A QoS mapping method and device for power distribution network 5G-TSN fusion communication, electronic equipment and storage medium

Abstract

The application discloses a QoS mapping method, device, electronic equipment and storage medium for power distribution network 5G-TSN fusion communication, and particularly relates to the technical field of communication. The method comprises the following steps: acquiring n service data streams to be mapped; taking a pre-set power distribution network service category semantic vector as a semantic prior vector, and selecting, in the n service data streams, service data streams closest to the semantic prior vector as k cluster centers; for each service data stream, calculating the weighted distance between the service data stream and the k cluster centers according to a pre-set weighted distance formula, and classifying the service data stream into a cluster cluster in which the weighted distance is the smallest; and mapping the k cluster clusters to corresponding 5QIs respectively, thereby avoiding the problem that the existing K-means clustering method is easily pulled to one side by GOOSE burst short packets due to random initialization and mean center points.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a QoS mapping method, apparatus, electronic device and storage medium for 5G-TSN converged communication in power distribution networks. Background Technology

[0002] In 5G-TSN converged communication scenarios in power distribution networks, critical traffic and best-effort traffic must be transmitted within the same 5G-TSN converged channel. The core issue for achieving accurate mixed traffic scheduling is how to correctly classify heterogeneous service flows with different security levels and flexibly allocate 5G-side QoS resources. As a TSN logical bridge, the 5G system needs to map TSN-side service data flows to 5QIs to achieve end-to-end QoS guarantees. To achieve accurate mapping, the service data flows are typically clustered first, and then the corresponding 5QI is determined for each service data flow based on the clustering results.

[0003] Existing clustering methods typically use the K-means algorithm to cluster and map business data streams. While this approach has some applicability in general network scenarios, it suffers from the following drawbacks when applied to power distribution network scenarios: On the one hand, the initial cluster centers of the K-means algorithm are randomly selected. After convergence, the cluster categories only reflect the similarity of data in numerical distribution and cannot effectively correspond to various services in the distribution network. This makes the clustering results difficult to use as a valid basis for 5QI mapping decisions, especially in determining which service data flows in a given cluster belong to the highest priority I zone GOOSE protection service. On the other hand, GOOSE messages in the I zone of the distribution network exhibit sudden characteristics at the moment of a fault, with a large number of abnormal short packets flooding in within a very short time. The K-means algorithm uses the mean of samples within a cluster as the cluster center, making it extremely sensitive to outliers. The center point is easily skewed by sudden short packets, leading to cluster boundary drift. GOOSE protection messages that should be classified into high-priority categories may be classified into low-priority categories, causing them to be forwarded along incorrect 5QI paths, directly affecting the timeliness and correctness of power grid protection actions. Summary of the Invention

[0004] This invention proposes a QoS mapping method, device, electronic device, and storage medium for 5G-TSN converged communication in distribution networks. It addresses the technical problems of existing K-means clustering methods in 5G-TSN converged communication scenarios in distribution networks. These problems arise because random initialization and the mean center point are easily skewed by GOOSE burst short packets, resulting in clustering results that cannot effectively correspond to distribution network service types. Furthermore, it addresses the issue that GOOSE protection messages in Zone I cannot obtain the high priority they deserve during the classification stage. This invention provides technical support for the integrated management and control of distribution network data from transmission to reception.

[0005] One embodiment of the present invention provides a QoS mapping method for 5G-TSN converged communication in power distribution networks, comprising: Obtain n business data streams to be mapped, each of which includes security partition level, packet latency budget, scheduling priority, and packet error rate; Using k pre-defined semantic vectors of power distribution network business categories as semantic prior vectors, the business data streams with the smallest distance from the semantic prior vectors are selected from the n business data streams as k cluster centers. For each of the aforementioned business data flows, the weighted distance between the business data flow and k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data flows are then assigned to the cluster containing the cluster center that minimizes the weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data flow and the k cluster centers. Perform iterative optimization on the k cluster centers until the convergence condition is met, and determine the final k cluster centers and the corresponding k clusters; The k clusters are mapped to their corresponding 5QI values.

[0006] Furthermore, mapping the k clusters to their corresponding 5QI values ​​includes: For each cluster, the security partition level and data packet delay budget corresponding to the cluster center are used as the security partition level and data packet delay budget of the cluster. Based on the security partition level and packet latency budget of each cluster, candidate 5QI values ​​for each cluster are determined from the preset 5QI set; The triangular membership function is used to calculate the membership degree between the cluster and each 5QI candidate value in three dimensions: packet delay budget, scheduling priority and packet error rate. The membership degree of the three dimensions is then weighted and merged into a comprehensive membership degree. 5QI candidate values ​​with a comprehensive membership degree greater than a preset first threshold are assigned to the preferred interval; 5QI candidate values ​​with a comprehensive membership degree less than the first threshold but greater than a second threshold are assigned to the flexible candidate interval; and 5QI candidate values ​​with a comprehensive membership degree less than the second threshold are excluded; wherein, the first threshold is greater than the second threshold. Obtain the channel resource utilization rate for each 5QI candidate value, and calculate the load score based on the channel resource utilization rate; The comprehensive membership degree and load score are weighted and summed to obtain the comprehensive score. Within the preferred range, each 5QI candidate value is sorted from high to low according to the comprehensive score. Based on the sorting result, the load score of each 5QI candidate value is compared with the preset overload threshold in order from high to low. If there is a 5QI candidate value with a load score less than the overload threshold within the preferred interval, then the first 5QI candidate value with a load score less than the overload threshold is selected as the mapping result. If the load score of each 5QI candidate value in the preferred interval is greater than or equal to the overload threshold, then switch to the elastic candidate interval and select the 5QI candidate value with the highest comprehensive score from the elastic candidate interval as the mapping result.

[0007] Furthermore, the semantic vector of the distribution network service category is a semantic feature vector constructed based on the typical distribution network services defined by the IEC 61850 standard; The semantic vectors of the distribution network service categories include semantic feature vectors corresponding to GOOSE tripping, GOOSE non-tripping, SV sampled value, remote control, and telemetry. Each semantic feature vector corresponds to the security partition level of the typical distribution network service.

[0008] Furthermore, the iterative optimization of the k cluster centers until the convergence condition is met, to determine the final k cluster centers and the corresponding k clusters, includes: For each cluster, each non-cluster center business data flow within the cluster is assumed to be a candidate replacement cluster center, and the overall clustering cost difference before and after replacing the current cluster center with each candidate replacement cluster center is calculated; wherein, the overall clustering cost difference is the sum of the weighted distances of all business data flows within the cluster to their respective cluster centers; If the calculated overall clustering cost difference is negative, the assumptions about each non-cluster center are terminated, and the current cluster center is replaced by a candidate alternative cluster center as the cluster center of the cluster. The business data streams of each non-cluster center in the cluster are then assumed as candidate alternative cluster centers and iterated again until the overall cost difference in the cluster is not negative. The final k cluster centers and their corresponding k clusters are determined when the overall clustering cost difference of all clusters is no longer negative.

[0009] Furthermore, the weighted distance is calculated using the following formula: ; in, This is the nth business data stream; Let k be the cluster center; For business data flow and cluster center Weighted distance between them; For the nth business data stream The security partition level ranges from 1 to 4, corresponding to zones one through four respectively. For the k-th cluster center Security partition level; For the nth business data stream Data packet latency budget; Cluster center Data packet latency budget; This is a set of QoS parameters, including scheduling priority and packet error rate; for The index; For the nth business data stream In the The values ​​of each QoS parameter; For the k-th cluster center In the The values ​​of each QoS parameter; These are the weighting coefficients for the security partition level dimension; These are the weighting coefficients for the data packet latency budget dimension; For the first The weight coefficients of each QoS parameter dimension, and .

[0010] Furthermore, when the security partition level corresponding to the business data flow is Zone 1, the priority of the business data flow is higher than the priority of business data flows in other security partition levels; wherein, the priority of the business data flow is based on the weighting coefficient of the security partition level of the business data flow. The difference between the weighting coefficients of the data packet delay budget is determined.

[0011] Furthermore, the formula for calculating the comprehensive membership degree is as follows: ; in, This represents the combined membership degree between the k-th cluster and the j-th 5QI candidate value; This represents the k-th cluster; This is the j-th 5QI candidate value; Membership degree for the data packet latency budget dimension; Membership degree is the dimension of scheduling priority; Membership degree for the packet error rate dimension; , , For the corresponding weight coefficients, and , .

[0012] One embodiment of the present invention provides a QoS mapping device for 5G-TSN converged communication in power distribution networks, comprising at least: The data acquisition module is used to acquire n business data streams to be mapped. Each business data stream includes the security partition level, data packet latency budget, scheduling priority, and data packet error rate. The clustering module is used to select the business data streams with the smallest distance to the semantic prior vectors from among the n business data streams, using pre-defined k semantic prior vectors as semantic prior vectors. For each business data stream, the weighted distance between the business data stream and the k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data streams are then assigned to the cluster containing the cluster center with the smallest weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data stream and the k cluster centers. The iterative optimization module is used to perform iterative optimization on the k cluster centers until the convergence condition is met, and to determine the final k cluster centers and the corresponding k clusters. The mapping module is used to map the k clusters to their corresponding 5QI values.

[0013] One embodiment of the present invention provides an electronic device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor; wherein, when the processor executes the computer program, it implements the QoS mapping method for 5G-TSN converged communication in power distribution networks as described in any one of the present invention. One embodiment of the present invention provides a storage medium, the storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to execute the QoS mapping method for 5G-TSN converged communication in power distribution networks as described in any one of the present invention.

[0014] The embodiments of the present invention have the following beneficial effects: This invention provides a QoS mapping method, apparatus, electronic device, and storage medium for 5G-TSN converged communication in power distribution networks. By using preset semantic vectors of power distribution network service categories as semantic priors, the invention selects the nearest real sample in the service data stream that has the smallest distance to each semantic prior vector as the initial cluster center. This anchors the clustering starting point to a location with clear power distribution network service meaning, ensuring that the clustering results converge to various power distribution network services from the initial stage. This solves the problem of ineffective correspondence between clustering categories and service types caused by traditional random initialization, enabling the clustering results to serve as an effective decision-making basis for 5QI mapping. The weighted distance formula is the sum of the weighted squared differences of the normalized security partition level, packet delay budget, scheduling priority, and packet error rate. The weight coefficient of the security partition level is greater than that of other items, making the security partition level dominant in the clustering process. This ensures that the GOOSE protection services in Zone I always move closer to the high-priority center point in the distance calculation. Even if a large number of GOOSE short packets flood in, the real sample center point will not be pulled off by outliers, and Zone I services will still be clustered into the high-priority category. This avoids the misclassification of protection services due to cluster boundary drift, and improves the timeliness and accuracy of power grid protection actions. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a QoS mapping method for 5G-TSN converged communication in a power distribution network, provided by an embodiment of the present invention.

[0016] Figure 2 This is a schematic diagram of the structure of a QoS mapping device for 5G-TSN converged communication in a power distribution network, provided by an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] like Figure 1 As shown, one embodiment of the present invention provides a QoS mapping method for 5G-TSN converged communication in power distribution networks, which may include at least: S1: Obtain n business data streams to be mapped. Each business data stream includes security partition level, packet latency budget, scheduling priority, and packet error rate.

[0021] Furthermore, the safety zoning levels are divided into four levels according to the "Four Zones and Two Networks" safety standard for power: Zone I corresponds to a value of 1, which is the control zone; Zone II corresponds to a value of 2, which is the non-control production zone; Zone III corresponds to a value of 3, which is the production management zone; and Zone IV corresponds to a value of 4. The scheduling priority value ranges from 1 to 7, with higher values ​​indicating higher priority.

[0022] S2: Using k pre-defined semantic vectors for distribution network service categories as semantic prior vectors, select the service data streams with the smallest distance from the semantic prior vectors from among the n service data streams as k cluster centers. In this embodiment, k semantic vectors for distribution network service categories are constructed in advance based on the typical distribution network services defined by the IEC 61850 standard as semantic prior vectors. Each semantic vector is composed of normalized four-dimensional features.

[0023] Specifically, for each semantic prior vector, the Euclidean distance between the business flow and the semantic prior vector is calculated one by one among the n business data flows to be mapped. The business data flow with the smallest distance is selected as the initial cluster center for the corresponding category, resulting in a total of k initial cluster centers. This ensures that the initial cluster anchor points correspond one-to-one with the actual business types of the distribution network, avoiding semantic ambiguity caused by random initialization and ensuring that the clustering results converge to various business types of the distribution network from the initial stage.

[0024] For each of the aforementioned business data flows, the weighted distance between the business data flow and k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data flows are then assigned to the cluster containing the cluster center that minimizes the weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data flow and the k cluster centers.

[0025] Furthermore, since the dimensions and value ranges of each parameter are significantly different, each parameter must be normalized according to its own value range before participating in the weighted distance calculation, so as to ensure that the weight coefficients can truly reflect the relative importance of each dimension in the distance metric.

[0026] By using pre-defined semantic vectors of distribution network business categories as semantic prior vectors, domain knowledge of distribution network business is explicitly injected into the semantic-aware clustering process, so that the clustering results are naturally aligned with the business semantics and security level system of the power system.

[0027] S3: Perform iterative optimization on the k cluster centers until the convergence condition is met, and determine the final k cluster centers and the corresponding k clusters.

[0028] S4: Map the k clusters to their corresponding 5QI values.

[0029] In one embodiment of the present invention, the semantic vector of the distribution network service category is a semantic feature vector constructed based on the typical distribution network services defined by the IEC 61850 standard; The distribution network service category semantic vector includes semantic feature vectors corresponding to GOOSE tripping, GOOSE non-tripping, SV sampled value, remote control, and telemetry categories. Each semantic feature vector corresponds to the security partition level of a typical distribution network service. Furthermore, the distribution network service category semantic vector also includes video inspection categories and other categories besides GOOSE tripping, GOOSE non-tripping, SV sampled value, remote control, and video inspection categories, as shown in Table 1 below: Table 1 Typical Service Classifications and QoS Requirements of Distribution Networks In one embodiment of the present invention, the weighted distance is calculated using the following formula: ; in, This is the nth business data stream; Let k be the cluster center; For business data flow and cluster center Weighted distance between them; For the nth business data stream The security partition level ranges from 1 to 4, corresponding to Zone 1 to Zone 4 (Zone 1 to Zone 4). For the k-th cluster center Security partition level; For the nth business data stream Data packet latency budget; Cluster center Data packet latency budget; This is a set of QoS parameters, including scheduling priority and packet error rate; for The index; For the nth business data stream In the The values ​​of each QoS parameter; For the k-th cluster center In the The values ​​of each QoS parameter; These are the weighting coefficients for the security partition level dimension; These are the weighting coefficients for the data packet latency budget dimension; For the first The weight coefficients of each QoS parameter dimension, and .

[0030] Specifically, the weighting coefficients are set differently for each zone according to the "Four Zones and Two Networks" security isolation standard: when the nth business data flow When it belongs to Zone 1 (Zone I), The value is significantly greater than , Greater than Security partition level is used as the primary criterion for attribution, data packet delay budget is used as the secondary criterion, and other parameters are only used for auxiliary correction.

[0031] When the business data flow belongs to zones two to four (II–IV), Still the largest, but with The gap is moderately narrowed, allowing delay budgets to participate in fine-grained classification. The above constraints can be summarized as follows: The specific value should be determined by those skilled in the art based on the business composition and security level distribution of the deployed distribution network. This design prioritizes the GOOSE protection services in Zone I towards high-priority cluster centers during distance calculation, structurally avoiding the problem of mixing protection-type services with low-priority services.

[0032] In one embodiment of the present invention, mapping the k clusters to their corresponding 5QI values ​​includes: S401: For each cluster, the security partition level and data packet delay budget corresponding to the cluster center are used as the security partition level and data packet delay budget of the cluster. Based on the security partition level and data packet latency budget of each cluster, candidate 5QI values ​​for each cluster are determined from the preset 5QI set.

[0033] S402: The triangular membership function is used to calculate the membership degree between the cluster and each 5QI candidate value in three dimensions: packet delay budget, scheduling priority and packet error rate. The membership degree of the three dimensions is then weighted and merged into a comprehensive membership degree.

[0034] For example, the formula for calculating the membership degree between a cluster and each 5QI candidate value in the packet delay budget using the triangular membership function is as follows: ; in, Let be the membership degree between the k-th cluster and the j-th 5QI candidate value in the packet delay budget; This represents the k-th cluster; This is the j-th 5QI candidate value; Let be the center value of the delay requirement for the k-th cluster; For the packet delay budget of the j-th 5QI candidate value, The acceptable maximum time delay deviation range is configured by the system according to the distribution network operation specifications, using a triangular function. The center value of the delay is the peak point, and the bottom width is the acceptable delay deviation range on both sides. The higher the value, the better the match between the 5QI and the latency requirements of the aggregated stream. The value range is [0,1].

[0035] In one embodiment of the present invention, the formula for calculating the comprehensive membership degree is: ; in, This represents the combined membership degree between the k-th cluster and the j-th 5QI candidate value; This represents the k-th cluster; This is the j-th 5QI candidate value; Membership degree for the data packet latency budget dimension; Membership degree is the dimension of scheduling priority; Membership degree for the packet error rate dimension; , , For the corresponding weight coefficients, and , In this embodiment, , =0.25, =0.15.

[0036] Specifically, the membership degrees for scheduling priority and packet error rate are calculated in the same way as those for packet latency budget, using triangular membership functions with scheduling priority and PER (logarithm) as their respective domains. The comprehensive membership degree serves as the core input for candidate interval partitioning and dynamic optimization, with a value range of [0,1]. A larger value indicates a higher degree of multidimensional adaptation between the aggregated flow and the candidate 5QI, thus more comprehensively reflecting the compatibility between the business and 5QI.

[0037] S403: 5QI candidate values ​​with a comprehensive membership degree greater than a preset first threshold are assigned to the preferred interval; 5QI candidate values ​​with a comprehensive membership degree less than the first threshold but greater than a second threshold are assigned to the flexible candidate interval; 5QI candidate values ​​with a comprehensive membership degree less than the second threshold are excluded and do not participate in subsequent decisions; wherein, the first threshold is greater than the second threshold. In this embodiment, the first threshold is set to 0.7 and the second threshold is set to 0.4.

[0038] This three-layer division replaces the more abstract upper and lower approximation concepts of rough concentration, and realizes the delineation of the elastic range in a more intuitive numerical gating method. In engineering implementation, it has low computational complexity and is easy to deploy.

[0039] S404: Obtain the channel resource utilization rate for each 5QI candidate value, and calculate the load score based on the channel resource utilization rate. The channel resource utilization rate is the ratio of the number of occupied PRBs to the total number of PRBs, with a value range of [0,1]. The load score is calculated using the following formula: ; in, This is the j-th 5QI candidate value; The load score for the j-th 5QI candidate value, The load score represents the channel resource utilization rate of the j-th 5QI candidate value. The higher the load score, the more abundant the channel resources corresponding to the 5QI candidate value and the lower the congestion risk.

[0040] S405: The weighted sum of the comprehensive membership degree and load score is used to obtain the comprehensive score, which is calculated using the following formula: ; in, As a balancing factor, 0.70 is used for Zone I and 0.50 for other Zones II-IV, prioritizing the compatibility of services with 5QI while also considering load balancing; It is the combined score of the k-th cluster and the j-th 5QI candidate value.

[0041] S406: Within the preferred range, sort the 5QI candidate values ​​from high to low according to the comprehensive score. Based on the sorting result, compare the load score of each 5QI candidate value with the preset overload threshold in order from high to low. The overload threshold is preferably set to 0.85, that is, a load score lower than 0.15 is judged as overload.

[0042] If there is a 5QI candidate value with a load score less than the overload threshold within the preferred interval, then the first 5QI candidate value with a load score less than the overload threshold is selected as the mapping result. If the load score of each 5QI candidate value in the preferred interval is greater than or equal to the overload threshold, then switch to the elastic candidate interval and select the 5QI candidate value with the highest comprehensive score from the elastic candidate interval as the mapping result.

[0043] In one embodiment of the present invention, the step of performing iterative optimization on the k cluster centers until the convergence condition is met, and determining the final k cluster centers and the corresponding k clusters, includes: S301: In each iteration, for each cluster, each non-cluster center business data flow within the cluster is assumed to be a candidate replacement cluster center, and the overall clustering cost difference before and after replacing the current cluster center with each candidate replacement cluster center is calculated. The overall clustering cost difference is the difference between the sum of the weighted distances of all business data flows within the cluster to their respective cluster centers and the sum of the weighted distances before the replacement. Specifically, for a given cluster, assuming its current cluster center, any non-center business data flow within the cluster is assumed to be a candidate replacement cluster center, then the overall clustering cost difference before and after the replacement is calculated. It can be represented as: ; in, 1 represents the sum of weighted distances from all business data flows within the replaced cluster to the candidate replacement cluster center; It is the sum of the weighted distances from all business data flows within the previous cluster to the current cluster center; if A negative value indicates that the overall clustering cost is reduced and the clustering quality is improved after the replacement.

[0044] It should be noted that the calculation of the overall clustering cost difference is based on the aforementioned weighted distance formula, that is, the distance between each business data stream and the cluster center is the sum of the weighted squared differences of the normalized security partition level, data packet delay budget, scheduling priority and data packet error rate.

[0045] S302: If the calculated overall clustering cost difference is negative, terminate the assumption of each non-clustering center, replace the current clustering center with a candidate alternative clustering center as the clustering center of the cluster, and iterate again by assuming each non-clustering center's business data stream as a candidate alternative clustering center until the overall cost difference within the clustering cluster is not negative. The final k cluster centers and their corresponding k clusters are determined when the overall clustering cost difference of all clusters is no longer negative.

[0046] Optionally, in one embodiment, if there are multiple candidate alternative cluster centers with negative overall clustering cost differences, the candidate alternative cluster center that reduces the overall clustering cost the most (i.e., the one with negative ΔTC and the largest absolute value) is selected to replace the current cluster center to accelerate the convergence process.

[0047] In one embodiment of the present invention, when the security partition level corresponding to the service data flow is Zone I, the priority of the service data flow is greater than the priority of service data flows in other security partition levels; wherein, the priority of the service data flow is based on the weighting coefficient of the security partition level of the service data flow. The difference between the weighting coefficients of the data packet delay budget is determined.

[0048] Specifically, when the security partition level corresponding to the business data stream is Zone I, the weighting coefficient of the security partition level dimension is... Weighting coefficients of the data packet latency budget dimension The difference between them is used as the first difference. When the security partition level corresponding to the business data stream is another zone, the weighting coefficient of the security partition level dimension is... Weighting coefficients of the data packet latency budget dimension The difference between the two is used as the second difference, and the first difference is greater than the second difference. This is used to ensure that the I area business is always assigned to a higher priority cluster at the cluster boundary, avoiding classification offset caused by sudden short packets.

[0049] Please see Figure 2 Based on the same inventive concept as the above embodiments, the present invention also provides a QoS mapping device for 5G-TSN converged communication in power distribution networks, comprising at least: The data acquisition module is used to acquire n business data streams to be mapped. Each business data stream includes the security partition level, data packet latency budget, scheduling priority, and data packet error rate.

[0050] The clustering module is used to select the business data streams with the smallest distance from the semantic prior vector among n business data streams, using a pre-defined semantic prior vector of the distribution network business category as the semantic prior vector, as k cluster centers. For each business data stream, the weighted distance between the business data stream and the k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data stream is then assigned to the cluster containing the cluster center with the smallest weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data stream and the k cluster centers.

[0051] In one embodiment of the present invention, the weighted distance is calculated using the following formula: ; in, This is the nth business data stream; Let k be the cluster center; For business data flow and cluster center Weighted distance between them; For the nth business data stream The security partition level ranges from 1 to 4, corresponding to zones one through four, respectively. For the k-th cluster center Security partition level; For the nth business data stream Data packet latency budget; Cluster center Data packet latency budget; This is a set of QoS parameters, including scheduling priority and packet error rate; for The index; For the nth business data stream In the The values ​​of each QoS parameter; For the k-th cluster center In the The values ​​of each QoS parameter; These are the weighting coefficients for the security partition level dimension; These are the weighting coefficients for the data packet latency budget dimension; For the first The weight coefficients of each QoS parameter dimension, and .

[0052] Specifically, the weighting coefficients are set differently for each zone according to the "Four Zones and Two Networks" security isolation standard: when the nth business data flow When it belongs to Zone 1 (Zone I), The value is significantly greater than , Greater than Security partition level is used as the primary criterion for attribution, data packet delay budget is used as the secondary criterion, and other parameters are only used for auxiliary correction.

[0053] When the business data flow belongs to zones two to four (II–IV zones), Still the largest, but with The gap is moderately narrowed, allowing delay budgets to participate in fine-grained classification. The above constraints can be summarized as follows: The specific value should be determined by those skilled in the art based on the business composition and security level distribution of the deployed distribution network. This design prioritizes the GOOSE protection services in Zone I towards high-priority cluster centers during distance calculation, structurally avoiding the problem of mixing protection-type services with low-priority services.

[0054] The iterative optimization module is used to perform iterative optimization on the k cluster centers until the convergence condition is met, and to determine the final k cluster centers and the corresponding k clusters.

[0055] In one embodiment of the present invention, the step of performing iterative optimization on the k cluster centers until the convergence condition is met, and determining the final k cluster centers and the corresponding k clusters, includes: S301: In each iteration, for each cluster, each non-cluster center business data flow within the cluster is assumed to be a candidate replacement cluster center, and the overall clustering cost difference before and after replacing the current cluster center with each candidate replacement cluster center is calculated. The overall clustering cost difference is the difference between the sum of the weighted distances of all business data flows within the cluster to their respective cluster centers and the sum of the weighted distances before the replacement. Specifically, for a given cluster, assuming its current cluster center, any non-center business data flow within the cluster is assumed to be a candidate replacement cluster center, then the overall clustering cost difference before and after the replacement is calculated. It can be represented as: ; in, 1 represents the sum of weighted distances from all business data flows within the replaced cluster to the candidate replacement cluster center; It is the sum of weighted distances from all business data flows within the cluster before the replacement to the current cluster center; if ΔTC is negative, it means that the overall clustering cost is reduced and the clustering quality is improved after the replacement.

[0056] It should be noted that the calculation of the overall clustering cost difference is based on the aforementioned weighted distance formula, that is, the distance between each business data stream and the cluster center is the sum of the weighted squared differences of the normalized security partition level, data packet delay budget, scheduling priority and data packet error rate.

[0057] S302: If the calculated overall clustering cost difference is negative, terminate the assumption of each non-clustering center, replace the current clustering center with a candidate alternative clustering center as the clustering center of the cluster, and iterate again by assuming each non-clustering center's business data stream as a candidate alternative clustering center until the overall cost difference within the clustering cluster is not negative. The final k cluster centers and their corresponding k clusters are determined when the overall clustering cost difference of all clusters is no longer negative.

[0058] Optionally, in one embodiment, if there are multiple candidate alternative cluster centers with negative overall clustering cost differences, the candidate alternative cluster center that reduces the overall clustering cost the most (i.e., the one with negative ΔTC and the largest absolute value) is selected to replace the current cluster center to accelerate the convergence process.

[0059] In one embodiment of the present invention, when the security partition level corresponding to the service data flow is Zone I, the priority of the service data flow is greater than the priority of service data flows in other security partition levels; wherein, the priority of the service data flow is based on the weighting coefficient of the security partition level of the service data flow. The difference between the weighting coefficients of the data packet delay budget is determined.

[0060] Specifically, when the security partition level corresponding to the business data stream is Zone I, the weighting coefficient of the security partition level dimension is... Weighting coefficients of the data packet latency budget dimension The difference between them is used as the first difference. When the security partition level corresponding to the business data stream is another zone, the weighting coefficient of the security partition level dimension is... Weighting coefficients of the data packet latency budget dimension The difference between the two values ​​is used as the second difference, and the first difference is greater than the second difference. This is used to ensure that services in Zone I are always assigned to higher-priority clusters at cluster boundaries, avoiding classification offsets caused by sudden short packets.

[0061] The mapping module is used to map the k clusters to their corresponding 5QI values.

[0062] In one embodiment of the present invention, mapping the k clusters to their corresponding 5QI values ​​includes: S401: For each cluster, the security partition level and data packet delay budget corresponding to the cluster center are used as the security partition level and data packet delay budget of the cluster. Based on the security partition level and data packet latency budget of each cluster, candidate 5QI values ​​for each cluster are determined from the preset 5QI set.

[0063] S402: The triangular membership function is used to calculate the membership degree between the cluster and each 5QI candidate value in three dimensions: packet delay budget, scheduling priority and packet error rate. The membership degree of the three dimensions is then weighted and merged into a comprehensive membership degree.

[0064] For example, the formula for calculating the membership degree between a cluster and each 5QI candidate value in the packet delay budget using the triangular membership function is as follows: ; in, Let be the membership degree between the k-th cluster and the j-th 5QI candidate value in the packet delay budget; This represents the k-th cluster; This is the j-th 5QI candidate value; Let be the center value of the delay requirement for the k-th cluster; For the packet delay budget of the j-th 5QI candidate value, The acceptable maximum time delay deviation range (configured by the system according to the distribution network operation specifications) is determined by the triangular function based on the distribution network operation specifications. The center value of the delay is the peak point, and the bottom width is the acceptable delay deviation range on both sides. The higher the value, the better the match between the 5QI and the latency requirements of the aggregated stream. The value range is [0,1].

[0065] In one embodiment of the present invention, the formula for calculating the comprehensive membership degree is: ; in, This represents the combined membership degree between the k-th cluster and the j-th 5QI candidate value; This represents the k-th cluster; This is the j-th 5QI candidate value; Membership degree for the data packet latency budget dimension; Membership degree is the dimension of scheduling priority; Membership degree for the packet error rate dimension; , , For the corresponding weight coefficients, and , In this embodiment, , =0.25, =0.15.

[0066] Specifically, the membership degrees for scheduling priority and packet error rate are calculated in the same way as those for packet latency budget, using triangular membership functions with scheduling priority and PER (logarithm) as their respective domains. The comprehensive membership degree serves as the core input for candidate interval partitioning and dynamic optimization, with a value range of [0,1]. A larger value indicates a higher degree of multidimensional adaptation between the aggregated flow and the candidate 5QI, thus more comprehensively reflecting the compatibility between the business and 5QI.

[0067] S403: 5QI candidate values ​​with a comprehensive membership degree greater than a preset first threshold are assigned to the preferred interval; 5QI candidate values ​​with a comprehensive membership degree less than the first threshold but greater than a second threshold are assigned to the flexible candidate interval; 5QI candidate values ​​with a comprehensive membership degree less than the second threshold are excluded and do not participate in subsequent decisions; wherein, the first threshold is greater than the second threshold. In this embodiment, the first threshold is set to 0.7 and the second threshold is set to 0.4.

[0068] This three-layer division replaces the more abstract upper and lower approximation concepts of rough concentration, and realizes the delineation of the elastic range in a more intuitive numerical gating method. In engineering implementation, it has low computational complexity and is easy to deploy.

[0069] S404: Obtain the channel resource utilization rate for each 5QI candidate value, and calculate the load score based on the channel resource utilization rate. The channel resource utilization rate is the ratio of the number of occupied PRBs to the total number of PRBs, with a value range of [0,1]. The load score is calculated using the following formula: ; in, This is the j-th 5QI candidate value; The load score for the j-th 5QI candidate value, The load score represents the channel resource utilization rate of the j-th 5QI candidate value. The higher the load score, the more abundant the channel resources corresponding to that 5QI and the lower the congestion risk.

[0070] S405: The weighted sum of the comprehensive membership degree and load score is used to obtain the comprehensive score, which is calculated using the following formula: ; in, As a balancing factor, the value is 0.70 for Zone I and 0.50 for other Zones II-IV, prioritizing the compatibility of services with 5QI while also considering load balancing; It is the combined score of the k-th cluster and the j-th 5QI candidate value.

[0071] S406: Within the preferred range, sort the 5QI candidate values ​​from high to low according to the comprehensive score. Based on the sorting result, compare the load score of each 5QI candidate value with the preset overload threshold in order from high to low. The overload threshold is preferably set to 0.85, that is, a load score lower than 0.15 is judged as overload.

[0072] If there is a 5QI candidate value with a load score less than the overload threshold within the preferred interval, then the first 5QI candidate value with a load score less than the overload threshold is selected as the mapping result. If the load score of each 5QI candidate value in the preferred interval is greater than or equal to the overload threshold, then switch to the elastic candidate interval and select the 5QI candidate value with the highest comprehensive score from the elastic candidate interval as the mapping result.

[0073] Based on the same inventive concept as the above embodiments, the present invention also provides a QoS mapping system for 5G-TSN converged communication in power distribution networks. The system consists of three parts: a TSN wired domain on the device terminal side, a 5G system logical bridge (5GS), and a TSN wired domain on the network side, designed with reference to the 3GPP standard architecture.

[0074] Firstly, at the device-side TSN, various service data streams (GOOSE, SV, MMS, and video inspection, etc.) are aggregated at the DS-TT. After aggregation, the data enters the 5G wireless network. In the 5G system control plane, the QoS mapping is completed by the SGK-FM module deployed in TSN-AF, and the policy is distributed to the 5G MAC scheduler through PCF / SMF. On the Radio Access Network (RAN) side, the scheduler completes resource allocation based on the assigned 5QI. On the network side TSN, after the user plane data is carried by the UPF, the NW-TT completes traffic separation, restores the traffic format on the TSN side, and forwards it to the SCADA, protection master station and other distribution master station applications in the network side TSN domain. In the entire mapping process, Zone I GOOSE / SV always obtains the highest priority through the weighting mechanism of SGK-M, and is mapped to high priority resources such as 5QI=82 / 83 by FM-5QI.

[0075] In one embodiment of the present invention, SGK-FM adopts a dual triggering mechanism that combines periodic triggering and event-driven mechanisms, taking into account both controllable system overhead under steady state and rapid response under fault scenarios.

[0076] The periodic triggering mechanism is as follows: The system maps every configured period T. _map(Typical values ​​are on the order of seconds) Perform an SGK-M convergence test on all aggregated flows. Specifically, TSN-AF obtains the cluster centers and intra-cluster member distributions of each aggregated flow, compares them with the cluster center positions recorded in the previous period, and calculates the offset of each cluster center. If the cluster center drift of an aggregated flow exceeds a preset threshold, it indicates that the business composition of that aggregated flow has changed significantly, triggering re-clustering and updating the FM-5QI mapping strategy accordingly; otherwise, only the load score of each 5QI candidate value is refreshed, and the 5QI selection is fine-tuned within the candidate interval to maintain the continuity of the mapping strategy. This mechanism avoids frequently performing complete clustering and mapping update operations under steady-state conditions, effectively controlling the computational overhead of the system.

[0077] The event triggering mechanism is as follows: When TSN-AF detects a GOOSE burst event, it triggers an emergency mapping review. GOOSE burst event detection is determined by the abrupt change in message arrival rate: according to the GOOSE retransmission mechanism of IEC 61850-8-1, messages are transmitted at a low frequency with a relatively large retransmission interval in steady state, and immediately after a state change, they are transmitted at the minimum retransmission interval. In continuous bursts, the rates of the two methods differ significantly. TSN-AF continuously counts packet arrivals using a sliding time window; when the number of arrivals within a unit window exceeds a configurable threshold... At that time, it was determined that a GOOSE emergency had occurred.

[0078] Once an emergency mapping review is triggered, the system will respond within milliseconds. Within a typical timeframe (<50ms, configured by the system according to distribution network fault response specifications), the system confirms or switches the preferred 5QI for protection-type aggregated flows. Specifically, the system immediately obtains the real-time load status of each high-priority 5QI (e.g., 5QI=82 / 83), recalculates the comprehensive score based on the membership information of the protection-type aggregated flows, and confirms whether the current mapping result is still optimal. If the preferred 5QI is overloaded, it quickly switches to the 5QI with the highest comprehensive score within the flexible candidate interval. This mechanism ensures that protection messages are always forwarded along the high-priority 5QI path, without mapping lag due to periodic review intervals, thereby ensuring the timeliness of protection actions in the event of a GOOSE fault.

[0079] This invention provides a QoS mapping method, apparatus, electronic device, and storage medium for 5G-TSN converged communication in power distribution networks. By using preset semantic vectors of power distribution network service categories as semantic priors, the method selects locations in the service data stream that correspond to the semantic prior meanings of various power distribution network services. This ensures that the real sample with the smallest distance from the clustering verification vector is used as the initial cluster center, anchoring the clustering starting point to a location with a clear distribution result that converges to various power distribution network services from the initial stage. This solves the problem of ineffective correspondence between clustering categories and service types caused by traditional random initialization, enabling the clustering results to serve as an effective decision-making basis for 5QI mapping. The weighted distance formula is the sum of the weighted squared differences of the normalized security partition level, packet delay budget, scheduling priority, and packet error rate. The weight coefficient of the security partition level is greater than that of other items, making the security partition level dominant in the clustering process. This ensures that the GOOSE protection services in Zone I always move closer to the high-priority center point in the distance calculation. Even if a large number of GOOSE short packets flood in, the real sample center point will not be pulled off by outliers, and Zone I services will still be clustered into the high-priority category. This avoids the misclassification of protection services due to cluster boundary drift, and improves the timeliness and accuracy of power grid protection actions.

[0080] Existing methods establish a static correspondence between service types and 5QIs, resulting in a fixed mapping strategy that cannot adapt to dynamic changes in network load. This invention introduces a fuzzy membership-based elastic mapping mechanism for 5QIs. In this embodiment, a triangular membership function is used to calculate the comprehensive membership degree between each cluster and the candidate 5QI values. Based on dual thresholds, the candidate 5QI values ​​are divided into a preferred interval, an elastic candidate interval, and an exclusion interval. Simultaneously, dynamic optimization is performed by considering the real-time load status of each 5QI. Within the preferred interval, 5QIs with load rates not exceeding the overload threshold are selected from high to low based on the comprehensive score. If the entire preferred interval is overloaded, the system switches to the elastic candidate interval. This mechanism allows the same cluster to correspond to different 5QIs under different load conditions. It can select the optimal matching 5QI to ensure latency under normal load and dynamically switch under heavy load to avoid congestion, overcoming the low resource utilization of fixed mapping tables under high load.

[0081] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0082] Those skilled in the art will clearly understand that, for convenience and simplicity, the specific working process of the device described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0083] One embodiment of the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the QoS mapping method for 5G-TSN converged communication in power distribution networks according to any one of the present invention.

[0084] Electronic devices can be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. Electronic devices may include, but are not limited to, processors and memory.

[0085] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting all parts of the electronic device through various interfaces and lines.

[0086] Memory can be used to store computer programs. The processor implements various functions of electronic devices by running or executing computer programs stored in memory and accessing data stored in memory. Memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc.; the data storage area can store data created based on the use of the mobile phone, etc. In addition, memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital cards (SD cards), flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0087] One embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the QoS mapping method for 5G-TSN converged communication in power distribution networks according to any one of the present invention.

[0088] The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added to or subtracted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0089] One embodiment of the present invention provides a computer program product that, when run on a terminal device, causes the electronic device to execute the QoS mapping method for 5G-TSN converged communication in power distribution networks as described above.

[0090] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A QoS mapping method for 5G-TSN converged communication in power distribution networks, characterized in that, include: Obtain n business data streams to be mapped, each of which includes security partition level, packet latency budget, scheduling priority, and packet error rate; Using k pre-defined semantic vectors of power distribution network business categories as semantic prior vectors, the business data streams with the smallest distance from the semantic prior vectors are selected from the n business data streams as k cluster centers. For each of the aforementioned business data flows, the weighted distance between the business data flow and k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data flows are then assigned to the cluster containing the cluster center that minimizes the weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data flow and the k cluster centers. Perform iterative optimization on the k cluster centers until the convergence condition is met, and determine the final k cluster centers and the corresponding k clusters; The k clusters are mapped to their corresponding 5QI values.

2. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 1, characterized in that, Mapping the k clusters to their corresponding 5QI values ​​includes: For each cluster, the security partition level and data packet delay budget corresponding to the cluster center are used as the security partition level and data packet delay budget of the cluster. Based on the security partition level and packet latency budget of each cluster, candidate 5QI values ​​for each cluster are determined from the preset 5QI set; The triangular membership function is used to calculate the membership degree between the cluster and each 5QI candidate value in three dimensions: packet delay budget, scheduling priority and packet error rate. The membership degree of the three dimensions is then weighted and merged into a comprehensive membership degree. 5QI candidate values ​​with a comprehensive membership degree greater than a preset first threshold are assigned to the preferred interval; 5QI candidate values ​​with a comprehensive membership degree less than the first threshold but greater than a second threshold are assigned to the flexible candidate interval; and 5QI candidate values ​​with a comprehensive membership degree less than the second threshold are excluded; wherein, the first threshold is greater than the second threshold. Obtain the channel resource utilization rate for each 5QI candidate value, and calculate the load score based on the channel resource utilization rate; The comprehensive membership degree and load score are weighted and summed to obtain the comprehensive score. Within the preferred range, each 5QI candidate value is sorted from high to low according to the comprehensive score. Based on the sorting result, the load score of each 5QI candidate value is compared with the preset overload threshold in order from high to low. If there is a 5QI candidate value with a load score less than the overload threshold within the preferred interval, then the first 5QI candidate value with a load score less than the overload threshold is selected as the mapping result. If the load score of each 5QI candidate value in the preferred interval is greater than or equal to the overload threshold, then switch to the elastic candidate interval and select the 5QI candidate value with the highest comprehensive score from the elastic candidate interval as the mapping result.

3. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 1, characterized in that, The semantic vector of the distribution network service category is a semantic feature vector constructed based on the typical distribution network services defined by the IEC 61850 standard; The semantic vectors of the distribution network service categories include semantic feature vectors corresponding to GOOSE tripping, GOOSE non-tripping, SV sampled value, remote control, and telemetry. Each semantic feature vector corresponds to the security partition level of the typical distribution network service.

4. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 1, characterized in that, The iterative optimization of the k cluster centers until the convergence condition is met, to determine the final k cluster centers and the corresponding k clusters, includes: For each cluster, each non-cluster center business data flow within the cluster is assumed to be a candidate replacement cluster center, and the overall clustering cost difference before and after replacing the current cluster center with each candidate replacement cluster center is calculated; wherein, the overall clustering cost difference is the sum of the weighted distances of all business data flows within the cluster to their respective cluster centers; If the calculated overall clustering cost difference is negative, the assumptions about each non-cluster center are terminated, and the current cluster center is replaced by a candidate alternative cluster center as the cluster center of the cluster. The business data streams of each non-cluster center in the cluster are then assumed as candidate alternative cluster centers and iterated again until the overall cost difference in the cluster is not negative. The final k cluster centers and their corresponding k clusters are determined when the overall clustering cost difference of all clusters is no longer negative.

5. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 1, characterized in that, The weighted distance is calculated using the following formula: ； in, This is the nth business data stream; Let k be the cluster center; For business data flow and cluster center Weighted distance between them; For the nth business data stream The security partition level ranges from 1 to 4, corresponding to zones one through four, respectively. For the k-th cluster center Security partition level; For the nth business data stream Data packet latency budget; Cluster center Data packet latency budget; This is a set of QoS parameters, including scheduling priority and packet error rate; for The index; For the nth business data stream In the The values ​​of each QoS parameter; For the k-th cluster center In the The values ​​of each QoS parameter; These are the weighting coefficients for the security partition level dimension; These are the weighting coefficients for the data packet latency budget dimension; For the first The weight coefficients of each QoS parameter dimension, and .

6. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 5, characterized in that, When the security partition level corresponding to the service data flow is Zone 1, the priority of the service data flow is higher than the priority of service data flows in other security partition levels; wherein, the priority of the service data flow is based on the weighting coefficient of the security partition level of the service data flow. The difference between the weighting coefficients of the data packet delay budget is determined.

7. The QoS mapping method for 5G-TSN converged communication in power distribution networks as described in claim 2, characterized in that, The formula for calculating the comprehensive membership degree is: ； in, This represents the combined membership degree between the k-th cluster and the j-th 5QI candidate value; This represents the k-th cluster; This is the j-th 5QI candidate value; Membership degree for the data packet latency budget dimension; Membership degree is the dimension of scheduling priority; Membership degree for the packet error rate dimension; , , For the corresponding weight coefficients, and , .

8. A QoS mapping device for 5G-TSN converged communication in power distribution networks, characterized in that, include: The data acquisition module is used to acquire n business data streams to be mapped. Each business data stream includes the security partition level, data packet latency budget, scheduling priority, and data packet error rate. The clustering module is used to select the business data streams with the smallest distance to the semantic prior vectors from among the n business data streams, using pre-defined k semantic prior vectors as semantic prior vectors. For each business data stream, the weighted distance between the business data stream and the k cluster centers is calculated based on the corresponding security partition level, data packet delay budget, scheduling priority, and data packet error rate. The business data streams are then assigned to the cluster containing the cluster center with the smallest weighted distance. Among these, the security partition level has the largest weight coefficient when calculating the weighted distance between the business data stream and the k cluster centers. The iterative optimization module is used to perform iterative optimization on the k cluster centers until the convergence condition is met, and to determine the final k cluster centers and the corresponding k clusters. The mapping module is used to map the k clusters to their corresponding 5QI values.

9. An electronic device, characterized in that, include: A processor, a memory, and a computer program stored in the memory and configured to be executed by the processor; wherein, when the processor executes the computer program, it implements the QoS mapping method for 5G-TSN converged communication in a power distribution network as described in any one of claims 1-7.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to perform the QoS mapping method for 5G-TSN converged communication in power distribution networks as described in any one of claims 1-7.

Images