SDN-oriented virtual network topology dynamic construction method

Abstract

The invention discloses an SDN (Software Defined Network)-oriented virtual network topology dynamic construction method, which comprises the following steps of: 1, acquiring multi-modal network data in a teaching scene in real time through an embedded probe, and constructing a multi-modal topology sensing matrix; step 2, inputting the multi-modal topology perception matrix into a pre-trained deep decision tree model to generate a dynamic topology descriptor; 3, converting the dynamic topology descriptor into a standardized configuration instruction set compatible with multiple protocols; step 4, dynamically adjusting the density parameters of the virtual nodes according to the stage characteristics of the teaching process, and realizing progressive reconstruction from basic single-domain topology to complex cross-domain topology; and 5, triggering a topology self-healing mechanism when detecting that the entropy value is suddenly changed, and synchronously updating the abnormal operation mode feature library in the knowledge graph. Through cooperative work of multiple links, the efficient, intelligent and flexible virtual network topology dynamic construction method is provided for a teaching scene, and the teaching effect and the network management efficiency are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of SDN virtual network topology construction, and in particular to an SDN-oriented virtual network topology dynamic construction method. Background Art

[0002] With the development of networks, virtualization technology has become a research hotspot. The current Internet faces problems such as "rigidity," severe waste of hardware resources, difficult maintenance, and complex network device configuration. The concept of network virtualization has emerged as a response. Using abstraction, allocation, and isolation mechanisms, it builds diverse, coexisting yet isolated virtual networks by sharing underlying physical resources. This allows for the rational allocation and management of physical resources and the construction of logical networks based on existing network architectures to meet specific business needs, such as virtual local area networks (VLANs) and virtual private networks (VPNs).

[0003] Software-defined networking (SDN), a new network architecture, separates the network control plane from the data plane. It uses a centralized controller and standard interfaces to manage network devices, creating an open, programmable network structure that facilitates new network architectures and service innovation. SDN also possesses inherent advantages in network virtualization, simplifying network design and operation. It allows network administrators to centrally control a large number of network devices, eliminating the need for manual configuration of each device.

[0004] When it comes to building virtual network topologies, network simulation platforms aim to provide users with near-realistic virtual network environments for conducting cybersecurity experiments and improving security skills. However, existing technologies are insufficient for dynamically building and modifying virtual network topologies, failing to adequately meet users' needs for modifying and building upon existing topologies. For example, in teaching scenarios, network topologies need to be dynamically adjusted to simulate various complex network environments based on varying teaching content and experimental requirements. However, existing network topology construction methods struggle to adapt to the dynamic and complex changes in teaching scenarios. Summary of the Invention

[0005] In view of this, the present invention proposes a method for dynamically constructing virtual network topology for SDN to solve the problem in the prior art that dynamic construction of network topology in teaching scenarios is difficult to adapt to the dynamic and complex changes of teaching scenarios.

[0006] The specific technical solutions of the present invention are as follows:

[0007] A method for dynamically constructing a virtual network topology for SDN, comprising:

[0008] Step 1: Use embedded probes to collect multimodal network data in the teaching scene in real time and construct a multimodal topology perception matrix. The multimodal network data includes network cognitive level data flow, device instruction operation feature vectors, and protocol interaction state matrix.

[0009] Step 2: Based on the dynamic reconstruction module of the adaptive rule engine, the multimodal topology perception matrix constructed in step 1 is input into the pre-trained deep decision tree model to generate a dynamic topology descriptor including node association parameters, link time-varying weight matrix, and protocol stack configuration vector;

[0010] Step 3: The dynamic topology descriptor generated in step 2 is converted into a standardized configuration instruction set compatible with OpenFlow, Restconf, and Netconf protocols through the protocol abstraction middleware of the heterogeneous controller interface adaptation layer.

[0011] Step 4: Based on the real-time topology calculation engine, an incremental construction algorithm with time window constraints is used to dynamically adjust the virtual node density parameters according to the characteristics of the teaching process stage, achieving progressive reconstruction from basic single-domain topology to complex cross-domain topology;

[0012] Step 5: Build a process visualization feedback loop, calculate the network state complexity index in real time through the topology entropy monitoring module, trigger the topology self-healing mechanism when an entropy mutation is detected, and synchronously update the abnormal operation mode feature library in the knowledge graph.

[0013] Furthermore, in step 1, the network cognitive level data flow is generated by comprehensively analyzing the students' operation steps, task completion time, accuracy and operation fluency indicators in the network experiment; the device instruction operation feature vector is generated by analyzing the instruction type, frequency and abnormal operation mark received by the router and switch; the protocol interaction state matrix is generated by dynamically analyzing the interaction messages between the control plane and the data plane, including the protocol type, interaction frequency and abnormal protocol behavior mark.

[0014] Furthermore, in step 1, let the network cognitive level data flow be D c , the device instruction operation feature vector is D0, and the protocol interaction state matrix is D p ;D c ∈R m×n , where m represents the number of students, n represents the number of indicators for measuring network cognition level; D0∈R k , k represents the number of device instruction operation features; D p ∈R l×l , l represents the number of network protocols, and the matrix element D p (i, j) represents the interaction state between the i-th protocol and the j-th protocol; the multimodal topology perception matrix M is constructed by the following formula:

[0015]

[0016] Among them, 0 a×bis a zero matrix with a rows and b columns. vec(·) is a vectorized operation that stacks matrices into vectors by columns.

[0017] Furthermore, the dynamic topology descriptor generated by the deep decision tree model in step 2 includes: node association parameters, which are dynamically updated according to the communication dependency between nodes and are used to optimize node connection planning; link time-varying weight matrix, which dynamically adjusts the weight value according to the link congestion or load capacity during the teaching process to guide traffic scheduling; protocol stack configuration vector, which records device protocol parameters to ensure the compatibility and communication continuity of the TCP / IP protocol stack.

[0018] Furthermore, in step 2, the multimodal topology perception matrix M is input into the pre-trained deep decision tree model f to generate a dynamic topology descriptor T = [R0, W(t), C], where the node association parameter It is the part used to generate node association parameters in the deep decision tree model, R0∈R s×s , s represents the number of network nodes, R0(i, j) represents the association between node i and node j; the link time-varying weight matrix W(t) = f W (M, t), f W It is the part of the deep decision tree model used to generate the time-varying weight matrix of the link, t represents time, W(t)∈R e×e , e represents the number of network links, W(t)(i, j) represents the weight between link i and link j at time t; protocol stack configuration vector C = f C (M), f C It is the part of the deep decision tree model used to generate the protocol stack configuration vector, C∈R q , q represents the number of protocol stack configuration parameters.

[0019] Furthermore, in step 3, the protocol abstraction middleware implements a unified semantic expression of multi-protocol interfaces using a hierarchical mapping mechanism, including: the semantic abstraction layer converts the dynamic topology descriptor into a protocol-independent intermediate representation; the protocol adaptation layer maps the intermediate representation into OpenFlow flow table entries, Restconf configuration templates, or Netconf XML instructions according to the target controller type.

[0020] Furthermore, the incremental construction algorithm in step 4 includes: a time window segmentation module, which divides the teaching process into basic stage, transition stage and advanced stage windows; a virtual node density parameter adjustment module, which sets the single-domain topology low density value or the cross-domain topology high density value according to the stage characteristics; and a progressive reconstruction module, which only updates the newly added or deleted virtual nodes and links within the current time window.

[0021] Furthermore, in step 4, let the time window be [t1, t2]. Within this time window, the incremental construction algorithm only processes the network topology information related to this time window. According to different teaching stages, the virtual node density parameter ρ(t) is dynamically adjusted as follows:

[0022]

[0023] Among them, ρ base It is the benchmark for calculating the virtual node density at all teaching stages, reflecting the topological complexity at the initial stage of teaching; ΔS is the increment of teaching progress; ΔC is the complexity of concurrent operations; α and β are adaptive adjustment coefficients.

[0024] Furthermore, in step 5, the topology self-healing mechanism includes dual modes: link redundancy rerouting and protocol stack dynamic rollback: link redundancy rerouting mode, when a link interruption is detected, it automatically switches to the backup link; protocol stack dynamic rollback mode, when the protocol configuration is abnormal, it restores the protocol parameters to the stable version.

[0025] Furthermore, in step 5, the topological entropy value E is calculated using the following formula:

[0026]

[0027] Where N is the number of network states, p i is the probability of the i-th network state occurring; when a topology entropy mutation |E(t)-E(t-Δt)|>∈ is detected, the topology self-healing mechanism is triggered, where ∈ is a preset threshold.

[0028] The beneficial effects of the present invention are:

[0029] 1. Using embedded probes to collect multimodal network data from teaching scenarios in real time, a multimodal topology perception matrix is constructed to capture dynamic information in a timely manner and accurately reflect the actual network operation. This integration of network cognitive level data streams, device command operation feature vectors, and protocol interaction state matrices provides rich data support for network topology construction, performance analysis, and fault diagnosis.

[0030] 2. Based on the dynamic reconstruction module of the adaptive rule engine, the multimodal topology perception matrix is input into a pre-trained deep decision tree model to generate a dynamic topology descriptor. This descriptor contains node association parameters, a time-varying link weight matrix, and a protocol stack configuration vector. The adaptive rule engine automatically adjusts the decision-making process to adapt the topology descriptor to new network conditions, optimizing the network topology to meet different teaching needs.

[0031] 3. The protocol abstraction middleware of the heterogeneous controller interface adaptation layer converts the dynamic topology descriptor into a standardized configuration instruction set compatible with multi-protocol interfaces, enabling seamless docking and collaborative work between different protocols.

[0032] 4. The real-time topology calculation engine uses an incremental construction algorithm with time window constraints. It dynamically adjusts virtual node density parameters based on the characteristics of the teaching process stage, achieving progressive reconstruction from basic single-domain topology to complex cross-domain topology. This improves topology construction efficiency, reduces computing resource consumption, and adapts to actual teaching, making the teaching process more scientific, reasonable, and efficient.

[0033] 5. Build a process visualization feedback loop. The topology entropy monitoring module calculates network state complexity indicators in real time. Sudden changes in entropy indicate network anomalies, automatically initiating a topology self-healing mechanism. This dual-mode system, link redundancy rerouting and protocol stack dynamic rollback, enables rapid network repair. Abnormal operation patterns are also updated in the knowledge graph feature library, providing a reference for subsequent fault diagnosis and self-healing, improving network reliability and stability and ensuring smooth teaching. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments or descriptions of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0035] Figure 1 This is a flow chart of a method for dynamically constructing a virtual network topology for SDN according to the present invention;

[0036] Figure 2 This is a structural diagram of the method for dynamically constructing virtual network topology for SDN in the present invention. DETAILED DESCRIPTION

[0037] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0038] The present invention proposes a method for dynamically constructing virtual network topology for SDN. Figure 1-2 As shown, the method includes:

[0039] Step 1: Use embedded probes to collect multimodal network data in real time during the teaching scenario and construct a multimodal topology perception matrix. Using embedded probes distributed across key nodes in the network, three types of multimodal network data are collected in real time during the teaching scenario: network cognitive level data streams, device command operation feature vectors, and protocol interaction state matrices.

[0040] The teaching environment is constantly changing, and student operations and network status are constantly evolving depending on factors such as the teaching process, learning needs, and external interference. Real-time multimodal data collection can capture these changes promptly, allowing the constructed topology to dynamically adapt to changes in the teaching environment and better reflect the actual network operation. For example, when students perform new operations or encounter network failures during a network experiment, the topology awareness matrix can quickly update data, providing network administrators with the latest network status information so that they can take appropriate measures.

[0041] The specific implementation form of the embedded probe can flexibly select hardware devices or software modules according to the network architecture and monitoring requirements of the teaching scenario. At the hardware device level, the embedded probe can be designed as a miniaturized network monitoring device, integrating a sensor module and an embedded processor, and accessing key network nodes through a physical interface (such as an Ethernet interface, a USB interface or a wireless communication module), such as the management port deployed on a router or switch or the serial interface of an experimental terminal. This type of hardware probe has the ability to capture real-time data and can accurately collect the underlying signals in the device instruction operation feature vector, such as by monitoring the console instruction stream or parsing the device firmware log to extract parameters such as instruction type, execution frequency and response time. Its advantage lies in the ability to directly access hardware layer operations, and it can capture the physical details of students operating the device in teaching experiments, such as the input timing of configuration commands, the device restart cycle, etc., providing a high-precision data source for device status analysis.

[0042] The implementation of software modules focuses on protocol parsing and behavioral analysis, and is usually embedded in network operating systems, experimental platform virtual machines, or teaching management software in the form of lightweight agents or virtual probes. This type of probe dynamically constructs a protocol interaction state matrix by calling system APIs or intercepting network stack data packets. For example, in a virtualized teaching environment, software probes can be deployed at the hypervisor layer to parse the protocol header information of inter-virtual machine communications in real time, and generate a dynamic view of protocol interactions based on traffic statistics. At the same time, software probes can integrate with the collection function of network cognitive level data streams, record the interface interaction events of students' operations through integration with online experimental platforms, and convert behavioral data into cognitive ability assessment indicators through knowledge graph mapping technology.

[0043] The network cognition level data stream reflects students' understanding and use of the Internet. It is an important indicator for measuring students' network knowledge absorption and practical operation ability. By analyzing various data such as students' operation steps, question answers, mastery of network concepts, and proficiency in operating network equipment in network experiments, their network cognition level is quantified. For example, based on indicators such as the time, accuracy, and fluency of students' operations in completing specific tasks in network experiments, their network cognition ability is comprehensively evaluated.

[0044] The device instruction operation feature vector records the various instructions and their characteristics received by network devices (such as routers, switches, etc.). By analyzing the device instruction operation feature vector, we can clearly understand the device's operating mode, such as whether the device is in normal operation or has undergone special configuration; grasp the frequency of device usage and determine whether the device is overloaded or idle; and promptly detect whether there are abnormal operations, such as unauthorized command input or frequent erroneous commands, which helps to analyze the device's operating status and user operating habits, and provide accurate guidance for device maintenance and management.

[0045] The protocol interaction state matrix describes the interaction between different network protocols. By dynamically parsing the interaction messages between the control plane and the data plane, it can intuitively display the interaction process and status of the protocols in the network. Specifically, the control plane is responsible for policy decisions (such as the OpenFlow controller issuing flow table rules), while the data plane is responsible for executing forwarding actions (such as the switch processing data packets). The protocol messages continuously exchanged between the two (such as LLDP discovery messages, flow table update requests / responses, etc.) carry key information about the network's operating status. By real-time monitoring of the protocol interaction status, abnormal protocol behaviors in the network, such as protocol conflicts and illegal protocol interactions, can be discovered in a timely manner. For example, when data transmission confusion or protocol message format errors are detected between different protocols, the problem can be quickly located to ensure the stable operation of the network and avoid network failures and data loss caused by protocol problems.

[0046] The network cognitive level data stream, device command operation feature vectors, and protocol interaction state matrix collected are deeply integrated to construct a multimodal topology perception matrix. This matrix comprehensively reflects the comprehensive state of the network in the teaching scenario, including user behavior, device operation status, and protocol interaction status. It provides rich data support for subsequent network topology construction, performance analysis, and fault diagnosis.

[0047] The following is an example to illustrate the construction process of the multimodal topology perception matrix: Let the network cognitive level data flow be D c , the device instruction operation feature vector is D0, and the protocol interaction state matrix is D p . D c ∈R m×n , where m represents the number of students and n represents the number of indicators for measuring network cognition level. For example, n indicators may include the time required to complete a specific task, accuracy, operation fluency, etc. D0∈R k , k represents the number of device instruction operation characteristics, such as instruction type, execution frequency, response time, etc. p ∈R l×l , l represents the number of network protocols, and the matrix element D p(i, j) represents the interaction state between the i-th protocol and the j-th protocol. The multimodal topology perception matrix M can be constructed by the following formula:

[0048]

[0049] Among them, 0 a×b is a zero matrix with a rows and b columns. vec(·) is a vectorized operation that stacks matrices into vectors by columns.

[0050] In step 2, the dynamic reconstruction module, based on the adaptive rule engine, inputs the constructed multimodal topology perception matrix into a pre-trained deep decision tree model to generate dynamic topology descriptors, including node association parameters, link time-varying weight matrices, and protocol stack configuration vectors. These descriptors accurately describe the dynamic characteristics of the network topology and provide a basis for subsequent topology construction and configuration decisions.

[0051] The dynamic reconstruction module, based on an adaptive rule engine, first uses the constructed multimodal topology perception matrix as input data and imports it into a pre-trained deep decision tree model. Trained with extensive historical network data, the deep decision tree model possesses powerful pattern recognition and decision-making capabilities, accurately learning and establishing a mapping between network status and topology. Simultaneously, the module's adaptive rule engine continuously monitors various changes in network status in real time, such as fluctuations in network traffic and changes in link performance, to prepare for potential subsequent decision adjustments.

[0052] The deep decision tree model generates a dynamic topology descriptor based on the data features in the multimodal topology perception matrix, which includes the following key parameters:

[0053] The node correlation parameter measures the closeness of connections between nodes in a network and directly reflects the interdependence between nodes during communication. For example, in a data center network, some server nodes frequently interact with each other, resulting in relatively high correlation parameter values. This clearly indicates a strong communication dependency between these nodes. This parameter provides insight into the communication patterns between nodes in the network, providing a strong basis for rationally planning node connections during subsequent topology construction, ensuring efficient and smooth network communication.

[0054] Considering that link status and performance change dynamically over time during actual network operation, particularly during teaching sessions where traffic bursts may occur, the time-varying link weight matrix dynamically adjusts the weight of each link based on these real-time characteristics. The weight reflects characteristics such as the link's importance or load capacity. For example, when a link is congested, its weight decreases, indicating to subsequent traffic scheduling mechanisms that it is not suitable for carrying excessive traffic. Conversely, when a link is lightly loaded and performing well, its weight increases, prioritizing it for subsequent traffic scheduling. This time-varying link weight matrix more accurately reflects the actual network operation, providing critical information for topology optimization and enabling more efficient allocation and utilization of network resources.

[0055] The protocol stack configuration vector records the configuration information of different protocol stacks in the network. In network communications, various protocol stacks need to be correctly configured to ensure the normal operation of network protocols and effective interaction between different devices. In a network that contains multiple network devices (such as routers, switches, servers, etc.), different devices may use different protocol stacks. The protocol stack configuration vector records in detail the specific configuration parameters of the protocol stack on each device, such as the IP address, subnet mask, gateway settings in the TCP / IP protocol stack, and specific parameters of each protocol layer. This ensures the correct operation of network protocols and smooth interaction between different devices, providing a solid guarantee for the stability and functionality of the network topology, and avoiding network failures or communication interruptions caused by protocol configuration errors.

[0056] Node affinity parameters, link time-varying weight matrices, and protocol stack configuration vectors together constitute the dynamic topology descriptor. These parameters comprehensively and dynamically describe the characteristics of the network topology, providing a solid basis for subsequent topology configuration and adjustment. For example, during network topology optimization, node affinity parameters can be used to adjust the connection between nodes, making the network structure more rational and improving network communication efficiency. The link time-varying weight matrix can be used to rationally distribute traffic, avoid link congestion, and ensure efficient network operation. The protocol stack configuration vector can be used to ensure the normal operation of network protocols, safeguard the integrity of network functions, and enable the network to stably support various business needs.

[0057] The adaptive rules engine plays a key role in automatically adjusting the decision-making process throughout the entire process. When it detects a change in network status, such as a sudden increase in network traffic, it automatically triggers a decision adjustment mechanism. At this point, it adjusts the time-varying link weight matrix in real time, ensuring that the weights more accurately reflect the current link status. This dynamic adjustment allows the generated topology descriptor to better adapt to new network conditions, thereby optimizing the network topology and ensuring that the network maintains good performance under varying teaching needs and network conditions, meeting the requirements of various practical application scenarios.

[0058] The following is an example to illustrate the generation process of dynamic topology descriptors: the multimodal topology perception matrix M is input into the pre-trained deep decision tree model f to generate a dynamic topology descriptor T = [R0, W(t), C], where the node association parameter It is the part used to generate node association parameters in the deep decision tree model, R0∈R s ×s , s represents the number of network nodes, R0(i, j) represents the association between node i and node j; the link time-varying weight matrix W(t) = f W (M, t), f W It is the part of the deep decision tree model used to generate the time-varying weight matrix of the link, t represents time, W(t)∈R e×e , e represents the number of network links, W(t)(i, j) represents the weight between link i and link j at time t; protocol stack configuration vector C = f C (M), f C It is the part of the deep decision tree model used to generate the protocol stack configuration vector, C∈R q , q represents the number of protocol stack configuration parameters.

[0059] Step 3: The protocol abstraction middleware at the heterogeneous controller interface adaptation layer converts the dynamic topology descriptor into a standardized configuration instruction set compatible with multi-protocol interfaces such as OpenFlow, Restconf, and Netconf to adapt to different types of controllers. The protocol abstraction middleware uses a layered mapping mechanism to achieve a unified semantic expression for multi-protocol interfaces.

[0060] The role of the heterogeneous controller interface adaptation layer is to ensure that generated configuration instructions are correctly executed on different controllers and network devices, enabling seamless integration and interoperability between different protocols. Because different network controllers in an SDN network may support different protocol interfaces, such as OpenFlow, Restconf, and Netconf, and protocol abstraction middleware can mask these differences, protocol abstraction middleware is introduced to convert and adapt dynamic topology descriptors, transforming them into a standardized set of configuration instructions compatible with multiple protocol interfaces. This allows any type of network controller to easily receive and process these configuration instructions, enabling dynamic adjustment of the virtual network topology.

[0061] The protocol abstraction middleware constructs a syntax-semantics decoupling architecture through a layered mapping mechanism, enabling unified expression of multi-protocol interfaces. Its semantic abstraction layer extracts topology intent based on the YANG model and generates intermediate instructions, shielding against protocol differences. The protocol adaptation layer dynamically compiles intermediate instructions into polymorphic configuration flows based on the target protocol's message format and interaction paradigm. This architecture utilizes a two-level mapping table, a semantic logic table, and a protocol syntax template library to efficiently convert topology descriptors into OpenFlow flow tables, Restconf RPC, and Netconf configuration blocks, ensuring consistent expression of northbound semantics while maintaining southbound interface differentiation.

[0062] Step 4: Based on the real-time topology calculation engine, an incremental construction algorithm with time window constraints is used to dynamically adjust the virtual node density parameter based on the characteristics of the teaching process stage, achieving progressive reconstruction from basic single-domain topology to complex cross-domain topology. The virtual node density parameter is nonlinearly correlated with the concurrent operation complexity of the teaching scenario at advanced stages of teaching.

[0063] Specifically, the real-time topology calculation engine accurately divides the timeline into multiple time windows of different lengths, each of which has its own specific time span and processing tasks. In each independent time window, the incremental construction algorithm only focuses on processing and updating the network topology information that is closely related to the current time window. For example, it only processes network connections that have changed in the current time window, new or deleted virtual nodes, etc., while ignoring stable topology information that has been processed in other time windows. This refined processing method cleverly avoids the frequent and time-consuming comprehensive reconstruction of the entire large and complex network topology, greatly improving the efficiency of topology construction and significantly shortening the time required for topology updates, so that the network topology can quickly respond to changes in the teaching environment. The incremental construction algorithm only updates the changed parts, avoiding the reconstruction of the entire topology, improving the efficiency of topology construction, and reducing the consumption of computing resources.

[0064] The incremental construction algorithm also deeply integrates time factors and dynamic changes in network status. By continuously monitoring the passage of time during the teaching process and real-time fluctuations in network status, such as increases and decreases in network traffic and fluctuations in node performance, it can more accurately and flexibly adjust the virtual node density parameter based on the unique characteristics of different stages of the teaching process, such as the speed of teaching progress and the shift in teaching focus. The virtual node density parameter, as a key quantitative indicator measuring the density of virtual nodes in the virtual network topology, directly and profoundly determines the complexity of the topology. It is not fixed but is dynamically and finely adjusted according to the stage characteristics of the teaching process.

[0065] The requirements for network topology complexity vary significantly at different teaching stages. During the initial phase of teaching, when students are first introduced to basic network knowledge, they lack an intuitive understanding of complex network structures and concepts. The real-time topology calculation engine sets the virtual node density parameter to a low value based on this teaching stage. This parameter setting creates a simple, clear single-domain topology with a relatively small number of nodes and links, and the connections between nodes are clearly understood. This provides students with a clear and accessible starting point for network learning, allowing them to quickly grasp network fundamentals and establish a basic cognitive framework, laying a solid foundation for subsequent in-depth study.

[0066] As instruction progresses, a transition phase begins. Students have mastered basic networking knowledge and gained a preliminary understanding of networking, eager to explore more advanced networking skills. The real-time topology calculation engine continuously monitors the characteristics of the teaching process, closely aligning it with students' learning progress and actual needs. It collects data on various aspects, including in-class performance, homework completion, and experimental results, to accurately assess students' learning status and needs. Based on this data, the engine dynamically adjusts virtual node density parameters, gradually increasing the number and distribution of virtual nodes as instruction progresses. Building on the basic single-domain topology, new nodes and links are cleverly added to gradually increase the complexity of the topology. These newly added nodes and links guide students to gradually master deeper networking knowledge and skills, enhancing their network awareness and operational capabilities, and enabling them to continuously consolidate and deepen their knowledge through practical application.

[0067] As instruction progresses to advanced levels, students will need to gain exposure to and master knowledge in complex network environments to better adapt to the network demands of future real-world work. Complex cross-domain topologies simulate real-world environments, such as large enterprise networks and internet data centers. These environments consist of multiple distinct network domains, each with its own specific functions and security policies, with widely distributed nodes and complex link relationships. The real-time topology calculation engine adjusts the virtual node density parameter to a higher value based on the characteristics of the current instructional stage. By adding a large number of nodes and links to the topology, a network topology with multiple domains, a widely distributed node structure, and complex link relationships is meticulously constructed. This complex cross-domain topology provides students with a practical scenario that closely resembles a real-world network environment, allowing them to experience various challenges and problems in complex network environments, such as network congestion, security vulnerabilities, and routing. By solving these problems, students can effectively enhance their ability to solve complex network problems, cultivate their innovative thinking and practical skills, and lay a solid foundation for their career development.

[0068] As teaching progresses and complexity increases, the adjustment of virtual node density parameters doesn't follow a simple linear growth pattern. Linear growth means the number of nodes increases at a fixed rate or rate as the teaching phase progresses. However, this simple adjustment approach fails to meet the diverse needs of actual teaching. In practice, virtual node density parameters are adjusted more flexibly and precisely based on actual teaching needs and scenario characteristics. For example, during the basic network knowledge introduction phase, only a small number of nodes may be needed to construct a simple topology. These nodes are sufficient to demonstrate basic network concepts and principles, such as clients, servers, and routers. However, during the complex network application practice phase, as teaching complexity increases, the number of nodes increases significantly, but the rate and pattern of increase are not linear. For example, when engaging in network security practice, a large number of simulated attack and defense nodes may be suddenly added. When studying network performance optimization, a large number of nodes may be added to simulate different traffic loads and device performance. This nonlinear adjustment approach is determined based on specific teaching objectives and content to better adapt to the teaching objectives of different stages, ensuring that students receive the most appropriate learning experience at each stage, making the teaching process more scientific, reasonable, and efficient.

[0069] The following example illustrates the process of incrementally reconstructing the virtual network topology: Assume a time window [t1, t2]. Within this time window, the incremental construction algorithm only processes the network topology information related to this time window. Depending on the teaching stage, such as the basic stage, transition stage, or advanced stage, the virtual node density parameter ρ(t) is dynamically adjusted. The adjustment rules are as follows:

[0070]

[0071] Among them, ρ base This is the benchmark for calculating virtual node density across all teaching phases, reflecting the topological complexity of the initial teaching phase. ΔS is the incremental teaching progress. ΔC is the concurrent operation complexity. α and β are adaptive adjustment coefficients.

[0072] Step 5: Build a process visualization feedback loop, calculate the network state complexity index in real time through the topology entropy monitoring module, trigger the topology self-healing mechanism when an entropy mutation is detected, and synchronously update the abnormal operation mode feature library in the knowledge graph.

[0073] Topological entropy is an indicator of the complexity and disorder of a network structure, providing a direct reflection of the network's operational status. Higher entropy values indicate greater network complexity and disorder. For example, in a simple star network, where node connections are relatively simple and data flows are relatively clear, the topological entropy may be low. In contrast, in a complex mesh network, where nodes have multiple connection paths and data transmission paths are diverse and complex, the topological entropy value is relatively high.

[0074] Anomaly detection relies heavily on data provided by the topology entropy monitoring module. This determination is based on continuous monitoring of topology entropy changes. When a significant, short-term change in entropy that falls outside the normal range is detected, known as a sudden change in entropy, the network is considered to have experienced an anomaly. This sudden change can be caused by a variety of factors. For example, device failure is a common cause. Hardware issues with a router or switch can lead to data interruptions or errors, which in turn can cause a sudden change in topology entropy. Link outages are also a concern. Accidentally unplugging a network cable or damaging a fiber optic line can disrupt normal network connectivity and cause entropy fluctuations. Furthermore, network attacks can also cause anomalies. Malware can generate large amounts of abnormal traffic, disrupting the network and causing sudden changes in topology entropy.

[0075] When the topology entropy monitoring module detects a sudden increase in entropy and confirms a network anomaly, the topology self-healing mechanism automatically activates, attempting to repair the network topology and restore it to normal. This mechanism features dual modes: link redundancy rerouting and protocol stack dynamic rollback to address various network anomalies. When a primary link fails, causing a sudden increase in entropy, the system quickly and automatically selects available redundant links to reroute data traffic. For example, core switches are typically connected by multiple fiber optic links. If one link fails, the system quickly switches data traffic to a functioning link, ensuring continued data transmission, avoiding network interruptions caused by the link failure, and ensuring business continuity. If the network anomaly is caused by a configuration error or software vulnerability in the network protocol stack, the protocol stack dynamic rollback mode automatically activates. For example, after a network device upgrade, network anomalies may occur due to compatibility issues. In this case, this feature automatically rolls back the protocol stack to its previously stable state, eliminating the anomaly caused by the protocol issue and quickly restoring normal network functionality, minimizing network downtime caused by the protocol failure.

[0076] When triggering the topology self-healing mechanism, the system records the detected abnormal operation patterns in detail and synchronously updates them to the abnormal operation pattern feature library in the knowledge graph. As a powerful structured model specifically designed to represent and organize knowledge, the knowledge graph stores a large amount of characteristic information about normal network operation and abnormal conditions. By continuously accumulating and updating the abnormal operation pattern feature library, the system can continuously accumulate experience with network failures and abnormal operations. In subsequent network operations, when similar problems are encountered, the system can rely on the relevant characteristic information stored in the knowledge graph to more quickly identify and handle abnormal situations, providing a strong reference for fault diagnosis and self-healing, thereby further improving the reliability and stability of the network and ensuring the smooth progress of teaching.

[0077] The following is an example to illustrate the process of topology entropy monitoring and self-healing: The topology entropy value E can be calculated using the following formula:

[0078]

[0079] Where N is the number of network states, p i is the probability of the i-th network state occurring. When a sudden change in the topology entropy value is detected, that is, |E(t)-E(t-Δt)|>∈ (∈ is a preset threshold), the topology self-healing mechanism is triggered.

[0080] The beneficial effects of the present invention are:

[0081] 1. Using embedded probes to collect multimodal network data from teaching scenarios in real time, a multimodal topology perception matrix is constructed to capture dynamic information in a timely manner and accurately reflect the actual network operation. This integration of network cognitive level data streams, device command operation feature vectors, and protocol interaction state matrices provides rich data support for network topology construction, performance analysis, and fault diagnosis.

[0082] 2. Based on the dynamic reconstruction module of the adaptive rule engine, the multimodal topology perception matrix is input into a pre-trained deep decision tree model to generate a dynamic topology descriptor. This descriptor contains node association parameters, a time-varying link weight matrix, and a protocol stack configuration vector. The adaptive rule engine automatically adjusts the decision-making process to adapt the topology descriptor to new network conditions, optimizing the network topology to meet different teaching needs.

[0083] 3. The protocol abstraction middleware of the heterogeneous controller interface adaptation layer converts the dynamic topology descriptor into a standardized configuration instruction set compatible with multi-protocol interfaces, enabling seamless docking and collaborative work between different protocols.

[0084] 4. The real-time topology calculation engine uses an incremental construction algorithm with time window constraints. It dynamically adjusts virtual node density parameters based on the characteristics of the teaching process stage, achieving progressive reconstruction from basic single-domain topology to complex cross-domain topology. This improves topology construction efficiency, reduces computing resource consumption, and adapts to actual teaching, making the teaching process more scientific, reasonable, and efficient.

[0085] 5. Build a process visualization feedback loop. The topology entropy monitoring module calculates network state complexity indicators in real time. Sudden changes in entropy indicate network anomalies, automatically initiating a topology self-healing mechanism. This dual-mode system, link redundancy rerouting and protocol stack dynamic rollback, enables rapid network repair. Abnormal operation patterns are also updated in the knowledge graph feature library, providing a reference for subsequent fault diagnosis and self-healing, improving network reliability and stability and ensuring smooth teaching.

[0086] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for dynamically constructing a virtual network topology for SDN, characterized in that: include: Step 1: Using embedded probes to collect multimodal network data in the teaching scene in real time, and construct a multimodal topology perception matrix. The multimodal network data includes network cognitive level data flow, device instruction operation feature vectors, and protocol interaction state matrix. Step 2: Based on the dynamic reconstruction module of the adaptive rule engine, the multimodal topology perception matrix constructed in step 1 is input into the pre-trained deep decision tree model to generate a dynamic topology descriptor including node association parameters, link time-varying weight matrix, and protocol stack configuration vector; Step 3: The dynamic topology descriptor generated in step 2 is converted into a standardized configuration instruction set compatible with OpenFlow, Restconf, and Netconf protocols through the protocol abstraction middleware of the heterogeneous controller interface adaptation layer. Step 4: Based on the real-time topology calculation engine, an incremental construction algorithm with time window constraints is used to dynamically adjust the virtual node density parameters according to the characteristics of the teaching process stage, achieving progressive reconstruction from basic single-domain topology to complex cross-domain topology; Step 5: Build a process visualization feedback loop, calculate the network state complexity index in real time through the topology entropy monitoring module, trigger the topology self-healing mechanism when an entropy mutation is detected, and synchronously update the abnormal operation mode feature library in the knowledge graph.

2. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 1, the network cognition level data flow is generated by comprehensively and quantitatively analyzing the students' operation steps, task completion time, accuracy and operation fluency indicators in the network experiment; the device instruction operation feature vector is generated by analyzing the instruction type, frequency and abnormal operation mark received by the router and switch; the protocol interaction state matrix is generated by dynamically analyzing the interaction messages between the control plane and the data plane, including the protocol type, interaction frequency and abnormal protocol behavior mark.

3. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 1, let the network cognitive level data flow be D c , the device instruction operation feature vector is D0, and the protocol interaction state matrix is D p ;D c ∈R m×n , where m represents the number of students, n represents the number of indicators for measuring network cognition level; D0∈R k , k represents the number of device instruction operation features; D p ∈R l×l , l represents the number of network protocols, and the matrix element D p (i, j) represents the interaction state between the i-th protocol and the j-th protocol; the multimodal topology perception matrix M is constructed by the following formula: Among them, 0 a×b is a zero matrix with a rows and b columns. vec(·) is a vectorized operation that stacks matrices into vectors by columns.

4. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: The dynamic topology descriptor generated by the deep decision tree model in step 2 includes: a node association parameter, which is dynamically updated based on the communication dependency between nodes and is used to optimize node connection planning; a link time-varying weight matrix, which dynamically adjusts the weight value according to the link congestion or load capacity during the teaching process to guide traffic scheduling; and a protocol stack configuration vector, which records device protocol parameters to ensure the compatibility and communication continuity of the TCP / IP protocol stack.

5. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 2, the multimodal topology perception matrix M is input into the pre-trained deep decision tree model f to generate a dynamic topology descriptor T = [R0, W(t), C], where the node association parameter It is the part used to generate node association parameters in the deep decision tree model, R0∈R s×s , s represents the number of network nodes, R0(i, j) represents the association degree between node i and node j; Link time-varying weight matrix W(t) = f W (M, t), f W It is the part of the deep decision tree model used to generate the time-varying weight matrix of the link, t represents time, W(t)∈R e×e , e represents the number of network links, W(t)(i, j) represents the weight between link i and link j at time t; Protocol stack configuration vector C = f C (M), f C It is the part of the deep decision tree model used to generate the protocol stack configuration vector, C∈R q , q represents the number of protocol stack configuration parameters.

6. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 3, the protocol abstraction middleware implements a unified semantic expression of the multi-protocol interface using a hierarchical mapping mechanism, including: a semantic abstraction layer converts the dynamic topology descriptor into a protocol-independent intermediate representation; and a protocol adaptation layer maps the intermediate representation into an OpenFlow flow table entry, a Restconf configuration template, or a Netconf XML instruction according to the target controller type.

7. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: The incremental construction algorithm in step 4 includes: a time window segmentation module, which divides the teaching process into basic stage, transition stage and advanced stage windows; a virtual node density parameter adjustment module, which sets a low density value for a single-domain topology or a high density value for a cross-domain topology based on stage characteristics; and a progressive reconstruction module, which updates only newly added or deleted virtual nodes and links within the current time window.

8. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 4, the time window is set to [t1, t2]. Within this time window, the incremental construction algorithm only processes the network topology information related to this time window. According to different teaching stages, the virtual node density parameter ρ(t) is dynamically adjusted as follows: Among them, ρ base It is the benchmark for calculating virtual node density in all teaching stages, reflecting the topological complexity in the early stages of teaching; ΔS is the increment of teaching progress; ΔC is the complexity of concurrent operations; α and β are adaptive adjustment coefficients.

9. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 5, the topology self-healing mechanism includes two modes: link redundancy rerouting and protocol stack dynamic rollback. In the link redundancy rerouting mode, when a link interruption is detected, the system automatically switches to a backup link. In the protocol stack dynamic rollback mode, when the protocol configuration is abnormal, the system restores the protocol parameters to the stable version.

10. The method for dynamically constructing a virtual network topology for SDN according to claim 1, wherein: In step 5, the topological entropy value E is calculated using the following formula: Where N is the number of network states, p i is the probability of the i-th network state occurring; when a topology entropy mutation |E(t)-E(t-Δt)|>∈ is detected, the topology self-healing mechanism is triggered, where ∈ is a preset threshold.

Images