the initial multi-topology routing when the real-time traffic data is less than the traffic threshold.

Images