High-voltage transformer remote metering and monitoring system based on 5G network slicing

Abstract

The invention relates to the technical field of wireless communication, and particularly discloses a high-voltage transformer remote metering monitoring system based on 5G network slicing, which senses network performance in real time by constructing a multi-dimensional network state matrix, performs path rehearsal and resource allocation calculation by using a digital twinning technology, and realizes remote metering monitoring of a high-voltage transformer. Generating a candidate transmission path set with optimal time delay consistency and a corresponding wireless resource block allocation scheme; calculating a path resource allocation factor and combining with network load prediction to generate an optimal resource scheduling strategy; and finally, converting the strategy into a specific network control instruction, establishing a special transmission channel with deterministic service quality, and implementing real-time monitoring and maintenance.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a high-voltage mutual inductor remote metering monitoring system based on 5G network slices. BACKGROUND

[0002] With the rapid development of smart grids, the remote metering and state monitoring of high-voltage mutual inductors put high requirements on communication networks. 5G networks, with their high bandwidth and low latency, have become ideal transmission carriers, and their network slicing technology can theoretically provide exclusive virtual channels for power business. However, in actual deployment, existing 5G-based network systems still use the traditional IP network's "best effort" routing mechanism and shared wireless resource scheduling strategy The existing 5G system encounters the deep technical problem of "path chaotic drift" when carrying the synchronous metering business of widely distributed high-voltage mutual inductors. Due to the dynamic optimization and load balancing mechanism of network routing, mutual inductor sampling data packets from different geographical locations with strict time sequence correlation will experience asymmetric and randomly changing network paths and time delays during transmission. This path uncertainty causes key applications such as traveling wave distance measurement, which rely on nanosecond-level time synchronization accuracy, to completely fail. The system cannot distinguish whether the time difference of the data packet is the real power grid fault propagation delay or the transmission time delay noise introduced by the inconsistent network path, causing the fault location function to fail. This problem is extremely hidden, and conventional network performance monitoring only focuses on connectivity and average delay, and cannot perceive and measure this inconsistency at the path level, thus becoming a serious technical obstacle to the deep application of 5G networks in precise power control. SUMMARY

[0003] The purpose of the present application is to provide a high-voltage mutual inductor remote metering monitoring system based on 5G network slices to solve the problems in the above background.

[0004] The purpose of the present application can be achieved by the following technical solutions: The high-voltage mutual inductor remote metering monitoring system based on 5G network slices comprises: a state monitoring and data acquisition module for real-time acquisition of performance parameters of each network node in the 5G network slice and construction of a multi-dimensional network state matrix containing time delay, packet loss rate and available bandwidth; a path calculation and analysis module for receiving the multi-dimensional network state matrix, performing resource availability analysis and outputting a candidate transmission path set and a wireless resource block allocation scheme for each path; a resource scheduling and decision module for calculating a path resource allocation factor based on the received candidate transmission path set and wireless resource block allocation scheme, and generating an optimal resource scheduling strategy in combination with network load prediction data; The control instruction generation module is configured to convert the received optimal resource scheduling strategy into network control instructions containing time-frequency resource reservation parameters, transmission priority configuration and core network function scheduling rules; The channel establishment and management module is configured to execute the received network control instructions, establish a dedicated transmission channel with a determined quality of service guarantee in the physical network, and perform real-time monitoring and dynamic maintenance on the dedicated transmission channel.

[0005] As a further scheme of the present application, the construction process of the multi-dimensional network state matrix is as follows: In a preset time window, a plurality of rounds of probe detection are performed on each network node to collect end-to-end delay, packet loss rate and available bandwidth data of the transmission path between nodes to form an original performance data set; The original performance data is organized into a space-time correlation matrix based on the network topology, wherein the row dimension of the space-time correlation matrix corresponds to different detection time points, the column dimension corresponds to transmission links in the network, and each matrix element contains a three-tuple measurement value of the delay, packet loss rate and available bandwidth of the transmission link at the corresponding time point; The space-time correlation matrix is subjected to sliding window normalization processing to form an enhanced feature vector; The enhanced feature vector is reorganized according to the hierarchical structure of the network topology to construct a multi-dimensional network state matrix with time continuity and spatial correlation.

[0006] As a further scheme of the present application, the output candidate transmission path set specifically includes: Based on the multi-dimensional network state matrix, the delay fluctuation coefficient and the packet loss risk index of each potential path are calculated, wherein the delay fluctuation coefficient is obtained by calculating the ratio of the standard deviation of the delay of the path to the average delay in each time dimension, and the packet loss risk index is obtained by analyzing the distribution characteristics and burst packet loss probability of the historical packet loss rate; A path comprehensive quality evaluation function is constructed, which considers both the basic performance indicators and the stability indicators of the path, wherein the basic performance indicators include the average delay and the available bandwidth, and the stability indicators include the delay fluctuation coefficient and the packet loss risk index, and the path quality score is formed by weighted combination; According to the specific requirements of the high-voltage mutual inductor remote metering service, double threshold conditions for path screening are set, the first threshold ensures that the basic performance of the path meets the minimum requirements of the metering data transmission, and the second threshold ensures that the stability of the path meets the reliability standards required by the metering service; From the path set satisfying the double threshold conditions, a number of paths with the highest comprehensive quality scores are selected to form the candidate transmission path set.

[0007] As a further scheme of the present application, the calculation process of the wireless resource block allocation scheme is as follows: The load state and interference of each link in the candidate transmission path are analyzed, and based on historical bandwidth utilization and real-time load data in the multi-dimensional network state matrix, the resource demand of each link in the future time period is predicted; According to the characteristics of the high-voltage transformer metering data, the wireless resource blocks are divided into two types of guarantee resource blocks and elastic resource blocks, wherein the guarantee resource blocks are used for transmitting key synchronous metering data, and the elastic resource blocks are used for transmitting non-real-time state monitoring data. Based on the load prediction results of each path and the service demand characteristics, the optimal resource block combination scheme is calculated for each candidate path.

[0008] As a further scheme of the application, the calculation of the path resource allocation factor specifically comprises: Based on the received wireless resource block allocation scheme, the resource allocation density characteristics of each path are extracted, and the resource allocation density index is obtained by calculating the ratio of the number of resource blocks allocated in a unit time to the total available resource blocks. The deviation degree of the actual use efficiency of the resource block from the theoretical capacity is monitored, and the resource utilization efficiency index is generated in combination with the idle time proportion statistics of the resource block. The fluctuation coefficient of the resource allocation scheme is calculated by statistically analyzing the adjustment frequency and amplitude of the resource allocation scheme in a preset time period, and the resource stability index is formed. The resource allocation density index, the resource utilization efficiency index and the resource stability index are normalized, and the path resource allocation factor is calculated by using a dynamic weight distribution method, wherein the weight coefficient is dynamically adjusted according to the actual demand of the high-voltage transformer metering service.

[0009] As a further scheme of the application, the generation of the optimal resource scheduling strategy in combination with the network load prediction data specifically comprises: The distribution law of the network load in the time and space dimensions is analyzed, and the load change trend of each network region in a future specific time period is predicted. The path resource allocation factor and the network load prediction data are analyzed in multiple dimensions, the matching degree of resource demand and supply is identified, and a key path set that needs to be preferentially scheduled is determined. Considering multiple optimization targets such as resource utilization efficiency, path stability and load balancing degree, an initial resource scheduling scheme is generated by using a dynamic programming algorithm. The feasibility of the initial resource scheduling scheme is verified and the effect is evaluated, and according to the evaluation result, the scheme is optimized and iterated by using a heuristic adjustment rule, and finally the optimal resource scheduling strategy meeting the quality of service requirements of the high-voltage transformer metering service is generated.

[0010] As a further scheme of the application, the generation process of the network control instruction is: Based on the resource allocation requirement in the optimal resource scheduling strategy, a resource block combination coding technology is used to generate time-frequency resource reservation parameters, adjacent resource blocks are combined and coded according to service demand characteristics to form a resource unit group with specific transmission characteristics; According to different service types and urgency of the high-voltage mutual inductor measurement data, a transmission priority configuration matrix is constructed, the transmission priority configuration matrix comprehensively considers the real-time requirement of data, the importance of service and the network state factors, and dynamically allocates differentiated transmission priorities for different types of data flow; For core network function scheduling, a user plane function selection mechanism based on service awareness is designed, and according to the transmission path characteristics and processing requirements of the measurement data, the optimal core network user plane function instance is selected and the corresponding traffic forwarding rule is configured; The generated time-frequency resource reservation parameters, transmission priority configuration and core network function scheduling rule are instruction encapsulation to form a network control instruction which can be directly issued to the network equipment.

[0011] As a further scheme of the application, the physical network establishes a dedicated transmission channel with a determined quality of service guarantee, specifically comprising: The received network control instruction is configured for parameter verification, the compatibility between instruction parameters is detected through simulation execution environment, potential execution conflicts are identified and corresponding parameter adjustment scheme is generated; The user plane function rule is configured at the core network level, the priority queue is set at the transmission network level, and the time-frequency resource block is allocated at the wireless access network level, so that the configuration of each network level is effectively coordinated; The actual establishment effect of the dedicated transmission channel is verified through end-to-end signaling detection, and the compliance degree of the measured channel parameters and the expected quality of service indicators is compared; The channel establishment process is implemented with fallback protection, when the channel establishment fails or the performance is not up to standard, the preset backup channel establishment scheme is automatically triggered to ensure the continuity of the high-voltage mutual inductor measurement service transmission.

[0012] As a further scheme of the application, the dedicated transmission channel is monitored and dynamically maintained in real time, specifically comprising: A multi-dimensional channel quality monitoring system is constructed, and the delay jitter, packet loss rate, bandwidth utilization and bit error rate indicators of the transmission channel are synchronously collected; The change trend of the channel quality indicators is analyzed, when the indicators abnormally fluctuate, the network section where the abnormality occurs and the reason are automatically located; An adaptive maintenance strategy library is designed, and corresponding maintenance schemes are preset for different types of channel abnormalities, including parameter adjustment, path switching and resource reallocation; Based on the channel quality historical data, preventive maintenance operation is actively performed before the channel performance declines.

[0013] Advantages of the present application: (1) The present application uses service awareness and digital twin technology to pre-calculate and establish a shared transmission path with optimal time delay consistency for an associated group of mutual inductor terminals, thereby fundamentally eliminating path randomness and ensuring that data from different nodes experiences nearly consistent network time delay during transmission. This ensures that the time difference of data received by the monitoring center accurately reflects the physical state of the power grid, thereby improving the accuracy of advanced analysis functions such as traveling wave fault location.

[0014] (2) In a shared wireless access network, the time-frequency resource blocks reserved for high-voltage metering services may experience unexpected, service-transparent scheduling conflicts or resource preemption due to "invisible" load balancing or resource restructuring operations performed by the network to optimize overall efficiency. This "implicit conflict" does not cause traditional link interruption alarms, but may result in random microsecond-level jitter or loss of metering data packets, thereby compromising the real-time and deterministic nature of metering control.

[0015] The present application establishes a "dynamic soft private line" based on service intent, thereby deeply integrating resource reservation and scheduling strategies. By performing a pre-rehearsal and conflict detection in the digital twin before any optimization operation is performed by the network, potential resource conflicts can be actively identified and avoided. This establishes a "virtual exclusive" transmission channel for high-voltage mutual inductor metering services, thereby ensuring resource exclusivity and transmission certainty for the service even when public network resources are dynamically adjusted. This reduces the risk of control command loss or abnormal sampling data caused by network-side factors by an order of magnitude, thereby improving the reliability of high-voltage power grid remote metering and control. BRIEF DESCRIPTION OF DRAWINGS

[0016] The present application will be further described below with reference to the accompanying drawings.

[0017] Figure 1 is a flowchart of the system of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of the present application.

[0019] Please refer to Figure 1 The present application is a high-voltage mutual inductor remote metering monitoring system based on 5G network slicing, as shown in the figure, which includes: The state monitoring and data acquisition module is configured to collect performance parameters of each network node in the 5G network slice in real time and construct a multi-dimensional network state matrix including time delay, packet loss rate and available bandwidth. The path calculation and analysis module is configured to receive the multi-dimensional network state matrix, perform resource availability analysis and output a candidate transmission path set and a radio resource block allocation scheme for each path. The resource scheduling and decision module is configured to calculate a path resource allocation factor based on the received candidate transmission path set and the radio resource block allocation scheme, and generate an optimal resource scheduling strategy in combination with network load prediction data. The control instruction generation module is configured to convert the received optimal resource scheduling strategy into network control instructions including time-frequency resource reservation parameters, transmission priority configuration and core network function scheduling rules. The channel establishment and management module is configured to execute the received network control instructions, establish a dedicated transmission channel with a determined quality of service guarantee in the physical network, and perform real-time monitoring and dynamic maintenance on the dedicated transmission channel.

[0020] In the state monitoring and data acquisition module, multiple rounds of probe detection are performed on each network node within a preset 1-minute time window. In specific implementation, a probe server deployed at the network edge sends detection data packets of a specific format to each network node. These data packets contain accurate timestamp information for calculating end-to-end transmission time delay. At the receiving end, the packet loss rate is calculated by counting the difference between the number of received data packets and the number of sent data packets. The available bandwidth is measured using a transmission rate-based detection method, which adjusts the sending rate to observe the change in receiving throughput to determine the available bandwidth range. Each round of detection lasts 5 seconds, and 12 rounds of detection are completed within a 1-minute time window to form an original performance data set including three dimensions of time delay, packet loss rate and available bandwidth.

[0021] The original performance data is organized into a space-time correlation matrix based on the network topology. In the specific implementation process, the collection time point is taken as the row dimension of the matrix, arranged in chronological order, and each time point corresponds to a detection period. The transmission links in the network are taken as the column dimension of the matrix, arranged according to the connection relationship in the network topology. Each matrix element stores the performance data triplet of the corresponding transmission link at a specific detection time point, including time delay value, packet loss rate value and available bandwidth value. The time delay value is recorded in milliseconds, accurate to three decimal places; the packet loss rate is recorded in percentage, with two decimal places; and the available bandwidth is recorded in megabits per second, with integer values. This organization structure enables the matrix to reflect the change characteristics of network state in both time and space dimensions.

[0022] Then, the spatiotemporal correlation matrix is normalized by a sliding window. A fixed-length sliding time window is adopted, and the window size is set to 10 consecutive detection time points. Within each sliding window, the statistical features of the performance indicators of each transmission link are calculated. For the delay data, the delay fluctuation coefficient in the current window is calculated, which is obtained by calculating the ratio of the standard deviation to the average of the delay values in the window. For the packet loss rate data, the packet loss change gradient is calculated, which is obtained by comparing the difference in packet loss rate between adjacent time points and taking the average. For the available bandwidth data, the bandwidth stability indicator is calculated, which is represented by the coefficient of variation of the bandwidth values in the window. These statistical features are combined with the original measurement values to form an enhanced feature vector containing six dimensions, including the original delay value, the delay fluctuation coefficient, the original packet loss rate, the packet loss change gradient, the original bandwidth value, and the bandwidth stability indicator.

[0023] The enhanced feature vector is reorganized according to the hierarchical structure of the network topology. According to the characteristics of the 5G network architecture, the network is divided into three levels: access layer, aggregation layer, and core layer. The enhanced feature vectors of the transmission links belonging to the same network level are combined into sub-matrices, each of which represents the state snapshot of the network level at a specific time granularity. These sub-matrices are combined vertically according to the network level relationship and horizontally expanded according to the time sequence to form a multi-dimensional network state matrix with temporal continuity and spatial correlation. The time dimension of this matrix retains all the detection time points, and the spatial dimension is organized according to the hierarchical structure of the network level and the transmission link, each matrix unit contains six-dimensional performance indicator data. Through this structure, the performance change trend of a single transmission link can be observed, and the state correlation between different network levels can be analyzed.

[0024] In the path calculation and analysis module, after obtaining the multi-dimensional network state matrix, the candidate transmission path set is first generated. In implementation, the delay data sequence of each potential path at 30 consecutive sampling time points is extracted from the multi-dimensional network state matrix. The calculation process of the delay fluctuation coefficient is as follows: for each potential path, the average delay value of the path at 30 time points is calculated, then the standard deviation of these delay values relative to the average value is calculated, and finally the delay standard deviation is divided by the delay average value, and the calculation result is multiplied by 100 to convert it to percentage form. The calculation of the packet loss risk index adopts a double evaluation method: first, the packet loss rate distribution of the path in the last 100 samples is counted, and the proportion of the number of abnormal times when the packet loss rate exceeds 1% to the total sampling times is recorded; second, the change trend of the packet loss rate in the consecutive sampling period is analyzed, and the average value of the difference in packet loss rate between adjacent periods is calculated. The two evaluation results are weighted and summed according to the weight of 6:4 to obtain the final packet loss risk index, and the larger the index value, the higher the packet loss risk.

[0025] The construction of the path synthesis quality evaluation function adopts a multi-index weighted evaluation method. The basic performance index includes two dimensions of average delay and available bandwidth, wherein the average delay takes the average value of the last 10 sampling periods, and the available bandwidth takes the current latest measurement value. The stability index includes the delay fluctuation coefficient and the packet loss risk index calculated as described above. The four indexes are respectively given different weight coefficients: the average delay weight is 0.3, the available bandwidth weight is 0.25, the delay fluctuation coefficient weight is 0.25, and the packet loss risk index weight is 0.2. The weight distribution ratio is determined according to the characteristics of the high-voltage transformer measurement business being sensitive to delay. The calculation formula of the path quality score is: the numerical value of each index is multiplied by its corresponding weight and then added, wherein the delay-related index (average delay, delay fluctuation coefficient) adopts a negative scoring method, that is, the larger the numerical value, the lower the score; the bandwidth index adopts a positive scoring method, the larger the numerical value, the higher the score.

[0026] The specific implementation of the double threshold condition of path screening is as follows: the first threshold is for the basic performance index, requiring that the average delay of the path is not more than 10 milliseconds, and the available bandwidth is not less than 5 Mbps. These two threshold values are determined according to the characteristics of the high-voltage transformer measurement data, and the measurement data packet size is usually in the range of 2-5KB, and the transmission frequency is 100 data packets per second, and it is calculated that at least 4Mbps of stable bandwidth is required. The second threshold is for the stability index, requiring that the delay fluctuation coefficient is less than 15% and the packet loss risk index is lower than 20 points. The delay fluctuation coefficient threshold ensures the time synchronization accuracy of the measurement data sampling, and the packet loss risk index threshold ensures the reliability of data transmission. In implementation, the first threshold is applied to screen out paths that meet the basic transmission requirements, and then the second threshold is applied to further screen among these paths.

[0027] The final determination of the candidate transmission path set adopts the quality score sorting method. For all paths screened by the double threshold, the paths are sorted from high to low according to the path quality score. According to the network size and application requirements, the top 5 paths are selected to form the candidate transmission path set. In actual implementation, when there are paths with the same score, the path with lower average delay is preferred. The candidate transmission path set contains at least 3 paths and at most not more than 8 paths, so as to ensure that there is enough selection margin and the management complexity is not increased due to too many paths.

[0028] The calculation of the wireless resource block allocation scheme first carries out load state analysis. Based on the historical data stored in the multi-dimensional network state matrix, the bandwidth utilization change of each candidate path in the past 60 minutes is extracted. A sliding window prediction method is adopted, with 15 minutes as a time window, to calculate the average bandwidth utilization in each window and analyze its change trend. Combined with real-time load data, the resource demand of each link in the next 15 minutes is predicted by comparing the difference between the current load and the historical load at the same period. The prediction result includes two values of basic resource demand and peak resource demand, the basic resource demand is 1.1 times the average value of the historical same period, and the peak resource demand is 0.9 times the maximum value of the historical same period.

[0029] The classification of wireless resource blocks is divided according to the characteristics of the transmission content. The guarantee resource block is used to transmit key synchronous metering data, which has the characteristics of small data volume, fixed transmission interval and high timeliness requirement. Each guarantee resource block allocates fixed time-frequency resources and adopts exclusive allocation method to ensure that metering data can be transmitted in time. The elastic resource block is used to transmit non-real-time state monitoring data, including device running state, environmental parameters, etc. This kind of data has relatively low transmission timeliness requirement, but the data volume may be large. The elastic resource block adopts shared allocation method, and multiple services can dynamically share these resource blocks.

[0030] The calculation of the resource block combination scheme adopts the demand matching method. For each candidate path, first, the number of guarantee resource blocks is allocated according to the predicted basic resource demand, and the calculation formula is guarantee resource block number equals to basic resource demand divided by the carrying capacity of a single resource block. The carrying capacity of a single resource block is determined according to the encoding method and modulation strategy. When 16QAM modulation and 3 / 4 encoding rate are adopted, a single resource block can carry about 1.2 Mbps of data volume. Then the number of elastic resource blocks is determined according to the difference between the peak resource demand and the basic resource demand, and the elastic resource block number equals to the difference between the peak and the basic resource demand divided by the carrying capacity of a single resource block. Finally, the resource block combination scheme is formed, which clearly marks the type, number and scheduling priority of each resource block. The guarantee resource block has the highest scheduling priority and is given priority to guarantee its transmission demand when resources are scarce.

[0031] In the resource scheduling and decision module, the calculation of the path resource allocation factor first starts from the processing of the wireless resource block allocation scheme. In implementation, 15 minutes is selected as a complete statistical period, and the number of resource blocks allocated to each path is recorded in this period. The calculation method of the resource allocation density index is: the total number of resource blocks allocated to a single path in the statistical period is divided by the total number of available resource blocks in the network in the period, and then multiplied by 100 to convert it to a percentage. For example, a certain path is allocated 150 resource blocks in 15 minutes, and the total available resource blocks of the network are 5000, then the resource allocation density index of the path is 3%. This calculation process is performed for each path in the candidate transmission path set respectively, to ensure that each path can obtain the corresponding resource allocation density index value.

[0032] The calculation of the resource utilization efficiency index adopts the comparison method of actual usage and theoretical capacity. First, monitor the actual data transmission volume of each resource block in the allocation period, and record its percentage of reaching the theoretical maximum capacity. The theoretical capacity is determined according to the modulation and coding scheme adopted, such as using 64QAM modulation and 5 / 6 coding rate, the theoretical capacity of a single resource block is 1.8 Mbps. The idle time proportion of the resource block is calculated by counting the time segment when the resource block is in the allocated state but no data is transmitted. The actual use efficiency and the idle time proportion are combined, and the calculation formula of the resource utilization efficiency index is: the actual use efficiency multiplied by (1 minus the idle time proportion). For example, the actual use efficiency of a certain resource block is 85%, and the idle time proportion is 10%, then its resource utilization efficiency index is 76.5%. After calculating the index of all resource blocks allocated to each path, the average value is taken to obtain the overall resource utilization efficiency index of the path.

[0033] The resource stability index is calculated by analyzing the historical adjustment records of the resource allocation scheme. The statistical time period is set to the last 2 hours, and 15 minutes is taken as an adjustment period, which contains 8 adjustment periods. The number of resource allocation scheme adjustments in this period and the change amplitude of the resource block quantity at each adjustment are counted. The calculation of the fluctuation coefficient is divided into two parts: the adjustment frequency coefficient is equal to the ratio of the actual adjustment times to the maximum possible adjustment times, and the adjustment amplitude coefficient is equal to the ratio of the average value of the absolute change amount of each adjustment to the initial allocation amount. The resource stability index is calculated by weighting the adjustment frequency coefficient and the adjustment amplitude coefficient according to the weight of 4:6, which is: 0.4 times (1 minus the adjustment frequency coefficient) plus 0.6 times (1 minus the adjustment amplitude coefficient). The closer the index value is to 1, the more stable the resource allocation scheme is.

[0034] The normalization of the three index values adopts the minimum-maximum normalization method. The minimum value in all path values of each index is mapped to 0, the maximum value is mapped to 1, and other values are converted in a linear relationship. The dynamic weight allocation is adjusted according to the specific needs of the high-voltage mutual inductor measurement business: in the business peak period, the weight of the resource allocation density index is set to 0.4, the weight of the resource utilization efficiency index is set to 0.35, and the weight of the resource stability index is set to 0.25; in the business stable period, the weights of the three indexes are adjusted to 0.3, 0.4, and 0.3, respectively. The calculation formula of the path resource allocation factor is the sum of the three normalized index values multiplied by their corresponding weights. The adjustment basis of the weight coefficient is the real-time monitoring result of the business data volume. When the data volume in the last 5 minutes exceeds the threshold value, it automatically switches to the peak period weight configuration.

[0035] The generation of network load prediction data adopts the space-time analysis method. In the time dimension, the load change law of each 15-minute time period in the past 24 hours is analyzed, and a load change curve with a period of one hour is established. In the spatial dimension, the network is divided into multiple regions, each region containing a number of adjacent network nodes, and the load correlation between regions is analyzed. The load prediction uses a weighted moving average method based on historical data, with the weight of the most recent time period being 0.5, the weight of the previous same period being 0.3, and the weight of the previous two same periods being 0.2. The prediction result includes the basic load value and the peak load value. The basic load value is the average value after weighted calculation, and the peak load value is 90% of the maximum value of the same period in history. The prediction time range is set to the next 30 minutes, and the prediction result is updated every 5 minutes.

[0036] The matching analysis of resource demand and supply is realized by establishing a correlation matrix. The rows of the matrix represent various paths, and the columns represent different network regions. The value of the matrix element represents the matching degree of the resource demand and supply of the path in the region, and the calculation method is the product of the path resource allocation factor and the load prediction value of the region. Paths with a matching degree value exceeding 0.8 are marked as critical paths, which have a higher priority for resource allocation. At the same time, for paths with a matching degree less than 0.3, resource allocation or path replacement is considered. The determination of the critical path set also needs to consider the redundancy relationship between paths to ensure that the failure of a single path does not affect the overall business operation.

[0037] The generation of the initial resource scheduling scheme adopts a multi-objective optimization method. The optimization objectives include: the resource utilization efficiency should reach more than 75%, the path stability index should not be less than 0.7, and the load balancing degree difference of each network area should be controlled within 20%. The state definition of the dynamic programming algorithm is the resource allocation of each time slice, and the state transition considers the resource allocation possibility between different paths. The algorithm starts from the initial state, gradually constructs the resource scheduling scheme through state transition, and records the achievement of the three optimization objectives at each decision step. The termination condition of the algorithm is to find a scheme that satisfies the three optimization objective constraints at the same time, or to reach the maximum iteration number of 100 times. The generated initial resource scheduling scheme includes specific resource block allocation details, scheduling time arrangement and priority setting of each path.

[0038] The scheme optimization iteration process adopts heuristic adjustment rules. The feasibility verification checks whether the resource scheduling scheme meets the network physical limitations, including node processing capacity, link bandwidth limitations and other constraint conditions. The effect evaluation is carried out by simulating the actual running environment, and the evaluation indexes include end-to-end delay, packet loss rate and throughput. The heuristic adjustment rules include: when the delay is not up to standard, the resource allocation of the key path is adjusted first; when the packet loss rate is too high, the resource configuration of the retransmission mechanism is increased; when the throughput is insufficient, the combination method of the resource block is optimized. After each adjustment, the effect is evaluated again until all indexes meet the quality of service requirements of the high-voltage mutual inductor measurement business. The finally generated resource scheduling scheme clearly specifies the allocation object, use period and scheduling priority of each resource block, ensuring the data transmission quality of the measurement business.

[0039] In the control instruction generation module, the generation process of the time-frequency resource reservation parameters is based on the resource allocation requirements determined in the optimal resource scheduling strategy. In implementation, first, the number of resource blocks required by each service flow and its distribution characteristics are analyzed. The specific operation of resource block combination coding is as follows: 3 adjacent resource blocks are taken as a basic combination unit, and each combination unit is assigned a unique identification code, which contains the time slot number, subcarrier index and combination type information of the resource block. The combination type is divided into four types according to the service demand characteristics: low latency type combination, high reliability type combination, large bandwidth type combination and balanced type combination. The low latency type combination adopts a shorter time slot length, and each time slot length is 0.5 milliseconds; the high reliability type combination adopts a repeated transmission mechanism, and the same data is repeatedly transmitted 2 times on adjacent resource blocks; the large bandwidth type combination occupies more subcarrier resources, and each combination contains 6 consecutive subcarriers; the balanced type combination is uniformly distributed in time domain and frequency domain. The formation of the resource unit group is realized by splicing multiple basic combination units according to the service demand, and each resource unit group contains 2-8 basic combination units, and the specific number is determined by the service data volume. Finally, these combination coding information is mapped to the actual time-frequency resource grid to generate a set of time-frequency resource reservation parameters containing detailed parameters such as starting time slot, duration, subcarrier range, combination type, etc.

[0040] The construction process of the transmission priority configuration matrix is based on a multi-dimensional evaluation method. The dimensions of the matrix include data real-time requirement, service importance and network state. The data real-time requirement is divided into three levels according to the type of service data: level 1 is real-time control instruction, requiring transmission delay less than 10 milliseconds; level 2 is real-time metering data, requiring transmission delay less than 20 milliseconds; level 3 is historical data and state information, requiring transmission delay less than 100 milliseconds. The service importance is divided into four levels according to the role of data: level 1 is protection control data; level 2 is metering and billing data; level 3 is monitoring and analysis data; level 4 is management and maintenance data. The network state factor is evaluated by three indicators of real-time monitoring network load rate, packet loss rate and delay jitter, and each indicator is divided into three state levels. The calculation of transmission priority uses the weighted scoring method, with data real-time requirement weight of 0.5, service importance weight of 0.3, and network state factor weight of 0.2. According to the weighted score result, the transmission priority is divided into 8 levels, level 1 is the highest priority, and level 8 is the lowest priority. The priority configuration matrix is updated in real time, with an update period of 1 second, ensuring that it can reflect the changes in network state in time.

[0041] The implementation of the core network user plane function selection mechanism includes two stages of function instance evaluation and selection. First, collect the performance data of all available user plane function instances, including processing delay, throughput capacity, current load rate and available port number, etc. The processing delay is measured by sending test data packets, and the average of 100 measurements is taken; the throughput capacity is determined by stress testing to determine the maximum processing capacity; the current load rate is obtained by monitoring CPU and memory usage; the available port number counts the number of service ports not currently occupied. Then, based on the transmission path characteristics and processing requirements of the measurement data, a selection factor calculation model is constructed. The selection factor is calculated by weighting four sub-factors: the delay factor weight is 0.4, the load factor weight is 0.3, the throughput factor weight is 0.2, and the location factor weight is 0.1. The delay factor is calculated according to the distance between the instance and the data source, the smaller the distance, the higher the delay factor value; the load factor is inversely proportional to the current load rate of the instance; the throughput factor is proportional to the remaining processing capacity of the instance; the location factor considers the network topology location advantage of the instance. The user plane function instance with the highest selection score is selected as the optimal selection. After selecting the instance, configure the corresponding traffic forwarding rules, including the exact matching conditions of source IP address range, destination IP address, protocol type, port number and service type, and the corresponding processing actions and next hop address.

[0042] The instruction encapsulation process adopts a hierarchical encapsulation method. First, the time-frequency resource reservation parameters are encoded according to the 3GPP standard format, including resource allocation type, resource block group size, time domain resource allocation, frequency domain resource allocation, etc. The resource allocation type field indicates the type of allocation scheme adopted, taking the value of 0 for continuous allocation and 1 for discrete allocation; the resource block group size field specifies the number of resource blocks contained in each resource block group, taking the value of 1-32; the time domain resource allocation field contains the starting symbol index and the number of symbols; the frequency domain resource allocation field contains the starting resource block index and the number of resource blocks. Second, the transmission priority configuration information is encoded into an 8-bit field, the first 3 bits represent the priority category, the middle 3 bits represent the scheduling weight, and the last 2 bits represent the preemption capability. The core network function scheduling rules are converted into a traffic description template and a set of policy rules, each rule containing matching conditions and execution actions. Finally, all parameters are assembled according to the predefined instruction format, the instruction header contains version number, instruction type, timestamp and sequence number, and the instruction body contains specific parameter sets. After encapsulation, the instruction format is verified to ensure that all mandatory fields are complete, the values are within a reasonable range, and there is no conflict between parameters. The instructions that pass the verification are marked as ready state and are ready to be issued to the corresponding network devices. The instruction issuance adopts a transaction mechanism to ensure atomic execution of the instructions, either all successful or all rolled back.

[0043] In the channel establishment and management module, the configuration parameter verification process is implemented by establishing a simulation test environment. The simulation environment contains virtual instances with the same configuration as the actual network equipment, which can simulate the execution process of instructions in the network. During verification, the received network control instructions are first parsed into various configuration parameters, including time slot allocation information in the time-frequency resource reservation parameter, subcarrier allocation information, queue scheduling parameters in the transmission priority configuration, and flow table rules in the core network function scheduling. Then the compatibility between parameters is checked, mainly including three aspects: first, whether the time-frequency resource allocation conflicts, checking whether the resource block allocation of different services overlaps; second, whether the priority configuration is reasonable, ensuring that high-priority services have sufficient resource guarantee; third, whether the core network rules match the access network configuration. Multiple detection points are set during the verification process, including resource conflict detection points, parameter range detection points, and service rule detection points. When parameter conflicts are found, the conflict type and conflict location are recorded, and corresponding adjustment suggestions are generated. For example, when there is overlap in time-frequency resource allocation, it is recommended to adjust the resource block allocation scheme; when the priority configuration is unreasonable, it is recommended to recalculate the priority weight. All verification results form a detailed verification report, including passed detection items, failed detection items, and corresponding adjustment suggestions.

[0044] The hierarchical configuration process is implemented in steps from the core network to the access network. At the core network level, traffic forwarding rules are first configured on the selected user plane function instances. The specific operations include: creating traffic classification rules to identify high-voltage transformer measurement service flows according to source IP address, destination IP address, protocol type, and port number; setting traffic policies to allocate corresponding quality of service levels for different types of measurement data; configuring forwarding paths to specify next-hop addresses and egress ports. At the transport network level, a priority-based queue scheduling mechanism is configured. The implementation steps include: creating 8 priority queues on each network node, corresponding to 8 different service priorities; setting different scheduling weights and cache management strategies for each queue; configuring queue mapping rules to map the priority tag of the service flow to the corresponding output queue. At the wireless access network level, the specific operation of allocating time-frequency resource blocks includes: reserving specified time-frequency resources in the resource manager of the base station; configuring the modulation and coding scheme and power control parameters of the resource block; establishing the mapping relationship between the service flow and the resource block. The configuration of each level is executed in a determined order, and each step is confirmed after execution to ensure the accuracy and integrity of the configuration.

[0045] The verification of the channel establishment effect is implemented by end-to-end signaling detection. The detection process is divided into three steps: first, a probe signal is sent, initiated from the access network device, passing through the transmission network, the core network, and finally reaching the monitoring center. The probe signal contains a timestamp sequence, and each network node records the receiving time and sending time when forwarding. Second, the performance data of each node is collected, including processing delay, queuing delay, and transmission delay. Finally, data analysis is performed to calculate key indicators such as end-to-end delay, delay jitter, and packet loss rate. The comparison of measured parameters and expected indicators uses a hierarchical evaluation method: the deviation of measured values from expected values is divided into three levels, with a deviation of less than 5% being excellent, 5%-10% being good, 10%-15% being qualified, and more than 15% being unqualified. The comparison includes delay indicators, bandwidth indicators, and reliability indicators. The delay indicator compares the difference between the actual transmission delay and the expected delay; the bandwidth indicator compares the compliance of the measured throughput with the expected bandwidth; and the reliability indicator compares the difference between the actual packet loss rate and the expected packet loss rate. According to the comparison results, a channel establishment quality evaluation report is generated, which details the compliance of each indicator.

[0046] The implementation of the fallback protection mechanism is achieved by a pre-device solution. When channel establishment failure or performance is detected, first determine the fault type and impact range. Fault types are divided into two categories: complete establishment failure and partial performance failure. For complete establishment failure, perform a complete backup channel establishment process: select the next best path from the candidate path set, regenerate the resource allocation scheme, and quickly execute the configuration instructions. For partial performance failure, take appropriate adjustment measures according to the specific failure indicators. When the delay is not up to standard, adjust the resource allocation scheme and increase the proportion of high-priority resource blocks; when the bandwidth is not up to standard, increase the number of allocated time-frequency resource blocks; when the reliability is not up to standard, enable data retransmission mechanism or path redundancy protection. Time limit is set during the backup channel establishment process, requiring the switching operation to be completed within 200 milliseconds. At the same time, channel switching records are established, detailing the fault information of the original channel, the selection basis of the backup channel, and the performance changes during the switching process, providing data support for subsequent optimization.

[0047] Multi-dimensional channel quality monitoring is implemented by deploying distributed collection points. Monitoring probes are set up at key nodes of the transmission channel, including access network base stations, transmission network switching nodes, and core network gateways. Monitoring indicators include latency jitter, packet loss rate, bandwidth utilization, and bit error rate. The calculation method of latency jitter is to continuously measure the transmission latency of 100 data packets and calculate the standard deviation of these latency values. The packet loss rate is calculated by using a sliding window method, with each 1000 data packets as a statistical window, and the proportion of lost data packets is calculated. Bandwidth utilization is monitored by comparing the actual data throughput with the theoretical maximum bandwidth. The bit error rate is calculated based on the cyclic redundancy check result, and the proportion of data frames with check errors in the total data frames is calculated. Monitoring data is collected at a period of 1 second and is stamped with accurate time stamps. All monitoring data is uploaded to the centralized processing unit in real time for unified storage and analysis. A data quality check mechanism is set up during the monitoring process to ensure the accuracy and integrity of the collected data.

[0048] Abnormality detection and localization are implemented by establishing a trend analysis model. This model is based on historical monitoring data and combines real-time data streams to identify abnormal fluctuations in channel quality. The specific implementation includes three steps: First, establish the normal fluctuation range of each indicator. By analyzing the monitoring data of the past 24 hours, the average value and standard deviation of each indicator in the same period are calculated, and the normal range is defined as three times the standard deviation of the average value. Second, set dynamic thresholds to dynamically adjust the abnormality judgment criteria according to network load and service characteristics. Finally, realize abnormal localization. When an indicator anomaly is detected, the specific section where the anomaly occurs is determined by analyzing the monitoring data of each node. The localization method uses a step-by-step elimination method, starting from the source node and sequentially checking the performance indicators of each network section until the section with performance mutation is found. The localization process takes into account the changes in multiple indicators to improve the accuracy of localization. All abnormal events are recorded in the abnormal log, including the occurrence time, abnormal type, impact range, and localization result.

[0049] The adaptive maintenance strategy library contains three main processing methods. The parameter adjustment scheme is aimed at slight performance degradation, including adjusting the transmission power, modifying the modulation and coding scheme, optimizing the buffer settings, etc. For example, when the delay jitter increases, the transmission power is appropriately increased; when the bit error rate rises, the modulation order is reduced. The path switching scheme is used to handle serious performance problems, and the implementation steps include: evaluating the availability of backup paths, calculating the switching cost, and performing path switching operations. The resource reallocation scheme solves the problem of insufficient resources or uneven load, and the specific operations include reallocating time-frequency resource blocks, adjusting queue scheduling weights, and optimizing traffic distribution strategies. Each processing scheme contains detailed implementation steps and expected effect evaluation. The selection of maintenance strategies is based on the type and severity of the anomaly, and the most suitable processing method is determined through a decision matrix. The input of the decision matrix includes factors such as anomaly type, impact range, and service priority, and the output is the recommended processing scheme and its priority ranking.

[0050] Preventive maintenance is implemented by establishing a performance prediction model. This model is based on historical channel quality data and combines real-time monitoring data to predict the trend of channel performance. The implementation process includes four steps: First, collect historical data, including long-term changes in various quality indicators and maintenance operation records. Second, analyze the performance degradation law, and find the correlation between indicators and change patterns through data mining methods. Then, establish a prediction model to predict the possible values of each indicator in the future period of time using time series analysis methods. Finally, develop a preventive maintenance plan and perform the corresponding maintenance operation in advance when the predicted performance is likely to decline. The specific operations of preventive maintenance include: adjusting resource allocation in advance, optimizing transmission parameters, and cleaning cache data. All preventive maintenance operations are recorded in the maintenance log, including maintenance time, maintenance content, and maintenance effect information, which is used to continuously optimize the maintenance strategy. The evaluation of maintenance effect is carried out by comparing the performance indicators before and after maintenance to ensure the effectiveness of the maintenance operation.

[0051] The working principle of the present application: through the state monitoring and data acquisition module, a multi-dimensional network state matrix reflecting the dynamic performance of the network is constructed in real time; the path calculation and analysis module performs resource availability analysis based on the matrix, outputs the candidate transmission path set and the corresponding wireless resource block allocation scheme; the resource scheduling and decision module calculates the path resource allocation factor and generates the optimal resource scheduling strategy combined with network load prediction; the control instruction generation module converts the strategy into network control instructions containing time-frequency resource reservation, transmission priority configuration, and core network scheduling rules; finally, the channel establishment and management module executes the instructions to establish a dedicated transmission channel with deterministic quality of service in the physical network, and implements real-time monitoring and dynamic maintenance, thereby ensuring the reliable and low-latency transmission of high-voltage mutual inductor measurement data in the 5G network.

[0052] The above has been described in detail one embodiment of the present application, but the content is only the preferred embodiment of the present application, cannot be considered for limiting the scope of the present application. Any equivalent changes and improvements made in the scope of the present application, should still belong to the scope of the present application.

Claims

1. A remote metering and monitoring system for high-voltage instrument transformers based on 5G network slicing, characterized in that, include: The status monitoring and data acquisition module is used to collect the performance parameters of each network node in the 5G network slice in real time and construct a multi-dimensional network status matrix including latency, packet loss rate and available bandwidth. The path calculation and analysis module is used to receive the multi-dimensional network state matrix, perform resource availability analysis, and output a set of candidate transmission paths and a wireless resource block allocation scheme for each path. The resource scheduling and decision-making module is used to calculate the path resource allocation factor based on the received candidate transmission path set and radio resource block allocation scheme, and generate the optimal resource scheduling strategy by combining network load prediction data. The control command generation module is used to convert the received optimal resource scheduling strategy into network control commands that include time-frequency resource reservation parameters, transmission priority configuration, and core network function scheduling rules. The channel establishment and management module is used to execute received network control commands, establish a dedicated transmission channel with a defined quality of service guarantee in the physical network, and perform real-time monitoring and dynamic maintenance of the dedicated transmission channel.

2. The high-voltage transformer remote metering and monitoring system based on 5G network slicing according to claim 1, characterized in that, The process of constructing the multidimensional network state matrix is ​​as follows: Within a preset time window, multiple rounds of probe detection are performed on each network node to collect end-to-end latency, packet loss rate, and available bandwidth data of the transmission path between nodes, forming the raw performance dataset. Based on the network topology, the raw performance data is organized into a spatiotemporal correlation matrix. The row dimension of the spatiotemporal correlation matrix corresponds to different detection time points, and the column dimension corresponds to the transmission links in the network. Each matrix element contains a triplet measurement of the transmission link's latency, packet loss rate, and available bandwidth at the corresponding time point. The spatiotemporal correlation matrix is ​​normalized using a sliding window to form an enhanced feature vector; The enhanced feature vectors are reorganized according to the hierarchical structure of the network topology to construct a multidimensional network state matrix with temporal continuity and spatial correlation.

3. The remote metering and monitoring system for high-voltage transformers based on 5G network slicing according to claim 1, characterized in that, The output candidate transmission path set specifically includes: Based on the multidimensional network state matrix, the latency fluctuation coefficient and packet loss risk index of each potential path are calculated. The latency fluctuation coefficient is obtained by calculating the ratio of the latency standard deviation to the average latency of the path in each time dimension, and the packet loss risk index is obtained by analyzing the distribution characteristics of historical packet loss rate and the probability of sudden packet loss. A comprehensive path quality evaluation function is constructed, which considers both the basic performance indicators and stability indicators of the path. The basic performance indicators include average latency and available bandwidth, while the stability indicators include latency fluctuation coefficient and packet loss risk index. The path quality score is formed by weighted combination. Based on the specific needs of remote metering services for high-voltage transformers, a dual threshold condition is set for path selection. The first threshold ensures that the basic performance of the path meets the minimum requirements for metering data transmission, while the second threshold ensures that the stability of the path meets the reliability standards required for metering services. From the set of paths that meet the dual threshold conditions, select the paths with the highest overall quality scores to form a candidate transmission path set.

4. The high-voltage transformer remote metering and monitoring system based on 5G network slicing according to claim 1, characterized in that, The calculation process for the wireless resource block allocation scheme is as follows: Analyze the load status and interference of each link in the candidate transmission path, and predict the resource requirements of each link in the future time period based on the historical bandwidth utilization and real-time load data in the multi-dimensional network state matrix. Based on the characteristics of high-voltage transformer metering data, wireless resource blocks are divided into two types: guaranteed resource blocks and flexible resource blocks. Guaranteed resource blocks are used to transmit critical synchronous metering data, while flexible resource blocks are used to transmit non-real-time status monitoring data. Based on the load prediction results and business demand characteristics of each path, the optimal resource block combination scheme is calculated for each candidate path.

5. The remote metering and monitoring system for high-voltage transformers based on 5G network slicing according to claim 1, characterized in that, The computation path resource allocation factor specifically includes: Based on the received wireless resource block allocation scheme, the resource allocation density characteristics of each path are extracted. The resource allocation density index is obtained by calculating the ratio of the number of resource blocks allocated per unit time to the total number of available resource blocks. By monitoring the deviation between the actual utilization efficiency and theoretical capacity of resource blocks, and combining the statistics of the idle time ratio of resource blocks, a resource utilization efficiency index is generated. By statistically analyzing the frequency and magnitude of adjustments to resource allocation schemes within a preset time period, the fluctuation coefficient of resource allocation schemes is calculated to form a resource stability index. The resource allocation density index, resource utilization efficiency index, and resource stability index are normalized, and the path resource allocation factor is calculated using a dynamic weight allocation method. The weight coefficients are dynamically adjusted according to the actual needs of high-voltage transformer metering services.

6. The remote metering and monitoring system for high-voltage transformers based on 5G network slicing according to claim 1, characterized in that, The process of generating the optimal resource scheduling strategy by combining network load prediction data specifically includes: Analyze the distribution patterns of network load in time and space, and predict the load change trends of each network region in a specific future time period; By conducting multi-dimensional correlation analysis between path resource allocation factors and network load prediction data, we can identify the degree of matching between resource demand and supply and determine the set of critical paths that need to be prioritized for scheduling. Simultaneously considering multiple optimization objectives such as resource utilization efficiency, path stability, and load balancing, an initial resource scheduling scheme is generated through a dynamic programming algorithm; The feasibility of the initial resource scheduling scheme is verified and its effectiveness is evaluated. Based on the evaluation results, heuristic adjustment rules are used to optimize and iterate the scheme, and finally the optimal resource scheduling strategy that meets the service quality requirements of high voltage transformer metering services is generated.

7. The remote metering and monitoring system for high-voltage transformers based on 5G network slicing according to claim 1, characterized in that, The process of generating the network control commands is as follows: Based on the resource allocation requirements in the optimal resource scheduling strategy, a resource block combination coding technique is used to generate time-frequency resource reservation parameters. By combining and coding adjacent resource blocks according to service demand characteristics, a resource unit group with specific transmission characteristics is formed. Based on the different service types and urgency of the high-voltage transformer metering data, a transmission priority configuration matrix is ​​constructed. The transmission priority configuration matrix comprehensively considers data real-time requirements, service importance and network status factors, and dynamically allocates differentiated transmission priorities to different types of data streams. For core network function scheduling, a service-aware user plane function selection mechanism is designed to select the optimal core network user plane function instance and configure the corresponding traffic forwarding rules based on the transmission path characteristics and processing requirements of metering data. The generated time-frequency resource reservation parameters, transmission priority configurations, and core network function scheduling rules are encapsulated into instructions to form network control instructions that can be directly issued to network devices.

8. The high-voltage transformer remote metering and monitoring system based on 5G network slicing according to claim 1, characterized in that, Establishing a dedicated transmission channel with guaranteed quality of service in the physical network specifically includes: The received network control commands are configured and parameters are verified. The compatibility between command parameters is detected by simulating the execution environment, potential execution conflicts are identified, and corresponding parameter adjustment schemes are generated. Configure user plane function rules at the core network level, set priority queues at the transport network level, and allocate time-frequency resource blocks at the radio access network level to ensure that the configurations at each network level work together effectively. The actual establishment effect of the dedicated transmission channel is verified by end-to-end signaling detection, and the degree of conformity between the measured channel parameters and the expected service quality indicators is compared. Implement fallback protection during the channel establishment process. When channel establishment failure or performance failure is detected, the preset backup channel establishment scheme is automatically triggered to ensure the continuity of high voltage transformer metering service transmission.

9. The remote metering and monitoring system for high-voltage transformers based on 5G network slicing according to claim 1, characterized in that, The real-time monitoring and dynamic maintenance of the dedicated transmission channel specifically includes: Construct a multi-dimensional channel quality monitoring system to simultaneously collect indicators such as latency jitter, packet loss rate, bandwidth utilization, and bit error rate of the transmission channel; Analyze the changing trends of channel quality indicators, and when abnormal fluctuations in the indicators are detected, automatically locate the network segment where the anomaly occurred and the cause; Design an adaptive maintenance strategy library to pre-configure corresponding maintenance schemes for different types of channel anomalies, including parameter adjustment, path switching and resource reallocation. Based on historical channel quality data, proactive preventative maintenance operations are performed before channel performance degrades.

Images