Intelligent conference room network cabling planning method and system and storage medium

Abstract

The invention provides an intelligent conference room network cabling planning method and system and a storage medium, and the method comprises the steps: representing a network QoS index and a security isolation level as a service demand vector through obtaining the topology and service demands of a conference room, and combining a priority vector generated based on the service importance, thereby achieving the network cabling planning of the conference room. Calculating a path penalty coefficient of each wiring path in the constructed network topological graph, performing VLAN initial division according to the security level, constructing a cost function including comprehensive wiring cost, switch port balance degree and cross-VLAN communication overhead, configuring the weight by a service priority vector, solving the function through iterative optimization, and performing VLAN initial division according to the security level. And solving the wiring scheme and predicting the flow based on the current VLAN division in each iteration, if equipment pairs with overlarge flow among VLANs exist, merging the equipment pairs to update the division until stable division is obtained, and generating the comprehensive wiring scheme comprising cable routing, port allocation and VLAN configuration according to the stable division.

Description

A method, system, and storage medium for planning network cabling in intelligent conference rooms. Technical Field

[0001] This disclosure pertains to the field of network cabling planning, and particularly relates to a method, system, and storage medium for planning network cabling in intelligent conference rooms. Background Technology

[0002] A typical smart conference room integrates video conferencing terminals, high-definition displays, wireless projection devices, intelligent control panels, IoT devices, and numerous user access points, supporting various services such as high-definition video streaming, real-time data collaboration, and device control signaling. These services have varying network performance requirements; for example, video conferencing requires low latency and high bandwidth, while device control demands higher reliability. Current conference room network cabling planning is typically done manually by network engineers based on architectural drawings and experience. This method heavily relies on the engineer's individual skills and subjective judgment, resulting in low efficiency, high time consumption, and difficulty in systematically balancing multiple interdependent factors such as cabling costs, network performance, security, and resource utilization. Consequently, the solutions are often not optimal and fail to meet increasingly complex business needs. In selecting cabling paths, estimations are often based solely on experience, lacking an evaluation mechanism that combines network service quality indicators with path characteristics, easily leading to compromised transmission quality for critical services. Regarding logical network isolation, VLANs are typically divided based on static rules such as department or device type. This method is rather coarse and does not consider the actual communication patterns between devices. When devices requiring frequent interaction are assigned to different VLANs, a large amount of cross-VLAN communication traffic is generated. This not only increases the processing burden on core switching equipment and introduces additional network latency, but also wastes network resources. Furthermore, switch port allocation is often done manually, making it difficult to achieve load balancing across all switch port resources. This often results in some device ports being overutilized while others remain largely idle, impacting overall network stability and future scalability. Therefore, current technology urgently needs an intelligent planning method that can comprehensively consider multi-dimensional needs to optimize network cabling solutions. Summary of the Invention

[0003] This disclosure proposes an intelligent conference room network cabling planning method to address the problem that existing technologies struggle to comprehensively consider multi-dimensional needs in intelligent planning for conference room networks, thereby achieving optimal network cabling solutions. The method includes the following steps:

[0004] Obtain the topology and service requirements of the conference room, including service importance, quality of service (QoS) indicators, and security isolation level; generate a service priority vector based on the service importance, and parse the QoS indicators and security isolation level into a service requirement vector;

[0005] Based on the aforementioned topology, a network topology diagram is constructed, and in conjunction with the QoS indicators in the service requirement vector, a path penalty coefficient is calculated for each potential cabling path between each device node in the network topology diagram.

[0006] Initial VLAN segmentation is performed based on the security isolation level in the service requirement vector; the weighted sum of the integrated cabling cost, switch port utilization balance, and cross-VLAN communication overhead is used as the multi-objective optimization cost function, wherein the integrated cabling cost is the sum of the path penalty coefficients of all selected cabling paths;

[0007] The cost function is iteratively solved to optimize VLAN partitioning until a stable partitioning is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between each VLAN; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold. If so, the device pairs are adjusted to the same VLAN to update the VLAN partitioning and proceed to the next iteration; if not, the current VLAN partitioning is determined to be a stable partitioning; and generating a comprehensive network cabling scheme that includes specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

[0008] Optionally, parsing the network service quality (QoS) indicators and security isolation levels into service requirement vectors includes:

[0009] The QoS metrics of bandwidth, latency, jitter, and security isolation level are numerically quantified to form a multi-dimensional service requirement vector.

[0010] Optionally, the initial VLAN partitioning based on the security isolation level in the service requirement vector includes:

[0011] Devices with the same security isolation level are grouped into the same initial VLAN.

[0012] Optionally, the weights of the weighted summation are configured according to the service priority vector, including:

[0013] Different weight combinations are configured for services with different service priorities. For services with higher service priorities, the weight of the integrated cabling cost in the weighted summation is reduced accordingly, while the weight of the switch port utilization balance and cross-VLAN communication overhead is increased.

[0014] Optionally, the step of obtaining a network planning scheme by solving the cost function based on the current VLAN partitioning and predicting the communication traffic between each VLAN includes:

[0015] Each iteration includes: calculating the optimal routing path between each device node based on the current VLAN division and using the path penalty coefficient; mapping service traffic to the optimal routing path according to the service interaction relationship in the service demand vector, thereby statistically analyzing and predicting the aggregated communication traffic between each VLAN.

[0016] Optionally, determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold includes:

[0017] Traverse all device pairs that communicate across VLANs and calculate their predicted communication traffic; filter out device pairs whose predicted communication traffic is greater than the preset merging threshold and mark them as candidate device pairs to be optimized.

[0018] Optionally, adjusting the devices to the same VLAN to update the VLAN division includes:

[0019] For the candidate device pair to be optimized, the following security and cost verification logic is executed:

[0020] Step A: Determine whether the security isolation levels of the two devices in the pair are consistent. If they are inconsistent, prohibit merging and keep the original VLAN division unchanged. If they are consistent, proceed to Step B.

[0021] Step B: Construct an exploratory partitioning scheme to adjust the two devices to the same VLAN, and calculate the multi-objective optimization cost function value under this scheme;

[0022] Step C: Compare the cost function values ​​of the tentative partitioning scheme and the current scheme. If the cost of the tentative partitioning scheme is lower, perform a merge operation to update the VLAN partitioning; otherwise, abandon the merge.

[0023] Optionally, generating a comprehensive network cabling scheme that includes specific cable routing, device port allocation, and VLAN configuration includes:

[0024] Output a data structure containing the identifier of each cabling, the starting device and port, the destination device and port, the VLAN identifier, the cable specification, and the path description.

[0025] In addition, this disclosure also relates to an intelligent conference room network cabling planning system, including the following modules:

[0026] The acquisition module is used to acquire the topology and business requirements of the conference room, wherein the business requirements include business importance, network service quality (QoS) indicators, and security isolation level; generate a business priority vector based on the business importance, and parse the network service quality (QoS) indicators and security isolation level into a business requirement vector;

[0027] The calculation module is used to construct a network topology map based on the topology, and calculate the path penalty coefficient for each potential cabling path between each device node in the network topology map by combining the QoS indicators in the service requirement vector.

[0028] The module is used to perform initial VLAN partitioning based on the security isolation level in the service requirement vector; and to use the weighted sum of the integrated cabling cost, switch port utilization balance and cross-VLAN communication overhead as a multi-objective optimization cost function, wherein the integrated cabling cost is the sum of the path penalty coefficients of all selected cabling paths;

[0029] The generation module is used to optimize VLAN partitioning by iteratively solving the cost function until a stable partitioning is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between each VLAN; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold. If so, the device pairs are adjusted to the same VLAN to update the VLAN partitioning and proceed to the next iteration; if not, the current VLAN partitioning is determined to be a stable partitioning; and generating a comprehensive network cabling scheme including specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

[0030] Preferably, the step of parsing the network service quality (QoS) indicators and security isolation levels into a service requirement vector includes:

[0031] The QoS metrics of bandwidth, latency, jitter, and security isolation level are numerically quantified to form a multi-dimensional service requirement vector.

[0032] Preferably, the initial VLAN partitioning based on the security isolation level in the service requirement vector includes:

[0033] Devices with the same security isolation level are grouped into the same initial VLAN.

[0034] Preferably, the weights of the weighted summation are configured according to the business priority vector, including:

[0035] Different weight combinations are configured for services with different service priorities. For services with higher service priorities, the weight of the integrated cabling cost in the weighted summation is reduced accordingly, while the weight of the switch port utilization balance and cross-VLAN communication overhead is increased.

[0036] Preferably, the step of obtaining a network planning scheme by solving the cost function based on the current VLAN partitioning and predicting the communication traffic between each VLAN includes:

[0037] Each iteration includes: calculating the optimal routing path between each device node based on the current VLAN division and using the path penalty coefficient; mapping service traffic to the optimal routing path according to the service interaction relationship in the service demand vector, thereby statistically analyzing and predicting the aggregated communication traffic between each VLAN.

[0038] Preferably, determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold includes:

[0039] Traverse all device pairs that communicate across VLANs and calculate their predicted communication traffic; filter out device pairs whose predicted communication traffic is greater than the preset merging threshold and mark them as candidate device pairs to be optimized.

[0040] Preferably, the step of adjusting the devices to the same VLAN to update the VLAN division includes:

[0041] For the candidate device pair to be optimized, the following security and cost verification logic is executed:

[0042] Step A: Determine whether the security isolation levels of the two devices in the pair are consistent. If they are inconsistent, prohibit merging and keep the original VLAN division unchanged. If they are consistent, proceed to Step B.

[0043] Step B: Construct an exploratory partitioning scheme to adjust the two devices to the same VLAN, and calculate the multi-objective optimization cost function value under this scheme;

[0044] Step C: Compare the cost function values ​​of the tentative partitioning scheme and the current scheme. If the cost of the tentative partitioning scheme is lower, perform a merge operation to update the VLAN partitioning; otherwise, abandon the merge.

[0045] Preferably, the generation of a comprehensive network cabling scheme that includes specific cable routing, device port allocation, and VLAN configuration includes:

[0046] Output a data structure containing the identifier of each cabling, the starting device and port, the destination device and port, the VLAN identifier, the cable specification, and the path description.

[0047] Furthermore, this disclosure also proposes a computer-readable storage medium storing a computer program, characterized in that the computer program, when executed by a processor, implements the method described above.

[0048] This disclosure comprehensively considers the conference room topology and diverse business needs, including business importance, network service quality (QoS) indicators, and security isolation levels. It constructs a multi-objective optimization model that integrates structured cabling costs, switch port utilization balance, and cross-VLAN communication overhead, and assigns weights based on business priorities to achieve global optimization of network planning solutions. This method balances security isolation requirements with high-frequency communication efficiency through iterative optimization of VLAN segmentation, reducing unnecessary network routing overhead and communication latency. It generates a precise and complete integrated network cabling solution, which not only improves planning efficiency and standardization but also enhances overall network resource utilization, ensures the service quality of critical businesses, and achieves an optimized balance between cost, performance, and security, making network deployment more scientific. Attached Figure Description

[0049] Figure 1 is a flowchart of the first embodiment;

[0050] Figure 2 is a schematic diagram of business requirement vectorization;

[0051] Figure 3 is a schematic diagram of the initial VLAN division based on security level;

[0052] Figure 4 is a schematic diagram of the weight configuration of the multi-objective cost function. Detailed Implementation

[0053] The technical solutions of the embodiments of this disclosure will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this disclosure, and not all embodiments. Based on the embodiments of this disclosure, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this disclosure.

[0054] In the first embodiment, this disclosure proposes a smart conference room network cabling planning method, as shown in Figure 1, which includes the following steps:

[0055] S1. Obtain the topology and service requirements of the conference room, wherein the service requirements include service importance, network service quality (QoS) indicators, and security isolation level; generate a service priority vector based on the service importance, and parse the network service quality (QoS) indicators and security isolation level into a service requirement vector;

[0056] Specifically, the system imports CAD drawing files of the conference room through an input interface, parses the drawings to obtain the precise coordinates of walls, floors, ceilings, cable trays, and information points, forming topology data. Business requirements are entered through a preset form template. For example, for video conferencing services, the importance of the service is entered as 5, the QoS indicators are entered as latency less than 20ms, bandwidth greater than 100Mbps, and security isolation level is entered as high. All importance values ​​of services, such as 1 to 5, are normalized to values ​​from 0.2 to 1.0 to form a business priority vector. The QoS indicators and security isolation level are digitized, for example, latency less than 20ms is recorded as 20, bandwidth greater than 100Mbps is recorded as 100, and high security isolation level is recorded as 3, forming a business requirement vector.

[0057] In an optional embodiment, resolving the network service quality (QoS) metrics and security isolation level into a service requirement vector includes:

[0058] The QoS metrics of bandwidth, latency, jitter, and security isolation level are numerically quantified to form a multi-dimensional service requirement vector.

[0059] For example, a high-definition video conferencing service has specific performance requirements: bandwidth no less than 20Mbps, latency no more than 50ms, and jitter no more than 10ms. These values ​​will be used as components of a vector. The second step is to represent the security isolation level. Assuming the network environment defines three security domains: a core R&D area, a general office area, and a visitor area, these areas can be represented as Level 3, Level 2, and Level 1, respectively.

[0060] Based on the above results, a business requirement vector can be constructed. For the high-definition video conferencing terminal located in the core R&D area, the business requirement vector can be represented as V=[20,50,10,3]. This vector structure is an ordered array, where the first element represents the minimum bandwidth requirement in Mbps; the second element represents the maximum latency tolerance in ms; the third element represents the maximum jitter tolerance in ms; and the fourth element represents the security isolation level. This vector serves as the basic input for all subsequent calculations and optimizations, as shown in Figure 2.

[0061] S2, Based on the topology, construct a network topology diagram, and combine the QoS indicators in the service requirement vector to calculate the path penalty coefficient for each potential cabling path between each device node in the network topology diagram;

[0062] Specifically, the servers, switches, routers, wireless access points, computers, projectors, and other devices in the conference room are used as nodes in the network topology diagram. Feasible cabling paths, such as along wall corners, through ceiling ducts, or using floor cable trays, are used as edges connecting the nodes. The initial weight of the edges is the distance, for example, 15m. The weights are adjusted according to QoS indicators. For example, if a path passes through a power well, which may cause electromagnetic interference and increase latency, failing to meet the requirement of less than 20ms latency for video conferencing services, then the path penalty coefficient for video conferencing services is set to a maximum value. For ordinary office services without this requirement, only a smaller penalty value is added, for example, a path penalty coefficient of 18 is obtained by multiplying the distance of 15m by the interference factor of 1.2. The path's performance, including maximum bandwidth, transmission delay, and reliability under electromagnetic interference, is compared with QoS indicators of service requirements, such as minimum bandwidth and maximum latency tolerance. The greater the difference, the higher the corresponding penalty factors for bandwidth, delay, and reliability. The base cost and each penalty factor are multiplied by their respective weights according to their importance, and the sum is the path penalty coefficient. The higher the coefficient value, the greater the overall cost of the potential cabling path, and the lower its priority in planning.

[0063] In an optional embodiment, the step of constructing a network topology map based on the topology and calculating a path penalty coefficient for each potential cabling path between device nodes in the network topology map, in conjunction with the QoS indicators in the service requirement vector, includes:

[0064] The calculation formula is: ,in, Let $\mathbf{i}$ be the path penalty coefficient connecting device nodes $i$ and $j$. For path length The relevant length penalty factor is a function that maps length to a dimensionless penalty value. The number of obstacles traversed by the path The relevant obstacle penalty factor is a function that maps the number of obstacles to a dimensionless penalty value. To match the cable type and path length used The relevant material and construction cost factors are converted into dimensionless penalty values; The penalty function generated for the difference between the actual QoS capability of the path and the QoS metrics required by the service; , , , These are the weighting coefficients for each item.

[0065] For example, length penalty factor The obstacle penalty factor can be determined by comparing the actual path length with the maximum permissible length in standard cabling specifications (such as 90 meters for horizontal cabling). It is a piecewise function, where different numbers of obstacles correspond to different obstacle penalty factors. The total material cost is obtained by multiplying the cost per meter of cable by the path length, and then dividing by a baseline cost value. Specifically Where B, L, and J represent bandwidth, latency, and jitter, respectively; the subscript req indicates the business requirement; and path indicates the actual capability of the path. , , The weights of each indicator.

[0066] S3, perform initial VLAN partitioning based on the security isolation level in the service requirement vector; use the weighted sum of the integrated cabling cost, switch port utilization balance and cross-VLAN communication overhead as a multi-objective optimization cost function, wherein the integrated cabling cost is the sum of the path penalty coefficients of all selected cabling paths;

[0067] Specifically, all devices with high security isolation levels, such as core servers and management terminals, are assigned to VLAN 10; devices with medium security isolation levels, such as employee computers and printers, are assigned to VLAN 20; and devices with low security isolation levels, such as visitor wireless network access points, are assigned to VLAN 30. The construction cost function F equals... Multiply by the cost of C-line cabling, plus Multiply by the C-port equalizer, plus Multiply by C for cross-VLAN communication; Overall cabling cost C is the sum of path penalty coefficients for all devices connecting to the switch via the selected path; Switch port utilization balance C is the variance of the number of used ports on all switches; the smaller the variance, the higher the balance; Cross-VLAN communication overhead C is the estimated total traffic requiring Layer 3 routing; If the service priority vector shows that service continuity is the most important, then configure the weights. It is 0.6. It is 0.2. It is 0.2.

[0068] In an optional embodiment, the initial VLAN partitioning based on the security isolation level in the service requirement vector includes:

[0069] Devices with the same security isolation level are grouped into the same initial VLAN.

[0070] Extract the service requirement vectors of all device nodes in the network and read the security isolation level dimension. For example, there are five devices in the network: devices A, B, C, D, and E. Their security levels in the service requirement vectors are 2, 3, 2, 1, and 3, respectively. Group the devices according to the security isolation level values. All devices with level 1 are grouped together, all devices with level 2 are grouped together, and so on. In the example above, device D is assigned to level 1 group; devices A and C are assigned to level 2 group; and devices B and E are assigned to level 3 group. Assign a unique VLAN ID to each group to complete the initial VLAN partitioning. For example, assign VLAN 10 to level 1 group, VLAN 20 to level 2 group, and VLAN 30 to level 3 group. The initial VLAN partitioning is: device D is in VLAN 10; devices A and C are in VLAN 20; and devices B and E are in VLAN 30, as shown in Figure 3.

[0071] In an optional embodiment, the weights of the weighted summation are configured according to the service priority vector, including:

[0072] Different weight combinations are configured for services with different service priorities. For services with higher service priorities, the weight of the integrated cabling cost in the weighted summation is reduced accordingly, while the weight of the switch port utilization balance and cross-VLAN communication overhead is increased.

[0073] Define service priority levels and their corresponding weight models. For example, define three priorities: high, medium, and low, corresponding to values ​​of 3, 2, and 1 respectively. Simultaneously, assign different weight sets to the three sub-objectives in the overall optimization objective function: structured cabling cost, port utilization balance, and cross-VLAN communication overhead. For example, the weight combination set for high-priority services. The weighting combination for low-priority services is set with a cost of 0.1, a balance of 0.5, and a communication overhead of 0.4. Cost: 0.7, Balance: 0.1, Communication overhead: 0.2, as shown in Figure 4.

[0074] When planning a network, select appropriate weight combinations based on the type of business being processed. For example, when planning network connections for the core database server, this business is marked as high priority. Weighted combinations are used to evaluate all alternatives. This indicates a preference for options that, even with slightly higher cabling costs, improve port load balancing and reduce cross-VLAN communication. Conversely, when planning a network for a typical office printer, a weighted combination is used... When weighting combinations, finding the lowest-cost cabling solution becomes the primary objective.

[0075] S4. The cost function is iteratively solved to optimize VLAN partitioning until a stable partitioning is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between each VLAN; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold. If so, the device pairs are adjusted to the same VLAN to update the VLAN partitioning and proceed to the next iteration; if not, the current VLAN partitioning is determined to be a stable partitioning; and generating a comprehensive network cabling scheme that includes specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

[0076] Specifically, in the first iteration, based on the initial VLAN 10, 20, and 30 partitions, Dijkstra's algorithm is used to calculate the cabling scheme that minimizes the cost function F. According to the service model, a long-term video data flow of up to 500Mbps is predicted between the video conferencing terminal in VLAN 10 and the file server in VLAN 20. The VLAN merging threshold is set to 100Mbps. Since 500 is greater than 100, it is determined that the communication overhead between the video conferencing terminal and the file server is too high, and it is decided to move the file server from VLAN 20 to VLAN 10. A new VLAN partitioning scheme is formed, and this new scheme enters the second iteration, where the cost function is recalculated and the inter-VLAN traffic is evaluated. If the communication traffic between all device pairs in all VLANs is less than 100Mbps in the second iteration, the iteration stops, and the current VLAN partitioning is considered stable. To avoid the arbitrariness of the greedy strategy and ensure the effectiveness of the optimization path, the order in which multiple candidate device pairs to be optimized are processed should be based on the principle of maximizing global benefits. In each iteration, all device pairs that meet the merging conditions (i.e., traffic exceeding the threshold and allowed by the security policy) are first traversed and identified. Merging is performed on each candidate pair, and the decrease in the global cost function after merging is calculated. Select candidate pairs that yield the greatest cost reduction for actual merging. Ensure that each iteration is the optimal choice at that moment, thus increasing the likelihood of approaching the global optimum.

[0077] Output a network cabling diagram showing the specific route of each network cable, for example, starting from port 3 under the conference table, along the floor cable tray to the corner, then up the wall cable tray to the ceiling, and finally connecting to the core switch in the low-voltage room; simultaneously generate a device configuration list, for example, the core switch, port GE1 / 0 / 1, connected to device PC01, assigned to VLAN 20; port GE1 / 0 / 2, connected to device projector 01, assigned to VLAN 10; and automatically generate the switch configuration script, which includes creating VLAN 10 named SERVER, creating VLAN 20 named STAFF, and command-line instructions to assign each port to the corresponding VLAN.

[0078] In an optional embodiment, the step of solving the cost function based on the current VLAN partitioning to obtain the network planning scheme and predicting the communication traffic between VLANs includes:

[0079] Each iteration includes: calculating the optimal routing path between each device node based on the current VLAN division and using the path penalty coefficient; mapping service traffic to the optimal routing path according to the service interaction relationship in the service demand vector, thereby statistically analyzing and predicting the aggregated communication traffic between each VLAN.

[0080] Assume that at the start of the Nth iteration, VLAN 10 contains devices A and B, and VLAN 20 contains device C. Using a minimum cost path algorithm such as Dijkstra's algorithm or A*, and employing the previously calculated path penalty coefficients as edge weights, the optimal cabling paths connecting devices A, B, and C to the core switching network are calculated. The set of these paths constitutes the provisional network planning scheme for the current iteration.

[0081] A pre-defined traffic simulation model is used, typically an N×N traffic matrix, where N is the total number of devices, and matrix element T(i,j) represents the expected communication data volume from device i to device j. For example, the model predicts traffic of 1.5Gbps from device A to device C and 0.8Gbps from device B to device C. Based on the current VLAN segmentation, all cross-VLAN traffic is aggregated. In this example, the total predicted traffic from VLAN 10 to VLAN 20 is the sum of traffic from A to C and from B to C, which is 2.3Gbps.

[0082] Optionally, the traffic simulation model is a graph neural network model. It consists of a three-layer graph convolutional network (GCN) and a decoder for prediction, where the decoder is a three-layer multilayer perceptron. The training set is derived from one year of historical traffic logs of the existing enterprise network, containing the actual communication data volume between all devices at each timestamp, as well as the device roles and service types. The model input is a graph data structure representing the current network topology, where each node represents a device, and the node's feature vector contains device type, service priority, security level, and the current hour and day of the week information. The model output is an N×N traffic matrix. Where N is the total number of devices in the network, and the elements in the matrix are... This represents the predicted average communication rate from device i to device j.

[0083] In some embodiments, the step of solving the cost function based on the current VLAN partitioning to obtain the network planning scheme and predicting the communication traffic between each VLAN includes:

[0084] Each iteration includes: calculating the optimal routing path between each device node based on the current VLAN division and using the path penalty coefficient; mapping service traffic to the optimal routing path according to the service interaction relationship in the service demand vector, thereby statistically analyzing and predicting the aggregated communication traffic between each VLAN.

[0085] Based on the business demand vector parsing, the business interaction matrix between devices is obtained. The existence of communication needs, demand types, and baseline traffic of device pairs are recorded. The Dijkstra algorithm is used to select the optimal routing path for each device pair, and the path with the smallest path penalty coefficient is the optimal one. Traffic prediction models are configured for different business types. For example, peak traffic for real-time video conferencing is predicted by bitrate × meeting duration × concurrency coefficient, and ordinary data transmission is predicted by historical average traffic × fluctuation coefficient. After mapping each business traffic to the corresponding optimal routing path, the aggregated traffic across VLANs is statistically analyzed according to VLAN affiliation, and the individual traffic value of each pair of cross-VLAN communication devices is recorded.

[0086] In an optional embodiment, determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold includes:

[0087] Traverse all device pairs that communicate across VLANs and calculate their predicted communication traffic; filter out device pairs whose predicted communication traffic is greater than the preset merging threshold and mark them as candidate device pairs to be optimized.

[0088] For example, based on the core router's processing capacity, this threshold is set to 1Gbps. This illustrates that any pair of devices in different VLANs, if their communication traffic exceeds 1Gbps, may bottleneck network performance, requiring them to be placed in the same VLAN. Assume simulation results show: device A in VLAN 10 and device C in VLAN 20 have a traffic of 1.2Gbps; device B in VLAN 10 and device D in VLAN 30 have a traffic of 500Mbps. Compare these cross-VLAN traffic values ​​one by one. During the comparison, it is found that the traffic between A and C (1.2Gbps) is greater than the preset 1Gbps threshold, therefore the device pair (A,C) is marked as needing VLAN merging adjustment. The traffic between B and D (500Mbps) is less than the 1Gbps threshold, so no operation is performed. The output is a list containing all device pairs that need VLAN merging, which will guide subsequent VLAN re-assignment operations.

[0089] In an optional embodiment, all device pairs that communicate across VLANs are traversed, and their predicted communication traffic is calculated; device pairs whose predicted communication traffic is greater than the preset merging threshold are selected and marked as candidate device pairs to be optimized.

[0090] Specifically, the traversal scope is limited to all device pairs in different VLANs that have predicted communication traffic. Device pairs without communication needs are excluded to reduce computational overhead. A single-device-pair traffic accumulation algorithm is used to calculate the predicted value, that is, to summarize the predicted peak traffic of the device pair under various service scenarios within the planning period, and take the maximum value to avoid misjudgment due to instantaneous traffic fluctuations. The preset merging threshold can be dynamically configured according to service priority. For example, the threshold for device pairs corresponding to high-priority services is set to 800Mbps, ordinary priority is set to 1Gbps, and low priority is set to 1.2Gbps to ensure that the network performance of core services is given priority. During the filtering process, the VLAN to which the device pair belongs, the security isolation level, and the traffic composition are recorded simultaneously to generate a candidate device pair list containing device ID, VLAN to which it belongs, security level, predicted traffic, and threshold.

[0091] In an optional embodiment, adjusting the devices to the same VLAN to update the VLAN division includes:

[0092] For a pair of devices that need to be merged, if the security isolation level of one device's VLAN is higher than that of the other device's VLAN, then the device with the higher security isolation level will be moved to the VLAN with the lower security isolation level.

[0093] If the two have the same security level, then all devices in the VLAN with fewer devices will be merged into the VLAN with more devices.

[0094] The first scenario involves merging VLANs with different security levels. Assume device A is in VLAN 20 with a security level of 3, and device B is in VLAN 10 with a security level of 2, and both A and B are marked as needing to be merged. According to the rules, higher-level devices are merged into lower-level ones; therefore, device A will be moved out of VLAN 20 and joined into VLAN 10. This operation follows the principle of security degradation, moving protected assets to a less protected environment, and this change is recorded for auditing purposes.

[0095] The second scenario involves merging VLANs with the same security level. Assume device C is in VLAN 30 and device D is in VLAN 40, both VLANs having a security level of 2. VLAN 30 has 5 devices, while VLAN 40 has 12 devices. When C and D need to be merged, the VLAN with fewer devices (VLAN 30) is merged entirely into VLAN 40. Therefore, all 5 devices originally belonging to VLAN 30, including C, will be reconfigured to VLAN 40. VLAN 30 is thus cleared. This updated VLAN partitioning scheme will serve as input for the next optimization iteration.

[0096] In an optional embodiment, adjusting the devices to the same VLAN to update the VLAN division includes:

[0097] For the candidate device pair to be optimized, the following security and cost verification logic is executed:

[0098] Step A: Determine whether the security isolation levels of the two devices in the pair are consistent. If they are inconsistent, prohibit merging and keep the original VLAN division unchanged. If they are consistent, proceed to Step B.

[0099] Step B: Construct an exploratory partitioning scheme to adjust the two devices to the same VLAN, and calculate the multi-objective optimization cost function value under this scheme;

[0100] Step C: Compare the cost function values ​​of the tentative partitioning scheme and the current scheme. If the cost of the tentative partitioning scheme is lower, perform a merge operation to update the VLAN partitioning; otherwise, abandon the merge.

[0101] A unique security isolation level is pre-configured for each VLAN and stored in the VLAN configuration table. The level ranges from 1 to 5, with higher values ​​indicating stricter security requirements. The comparison is performed by reading the level field of the VLAN to which the device belongs. If, in device pair (X, Y), VLAN 5 (to which X belongs) has a security level of 4, and VLAN 8 (to which Y belongs) has a security level of 3, the two levels are inconsistent, and the merging prohibition mechanism is directly triggered. At the same time, a log record is generated stating that the merging of device pair (X, Y) is rejected due to the security level mismatch, ensuring that the security isolation principle is not violated and preventing high-security-level devices from being exposed to low-security environments.

[0102] When security levels are consistent, two exploratory schemes are constructed for device pair (X, Y): Scheme 1: migrate X to the VLAN of Y; Scheme 2: migrate Y to the VLAN of X. The multi-objective optimization cost function values ​​for both schemes are calculated. The cost function defined earlier is used in the calculation. Assuming that device pair (X, Y) both belong to VLANs with security level 3, the cost function value of Scheme 1 is 86, and the cost function value of Scheme 2 is 92. Scheme 1, with the lower cost, is selected as the candidate exploratory scheme. A cost optimization threshold is set, such as a cost reduction of ≥5%, to avoid frequent iterations due to small differences. The exploratory scheme (Scheme 1, cost 86) is compared with the current scheme (cost 95). The calculated cost reduction is approximately 9.4%, meeting the optimization threshold. A merge operation is performed, migrating X to the VLAN of Y, and the VLAN configuration table, routing path, and port allocation scheme are updated synchronously. If the cost difference between the exploratory scheme (e.g., 93) and the current scheme (95) is less than 5%, the merge benefit is deemed insufficient, the operation is abandoned, and the original VLAN division is maintained. In addition, after merging, it is necessary to verify whether the QoS requirements of all devices in the new VLAN are met, such as whether the bandwidth and latency meet the standards, to ensure that the merging operation does not affect the operation of existing services. Only after the verification is passed can the new solution be used as input for the next iteration.

[0103] Optionally, cohesion is obtained by calculating the proportion of communication traffic between devices within each VLAN to the total communication traffic of that VLAN. If the cohesion of a VLAN is lower than a preset splitting threshold, it indicates that its internal members are loosely associated, and it can be considered as a splitting candidate. For candidate VLANs, graph partitioning or community discovery algorithms can be used to split them into two or more new VLANs based on the actual communication relationships between their internal devices. The splitting operation is only executed if it reduces the global multi-objective optimization cost function value, thereby achieving bidirectional dynamic optimization of VLAN merging and splitting.

[0104] In an optional embodiment, generating a comprehensive network cabling scheme that includes specific cable routing, device port allocation, and VLAN configuration includes:

[0105] Output a data structure containing the identifier of each cabling, the starting device and port, the destination device and port, the VLAN identifier, the cable specification, and the path description.

[0106] The output is a structured dataset, such as a JSON file or database table, providing network engineers with a guide for installation and configuration. The dataset's structure can be viewed as an array of objects, each representing complete information about a cable. For example, one object in the array might contain the following data: Cable identifier L001. Starting device and port: Port 23 of core switch A. Destination device and port: eth1 port of database server B. VLAN identifier: 50, indicating that this link belongs to VLAN 50. Cable specification: Category 6 unshielded twisted pair. Path description: Starting from the top cable tray of rack A, proceeding 15m along the east side cable tray, passing pillar 3, and descending to rack B, entered from the back of the rack. The entire output file will contain similar entries for all cables that need to be deployed in the network. For example, the next record, L002, might represent a connection from port 24 of core switch A to an office area AP, belonging to VLAN 60 and using fiber optic patch cords. This detailed plan ensures accurate cabling and logical configuration and is the deliverable of the entire planning process.

[0107] In a second embodiment, this disclosure also provides an intelligent conference room network cabling planning system, including the following modules:

[0108] The acquisition module is used to acquire the topology and business requirements of the conference room, wherein the business requirements include business importance, network service quality (QoS) indicators, and security isolation level; generate a business priority vector based on the business importance, and parse the network service quality (QoS) indicators and security isolation level into a business requirement vector;

[0109] The calculation module is used to construct a network topology map based on the topology, and calculate the path penalty coefficient for each potential cabling path between each device node in the network topology map by combining the QoS indicators in the service requirement vector.

[0110] The module is used to perform initial VLAN partitioning based on the security isolation level in the service requirement vector; and to use the weighted sum of the integrated cabling cost, switch port utilization balance and cross-VLAN communication overhead as a multi-objective optimization cost function, wherein the integrated cabling cost is the sum of the path penalty coefficients of all selected cabling paths;

[0111] The generation module is used to optimize VLAN partitioning by iteratively solving the cost function until a stable partitioning is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between each VLAN; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold. If so, the device pairs are adjusted to the same VLAN to update the VLAN partitioning and proceed to the next iteration; if not, the current VLAN partitioning is determined to be a stable partitioning; and generating a comprehensive network cabling scheme including specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

[0112] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0113] The above description of the disclosed embodiments enables those skilled in the art to make or use this disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for planning network cabling in an intelligent conference room, characterized in that, Includes the following steps: Obtain the topology and service requirements of the conference room, including service importance, network service quality (QoS) indicators, and security isolation level; A service priority vector is generated based on the service importance, and the network service quality (QoS) index and security isolation level are parsed into a service requirement vector. A network topology graph is constructed based on the topology, and a path penalty coefficient is calculated for each potential cabling path between each device node in the network topology graph in combination with the QoS index in the service requirement vector. Initial VLAN segmentation is performed based on the security isolation level in the aforementioned business requirement vector; The multi-objective optimization cost function is a weighted sum of structured cabling cost, switch port utilization balance, and cross-VLAN communication overhead. The structured cabling cost is the sum of path penalty coefficients for all selected cabling paths. The cost function is iteratively solved to optimize VLAN partitioning until a stable partition is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between VLANs; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold; if so, adjusting the device pairs to the same VLAN to update the VLAN partitioning and proceeding to the next iteration; if not, determining that the current VLAN partitioning is stable; and generating a comprehensive network cabling scheme including specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

2. The intelligent conference room network cabling planning method as described in claim 1, characterized in that, The step of parsing the network service quality (QoS) indicators and security isolation levels into a service requirement vector includes: quantifying the QoS indicators of bandwidth, latency, and jitter, as well as the security isolation level, to form a multi-dimensional service requirement vector.

3. The intelligent conference room network cabling planning method as described in any one of claims 1-2, characterized in that, The initial VLAN partitioning based on the security isolation level in the service requirement vector includes: partitioning devices with the same security isolation level into the same initial VLAN.

4. The intelligent conference room network cabling planning method as described in claim 1, characterized in that, The weights of the weighted summation are configured according to the service priority vector, including: configuring different weight combinations for services with different service priorities; for services with higher service priorities, reducing the weight of the integrated cabling cost in the weighted summation accordingly; and increasing the weight of the switch port utilization balance and cross-VLAN communication overhead.

5. The intelligent conference room network cabling planning method as described in claim 1, characterized in that, The process of obtaining a network planning scheme by solving the cost function based on the current VLAN division and predicting the communication traffic between each VLAN includes: each iteration includes: calculating the optimal routing path between each device node based on the current VLAN division and using the path penalty coefficient; mapping the service traffic to the optimal routing path according to the service interaction relationship in the service demand vector, thereby statistically analyzing and predicting the aggregated communication traffic between each VLAN.

6. The intelligent conference room network cabling planning method as described in any one of claims 1-3 and 5, characterized in that, The step of determining whether there are device pairs whose inter-VLAN communication traffic exceeds the preset merging threshold includes: traversing all device pairs that are communicating across VLANs and calculating their predicted communication traffic; filtering out device pairs whose predicted communication traffic is greater than the preset merging threshold and marking them as candidate device pairs to be optimized.

7. The intelligent conference room network cabling planning method as described in claim 1 or 2, characterized in that, The step of adjusting the device pair to the same VLAN to update the VLAN division includes: for the candidate device pair to be optimized, performing the following security and cost verification logic: Step A, determining whether the security isolation levels of the VLANs containing the two devices in the device pair are consistent; if they are inconsistent, merging is prohibited, and the original VLAN division remains unchanged; if they are consistent, proceed to Step B; Step B, constructing a tentative partitioning scheme to adjust the two devices to the same VLAN, and calculating the multi-objective optimization cost function value under the scheme; Step C, comparing the cost function value of the tentative partitioning scheme with the current scheme; if the cost of the tentative partitioning scheme is lower, performing a merging operation to update the VLAN division; otherwise, abandoning the merging.

8. The intelligent conference room network cabling planning method as described in claim 1, characterized in that, The generation of a comprehensive network cabling scheme that includes specific cable routes, device port allocations, and VLAN configurations includes: outputting a data structure that includes the identifier of each cabling, the starting device and port, the target device and port, the VLAN identifier, the cable specifications, and the path description.

9. A smart conference room network cabling planning system, characterized in that, It includes the following modules: an acquisition module, used to acquire the topology and service requirements of the conference room, wherein the service requirements include service importance, network service quality (QoS) indicators and security isolation level; A service priority vector is generated based on the service importance, and the network service quality (QoS) indicators and security isolation levels are parsed into a service requirement vector. The calculation module is used to construct a network topology map based on the topology, and calculate the path penalty coefficient for each potential cabling path between each device node in the network topology map by combining the QoS indicators in the service requirement vector. The module is used to perform initial VLAN partitioning based on the security isolation level in the business requirement vector; The multi-objective optimization cost function is a weighted sum of the structured cabling cost, switch port utilization balance, and cross-VLAN communication overhead. The structured cabling cost is the sum of path penalty coefficients for all selected cabling paths. A generation module iteratively solves the cost function to optimize VLAN partitioning until a stable partition is obtained. Each iteration includes: solving the cost function based on the current VLAN partitioning to obtain a network planning scheme and predicting the communication traffic between VLANs; determining whether there are device pairs whose inter-VLAN communication traffic exceeds a preset merging threshold; if so, adjusting the device pairs to the same VLAN to update the VLAN partitioning and proceeding to the next iteration; if not, determining that the current VLAN partitioning is stable; and generating a comprehensive network cabling scheme including specific cable routes, device port allocations, and VLAN configurations based on the network planning scheme under the stable partitioning.

10. A computer-readable storage medium storing a computer program thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-8.

Images