5G message transmission anomaly intelligent detection method and system

Abstract

The application relates to the field of new-generation mobile communication access network resource scheduling, and discloses a 5G message transmission anomaly intelligent detection method and system, which comprises the following steps: acquiring a media access control layer downlink control information format and a modulation coding strategy level; monitoring a switching action of the downlink control information format from a non-backoff format to a backoff format, determining a scheduling response delay feature by associating a variation amount of the modulation coding strategy level in a transmission time interval, and calculating a scheduling degradation index; when a channel quality indicator drop amount is in a non-fading fluctuation range, if the scheduling degradation index reaches an anomaly judgment threshold value, a transmission anomaly instruction is generated, and a physical resource block reservation queue is suspended; the application eliminates physical lag generated by layer-by-layer analysis of a protocol stack, predicts a link deterioration trend within a scheduling cycle order, cuts off a path in which a scheduler allocates invalid resources for a failed bearer, and improves spectrum utilization.

Description

Technical Field

[0001] This invention relates to a method and system for intelligent detection of 5G message transmission anomalies, belonging to the field of next-generation mobile communication access network resource scheduling technology. Background Technology

[0002] In the current next-generation mobile communication network architecture, reliability monitoring for messaging services is typically deployed on the core network side or application server side. The service status is determined by collecting transport layer acknowledgment messages or monitoring the round-trip latency of end-to-end heartbeat packets. This monitoring method, triggered from the bottom up based on the application layer protocol stack, is well-suited for handling routine low-frequency interactive services. However, when facing 5G message traffic with high-frequency transient concurrency characteristics, this cross-layer monitoring logic exhibits physical response lag due to the layer-by-layer unpacking of the protocol stack. When the radio access network encounters sudden congestion or scheduling conflicts, the latency deviation perceived by the service layer is usually on the order of hundreds of milliseconds. During this period, the base station media access control layer scheduler continues to reserve physical resource blocks for radio bearers that have already experienced transmission obstruction, based on outdated channel state information.

[0003] The asynchronous nature of this scheduling decision-making with the air interface physical state prevents base stations from taking intervention measures in the early stages of abnormal evolution. This not only leads to the invalidation of resource blocks but also easily induces signaling plane overload on the access network side due to frequent retransmission requests from terminals. Existing technologies not only have latency bottlenecks at the cross-layer interaction level of the protocol stack but also have limitations in their single-dimensional control strategies for air interface congestion. For example, Chinese invention patent CN113099483B discloses a cell congestion handling method, terminal, and network-side equipment. It guides the terminal to reselect or handover cells by issuing congestion indication information from the network side. The premise for the technical solution to work is that the base station identifies cell congestion at the macro level, which is a passive avoidance mechanism based on known states. Faced with millisecond-level bursts of 5G message traffic, relying on base stations to actively notify external triggering logic makes it difficult to capture instantaneous link degradation caused by micro-scheduling conflicts. Furthermore, the processing path involves cross-cell reselection actions, which lacks specificity for maintaining service flow continuity and fine-grained resource utilization recovery, and cannot achieve real-time capture and inversion of anomalies at the scheduling command level. If we try to improve sensitivity by shortening the heartbeat detection cycle, it will introduce additional air interface signaling overhead and increase terminal processing power consumption. If we simply adopt a strategy of increasing resource allocation margin, it will cause a decrease in spectrum utilization. This inherent constraint between performance improvement and engineering cost stems from the lack of in-depth utilization of the anomaly characteristics of the underlying native scheduling commands of the radio access network in existing technologies.

[0004] Therefore, the technical problem to be solved by this invention is how to achieve a closed-loop determination of transmission anomalies without parsing the service payload, based on the back-off characteristics of the native downlink control information format of the base station media access control layer and the trend of resource allocation deviation. Summary of the Invention

[0005] To address the problems in the background technology, the technical solution of the present invention is as follows: A 5G message transmission anomaly intelligent detection method, comprising the following steps: Step 101: Obtain the downlink resource allocation characteristics of the radio access network media access control layer. The downlink resource allocation characteristics include the downlink control information format and the modulation and coding scheme level. Step 102: Monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, and determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and determine the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. Step 103: Combine the format degradation weights and scheduling response delay characteristics corresponding to the handover action to calculate the scheduling degradation index that characterizes the air interface resource congestion state. Step 104: Obtain the channel quality indication fed back by the terminal device, and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range; Step 105: Under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, if the scheduling degradation index reaches the abnormal judgment threshold dynamically determined according to the service priority identifier of 5G message, an air interface transmission abnormal command for the target logical channel is output, and the radio access network media access control layer is triggered to suspend the physical resource block reservation queue corresponding to the target logical channel.

[0006] Preferably, before determining that a 5G message transmission anomaly has occurred, the method further includes the following steps: Step 201, obtaining the scheduling request collision density and the negative response density of the hybrid automatic repeat request of the terminal device; Step 202, verifying the scheduling request collision density and the negative response density by correlation to determine the logical link blocking state of the terminal device in the physical resource block allocation process; Step 203, generating a correction coefficient based on the logical link blocking state, and using the correction coefficient to correct the scheduling degradation index, so that Step 105 executes the determination logic based on the corrected scheduling degradation index.

[0007] Preferably, after outputting the air interface transmission anomaly command for the target logical channel, the method further includes the following steps: Step 301, extracting the service priority identifier corresponding to the 5G message; Step 302, allocating dedicated scheduling request resources that are in an orthogonal state and do not generate spectrum overlap to the terminal device on the base station side based on the service priority identifier.

[0008] Preferably, in step 102, the step change of the modulation and coding strategy level within adjacent transmission time intervals determines the scheduling response delay characteristics of the downlink, including the following sub-steps: Step 1021, count the total number of continuous transmission time intervals in which the modulation and coding strategy level is within a preset low-order value range; Step 1022, when the total number of continuous transmission time intervals exceeds a preset time threshold and the slope of the first block error rate of the corresponding downlink physical shared channel reaches a preset change rate limit, determine that the downlink has entered the scheduling response delay state.

[0009] Preferably, determining whether the drop in the channel quality indicator is within a preset non-fading fluctuation range in step 104 includes the following sub-steps: Step 1041, synchronously acquiring the reference signal received power and the signal-to-interference-plus-noise ratio of the terminal device; Step 1042, when the rate of change of the reference signal received power is below a first preset threshold and the variance of the signal-to-interference-plus-noise ratio is below a second preset threshold, determining that the drop in the channel quality indicator is within a non-fading fluctuation range.

[0010] Preferably, after allocating dedicated scheduling request resources to the terminal device, the method further includes the following steps: Step 601, monitoring the scheduling request reporting frequency of the terminal device on the dedicated scheduling request resources; Step 602, when the scheduling request reporting frequency is always 0 within a preset observation period, sending a radio access bearer release request to the core network user plane function network element to reclaim the physical resources corresponding to the physical resource block reserved queue.

[0011] Preferably, step 105, which triggers the radio access network media access control layer to suspend the physical resource block reservation queue, includes the following sub-steps: step 1051, locating the hybrid automatic repeat request process identifier corresponding to the target logical channel; step 1052, clearing the soft repeat buffer corresponding to the hybrid automatic repeat request process identifier, and restricting the downlink resource allocation authority of the terminal device in the subsequent 3 scheduling cycles.

[0012] Preferably, the method is applied to the distributed unit of a 5G base station to determine the abnormal state of air interface transmission based on the abnormal quantization characteristics of the radio access network control plane signaling without parsing the 5G message application layer payload carrier.

[0013] A 5G message transmission anomaly intelligent detection system, which implements a 5G message transmission anomaly intelligent detection method, includes: The feature acquisition module is used to acquire the downlink resource allocation features of the media access control layer of the radio access network. The downlink resource allocation features include the downlink control information format and the modulation and coding scheme level. The delay detection module, connected to the feature acquisition module, is used to monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and associate the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. The degradation assessment module, connected to the latency detection module, is used to fuse the format degradation weights corresponding to the handover action with the scheduling response latency characteristics to calculate the scheduling degradation index that characterizes the air interface resource congestion state. The channel monitoring module is used to acquire the channel quality indication fed back by the terminal device and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range. The anomaly determination module, connected to the degradation assessment module and the channel monitoring module respectively, is used to output an air interface transmission anomaly command for the target logical channel if the scheduling degradation index reaches the anomaly determination threshold dynamically determined according to the service priority identifier of the 5G message, under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, and triggers the radio access network media access control layer to suspend the physical resource block reservation queue corresponding to the target logical channel.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. In the intelligent detection of 5G message transmission anomalies, the technical features of the downlink control information scheduling log within a continuous time window are extracted. By calculating the product of the modulation and coding strategy level step difference and the weights of different downlink control information formats, a mapping relationship between scheduling degradation feature values ​​and air interface transmission status is established. This allows the anomaly identification process to directly act on the media access control layer of the radio access network. The detection contact point is shifted from the passive response after the application layer confirms the loss of the message to the immediate capture when the scheduling command generates a micro offset. By eliminating the inherent delay caused by the bottom-up unpacking of the protocol stack, the prediction of the transmission link deterioration trend is achieved within the order of the underlying scheduling cycle. This ensures that the access network scheduling unit can perceive and handle potential congestion risks earlier than the service layer.

[0015] 2. Combining the dynamic adjustment trend of modulation and coding strategies with the physical degradation characteristics of downlink control information formats, an anomaly inversion logic based on media access control layer scheduling behavior is constructed. This logic utilizes the inherent defensive action of the base station scheduler to forcibly reduce the scheduling format level when facing link congestion. By fusing and analyzing the deviation of multi-dimensional scheduling parameters, it achieves qualitative judgment of the transmission status of high-frequency concurrent messages without parsing service layer payload data. This approach not only improves the system's sensitivity to message content, but also avoids the lag in physical resource reservation caused by the time consumption of deep packet parsing at the service layer by directly extracting underlying signaling anomalies.

[0016] 3. By introducing the drop amplitude of channel quality indicator as a precondition for anomaly detection, and combining it with the scheduling request collision density and retransmission request negative response density of terminal equipment, a multi-criteria overlapping verification judgment system is formed. This system can effectively distinguish from the normal performance fluctuations caused by deep fading of wireless channels and the substantial transmission anomalies caused by resource allocation conflicts from a physical mechanism perspective. This precise removal of physical noise ensures that the base station scheduling unit only takes suspension actions for logical link blockage, avoiding the overall system throughput reduction caused by false triggering, and enhancing the scheduling accuracy of the access network in complex air interface environments. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the closed-loop implementation of the underlying detection method for 5G message transmission anomalies in this invention. Figure 2 This is a logical architecture diagram of the transmission anomaly determination method for multi-dimensional parameter collaborative verification according to the present invention.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0020] A method for intelligent detection of 5G message transmission anomalies includes the following steps: Step 101: Obtain the downlink resource allocation characteristics of the radio access network media access control layer. The downlink resource allocation characteristics include the downlink control information format and the modulation and coding scheme level. Step 102: Monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, and determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and determine the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. Step 103: Combine the format degradation weights and scheduling response delay characteristics corresponding to the handover action to calculate the scheduling degradation index that characterizes the air interface resource congestion state. Step 104: Obtain the channel quality indication fed back by the terminal device, and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range; Step 105: Under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, if the scheduling degradation index reaches the abnormal judgment threshold dynamically determined according to the service priority identifier of 5G message, an air interface transmission abnormal command for the target logical channel is output, and the radio access network media access control layer is triggered to suspend the physical resource block reservation queue corresponding to the target logical channel.

[0021] Preferably, before determining that a 5G message transmission anomaly has occurred, the method further includes the following steps: Step 201, obtaining the scheduling request collision density and the negative response density of the hybrid automatic repeat request of the terminal device; Step 202, verifying the scheduling request collision density and the negative response density by correlation to determine the logical link blocking state of the terminal device in the physical resource block allocation process; Step 203, generating a correction coefficient based on the logical link blocking state, and using the correction coefficient to correct the scheduling degradation index, so that Step 105 executes the determination logic based on the corrected scheduling degradation index.

[0022] Preferably, after outputting the air interface transmission anomaly command for the target logical channel, the method further includes the following steps: Step 301, extracting the service priority identifier corresponding to the 5G message; Step 302, allocating dedicated scheduling request resources that are in an orthogonal state and do not generate spectrum overlap to the terminal device on the base station side based on the service priority identifier.

[0023] Preferably, in step 102, the step change of the modulation and coding strategy level within adjacent transmission time intervals determines the scheduling response delay characteristics of the downlink, including the following sub-steps: Step 1021, count the total number of continuous transmission time intervals in which the modulation and coding strategy level is within a preset low-order value range; Step 1022, when the total number of continuous transmission time intervals exceeds a preset time threshold and the slope of the first block error rate of the corresponding downlink physical shared channel reaches a preset change rate limit, determine that the downlink has entered the scheduling response delay state.

[0024] Preferably, determining whether the drop in the channel quality indicator is within a preset non-fading fluctuation range in step 104 includes the following sub-steps: Step 1041, synchronously acquiring the reference signal received power and the signal-to-interference-plus-noise ratio of the terminal device; Step 1042, when the rate of change of the reference signal received power is below a first preset threshold and the variance of the signal-to-interference-plus-noise ratio is below a second preset threshold, determining that the drop in the channel quality indicator is within a non-fading fluctuation range.

[0025] Preferably, after allocating dedicated scheduling request resources to the terminal device, the method further includes the following steps: Step 601, monitoring the scheduling request reporting frequency of the terminal device on the dedicated scheduling request resources; Step 602, when the scheduling request reporting frequency is always 0 within a preset observation period, sending a radio access bearer release request to the core network user plane function network element to reclaim the physical resources corresponding to the physical resource block reserved queue.

[0026] Preferably, step 105, which triggers the radio access network media access control layer to suspend the physical resource block reservation queue, includes the following sub-steps: step 1051, locating the hybrid automatic repeat request process identifier corresponding to the target logical channel; step 1052, clearing the soft repeat buffer corresponding to the hybrid automatic repeat request process identifier, and restricting the downlink resource allocation authority of the terminal device in the subsequent 3 scheduling cycles.

[0027] Preferably, the method is applied to the distributed unit of a 5G base station to determine the abnormal state of air interface transmission based on the abnormal quantization characteristics of the radio access network control plane signaling without parsing the 5G message application layer payload carrier.

[0028] A 5G message transmission anomaly intelligent detection system includes: The feature acquisition module is used to acquire the downlink resource allocation features of the media access control layer of the radio access network. The downlink resource allocation features include the downlink control information format and the modulation and coding scheme level. The delay detection module, connected to the feature acquisition module, is used to monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and associate the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. The degradation assessment module, connected to the latency detection module, is used to fuse the format degradation weights corresponding to the handover action with the scheduling response latency characteristics to calculate the scheduling degradation index that characterizes the air interface resource congestion state. The channel monitoring module is used to acquire the channel quality indication fed back by the terminal device and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range. The anomaly determination module, connected to the degradation assessment module and the channel monitoring module respectively, is used to output an air interface transmission anomaly command for the target logical channel if the scheduling degradation index reaches the anomaly determination threshold dynamically determined according to the service priority identifier of the 5G message, under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, and triggers the radio access network media access control layer to suspend the physical resource block reservation queue corresponding to the target logical channel.

[0029] Example 1: In this example, under the condition of deploying an ultra-reliable low-latency communication network slice and facing a sudden high-frequency 5G message concurrency in a cellular environment, the physical downlink control channel (PDCCH) on the access network side experiences implicit packet loss due to instantaneous overload of control signaling. The media access control layer scheduling module in the base station's distributed unit continuously acquires the 5G message service flow to be detected, and extracts the physical downlink control channel scheduling log within a continuous time window based on the logical channel identifier mapped to the service flow. It obtains the actual modulation and coding strategy level (MCS level) and the actual downlink control information format (DCI format) within each scheduling period. The media access control layer scheduling module monitors in real time within the continuous time window whether there is a switch action from a non-backoff format to a backoff format in the actual DCI format. It determines that the actual DCI format has switched from DCI format 1_1 to DCI format 1_0. At this time, it calculates the step difference between the actual MCS level and the benchmark level corresponding to the target service quality identifier, and integrates the format degradation weight and the step difference to calculate the scheduling degradation index characterizing the air interface resource congestion state. Dispatch degradation index The following calculation relationship must be satisfied: ,in, To regulate the degradation index, For the first The format degradation weights corresponding to the actual DCI format within each scheduling cycle. For the first Step difference value within a scheduling cycle The total number of scheduling cycles contained within a continuous time window, when the scheduling degradation index... When the preset anomaly detection threshold is exceeded, the system synchronously acquires the Channel Quality Indicator (CQI) fed back by the terminal device and determines whether the CQI dropout is within the preset non-fading fluctuation range. If the CQI dropout is within the non-fading fluctuation range and the Hybrid Automatic Repeat Request (HARQ) negative response density exceeds the preset collision trigger line, the system generates an air interface transmission anomaly command for the logical channel identifier. This triggers the Media Access Control (MAC) layer to suspend the physical resource block reservation queue corresponding to the logical channel identifier. Based on the service priority identifier of the 5G message, the MAC layer allocates orthogonal and non-overlapping spectrum resources to the terminal device at the base station side. This embodiment utilizes dedicated scheduling requests for resources. By acquiring the underlying scheduling degradation facts—DCI format backoff and MCS level step drop—in downlink scheduling signaling, the anomaly detection node is moved from traditional application layer cross-layer detection to the access network scheduling cycle level. Within the scheduling cycle, the media access control layer identifies the stability degradation actions taken by the scheduler to maintain link gain. Combined with CQI fluctuation characteristics, it eliminates pure physical channel fading interference, achieving precise characterization of 5G message transmission anomalies within 10ms. This cuts off the path for the scheduler to continuously reserve radio resource blocks for failed bearers, reducing the air interface signaling plane load and improving the effective utilization of frequency resources.

[0030] In macro base station deployments with narrowband IoT access services, to determine scheduling response delay characteristics, the media access control layer scheduling module extracts modulation and coding strategy level change curves over 10 consecutive transmission time intervals, calculates the sum of level differences between adjacent intervals, and simultaneously obtains the first block error rate of the downlink physical shared channel (PDSCH). Based on channel capacity evolution patterns, downlink packet drop caused by air interface physical resource congestion manifests as a step increase in error rate over time. The media access control layer calls the base station's internal performance monitoring unit to extract the first block error rate log from the current cell's non-congested historical periods and calculates the standard sampling window. The system filters out abnormal communication records with resource allocation conflict labels by using the slope of the reference change within the port. It calculates the mean difference of outlier error rate within the corresponding interval and determines the lower quartile of the mean difference of outlier error rate as the preset change rate limit. Under this condition, the preset change rate limit is calibrated to 0.12 per transmission time interval by actual measurement. When the sum of the level difference values ​​continues to show a negative growth trend and the rising slope of the first transmission block error rate exceeds the preset change rate limit, the system determines that there is a time axis mismatch between the active speed reduction action performed by the scheduler to maintain link stability and the rate of wireless channel deterioration, and records the value of the time axis mismatch as the scheduling response delay characteristic.

[0031] Example 2: This experiment verifies the response speed and resource recovery efficiency of the access network media access control layer when facing air interface resource congestion caused by high-frequency concurrent 5G messages, based on the downlink scheduling signaling characteristics to invert transmission anomalies. The experimental platform is built in an access network system-level simulation environment conforming to the O-RAN architecture specification, including a distributed unit (DU) simulation module and a physical layer simulation channel. The data acquisition terminal is integrated into the media access control layer scheduler interface of the DU simulation module, and its sampling frequency is set to 1000Hz to ensure that the modulation and coding strategy level and downlink control information format are captured in each 1ms transmission time interval. To simulate real cellular conditions, additive white Gaussian noise with a signal-to-noise ratio of 20dB is injected into the physical layer simulation channel during the experiment, and multipath Rayleigh fading disturbance is superimposed to simulate the physical layer characteristics of terminal equipment in a mobile environment. In the experimental parameter setting phase, the total number of consecutive time windows is... The value of is controlled by the air interface signaling jitter frequency and the anomaly detection delay target. Its technical trade-off lies in balancing the accuracy of the detection results with the response speed. When the detected signal jitter spectrum bandwidth is wide, to avoid false triggering caused by transient physical layer deep fading, the total number of continuous time windows is... It tends towards the upper limit of the value range; in the 5G message burst scenario simulated in this experiment, in order to meet the resource reclamation requirement within 10ms, the total number of continuous time windows is... The value is set to 20, meaning 20 consecutive scheduling cycles are observed. Simultaneously, the format degradation weights are... The setting is determined based on the aggregation level decoding threshold of different DCI formats under the current physical layer configuration. The first weight based on DCI format 1_1 is 1.0, and the second weight based on DCI format 1_0 is 2.5. This difference reflects the spectral efficiency cost lost by the base station scheduler in order to maintain the control channel connection.

[0032] The experimental group using the method of this invention detected a switch in the actual DCI format of the Physical Downlink Control Channel (PDCCH) from 1_1 to 1_0 during the fourth transmission time interval. Simultaneously, the actual MCS level dropped from the baseline level of 22 to 8. Observational data showed that the step difference occurred between the fourth and twelfth transmission time intervals. The index remained consistently around 14. Calculations showed that at 9.2ms, the scheduling degradation index... Reaching 28.64, exceeding the preset anomaly threshold of 25.0, the synchronously acquired Channel Quality Indicator (CQI) drop is 2.1 dB, within the preset non-fading fluctuation range of 1.5 dB to 3.0 dB. Furthermore, the negative acknowledgment density of the Hybrid Automatic Repeat Request (HARQ) remains constant at 1.0 for five consecutive time slots. In response to these conditions, the system issues an air interface transmission anomaly command, suspending the physical resource block reservation queue. To demonstrate the synergistic effect between technical features, a partially missing control group is set up. This group removes the format degradation weight when calculating the scheduling degradation index. The data shows that the index value calculated by control group 1 was only 15.42 after the anomaly occurred, which failed to trigger the threshold judgment. As a result, the system continued to allocate wireless resources to the failed logical channel for the next 150ms, and the spectrum utilization rate decreased by 42.5%. This confirms the deep logical coupling between the DCI format switching action and the MCS step change. Meanwhile, control group 2 adopted the application layer deep packet detection method, and its anomaly detection time occurred at 246.8ms after the 5G message was initiated. Compared with the 9.2ms of the experimental group of this invention, the response delay increased by more than 20 times.

[0033] Validation of the key parameter boundaries shows that when the total number of consecutive time windows... When set to 100, the output time of the exception instruction is delayed to 48.5ms, resulting in a 180% increase in the invalid occupancy of physical resource blocks; while when When set to 3, due to Rayleigh fading noise interference, the system generated an alarm deviation of 12.8% under normal transmission conditions. The experimental data showed the correlation of the problem intensity gradient. As the 5G message concurrency density increased from 100 messages per second to 5000 messages per second, the scheduling degradation index... The rising slope of the concurrency density exhibits a non-linear positive correlation. This experiment verifies that the scheme of the present invention establishes a deterministic causal chain from signaling format degradation to resource reclamation by capturing the physical facts of scheduling degradation in the media access control layer of the access network. Under the premise of excluding random fluctuations in the physical channel, it achieves accurate determinism of transmission anomalies and solves the time delay problem of cross-layer detection mechanism when dealing with high-frequency message concurrency.

[0034] Example 3: This example combines Figures 1 to 2 This document describes a method and system for intelligent detection of 5G message transmission anomalies, such as... Figure 1As shown, in the downlink resource allocation characteristic acquisition stage, radio access network media access control layer characteristics, including downlink control information (DCI format) and modulation and coding strategy (MCS level), are extracted. In the downlink scheduling response delay characteristic determination stage, by real-time monitoring of whether there is a handover action from DCI format 1_1 to DCI format 1_0, and by associating the step change of the MCS level within adjacent transmission time intervals (TTI), a scheduling degradation index characterizing the air interface resource congestion state is calculated. This index calculation integrates the format degradation weight corresponding to the handover action and the scheduling response delay. Simultaneously, in another parallel logic branch, the channel quality indicator (CQI) fed back by the terminal device is obtained, and it is determined whether the drop in CQI is within the preset non-fading fluctuation range. After entering the logic node that determines the abnormal triggering conditions, if the two criteria of CQI drop being within the non-fading fluctuation range and scheduling degradation index reaching the abnormal judgment threshold dynamically determined by the service priority identifier are met, an air interface transmission abnormal command for the target logical channel is output, thereby triggering the radio access network media access control layer to suspend the physical resource block (PRB) reservation queue corresponding to the target logical channel.

[0035] like Figure 2 As shown, the reference signal received power (RSRP) and the signal-to-interference-plus-noise ratio (SINR) jointly point to and are used to determine whether the CQI drop is within the non-fading fluctuation range. At the same time, the downlink control information format and modulation and coding strategy level jointly point to and are used to determine whether the scheduling degradation index has reached the anomaly judgment threshold. In addition, the negative response density of the hybrid automatic repeat request and the scheduling request collision density jointly point to and are used to determine the logical link blocking state. The intermediate conclusions and state characteristics of the above judgments converge into the judgment system of multi-criteria overlapping verification. This system performs logical overlapping verification and finally outputs the air interface transmission anomaly command.

[0036] Example 4: In an access network deployment scenario involving large-scale machine-type communication services and facing non-stationary random interference, the system ensures the scheduling degradation index... To ensure the accuracy of the judgment, a format degradation weight is established through the media access control layer scheduling module. The mapping relationship with the demodulation threshold gain is determined by measuring the difference in signal-to-noise ratio requirements between DCI format 1_1 and DCI format 1_0 under the same bit error rate target in a physical layer simulation environment. When the system detects that the physical layer uses 1 / 2 code rate polar code encoding, the demodulation gain advantage of DCI format 1_0 compared to DCI format 1_1 is 4.0dB. At this point, the format degradation weight is adjusted. Defined as the linear mapping function value of demodulation gain advantage, in the specific deployment of this embodiment, the first weight of DCI format 1_1 is set to 1.0, and the second weight of DCI format 1_0 is locked to 2.5 according to the demodulation stability improvement factor. To determine the scheduling degradation index To determine the threshold for anomaly detection, the access network's radio resource management unit obtains the historical scheduling logs of the target cell over the past 24 hours and extracts the scheduling degradation index distribution sequence of the logical channel under normal transmission conditions. The system uses a Gaussian mixture model to fit the probability density of the distribution sequence and selects the exponential value corresponding to the cumulative distribution function reaching 0.95 as the benchmark value for anomaly detection. Specifically, regarding the internal construction mechanism of the Gaussian mixture model, the system fixes the total number of Gaussian distribution components in the model to three to accurately match the potential distribution characteristics of the communication network under three different states: light load, normal time period, and sudden congestion. In the initialization phase of the fitting algorithm, the system uses K-means clustering to determine the central origin corresponding to the three components and presets the variance within each scattered sample cluster as the initial covariance array. In the subsequent maximum likelihood estimation stage, the convergence of the continuous logarithmic probability deviation to a rate of change of one-thousandth is used as the calculation jump boundary to maintain the data in the derivation process of the model parameter matrix. To ensure convergence stability, in the application scenario of this embodiment, the calculated anomaly judgment threshold baseline value is 22.8. The system performs dynamic compensation based on the current physical resource block utilization rate of the cell. In the congestion state where the utilization rate exceeds 80%, the anomaly judgment threshold is adjusted upward by 10% to obtain the final anomaly judgment threshold of 25.0. After obtaining the judgment line that has been initially corrected in combination with the physical resource block utilization rate, the media access control layer scheduler extracts the service priority identifier quantization value corresponding to the target 5G message to carry out end-point dynamic calibration. The scheduling module multiplies the service priority identifier quantization value with the preset unit reduction scalar to obtain the offset deduction amount. The offset deduction amount is subtracted from the aforementioned initially corrected judgment line to output the final anomaly judgment threshold based on priority adaptation. The higher the priority quantization value of the message type service, the greater the judgment downward space is generated, ensuring that high-priority services can trigger the intervention mechanism in advance to cut off invalid scheduling paths in the early stage of resource congestion deterioration.

[0037] In the underlying execution path of the anomaly detection logic, the media access control layer scheduler connects to the physical layer resource allocation register through its internal in-situ control interface, responding to the scheduling degradation index. If the signal strength exceeds 25.0 and the Channel Quality Indicator (CQI) drop is within the non-fading fluctuation range of 1.5dB to 3.0dB, the scheduling module generates a 16-bit bitmask instruction. The high-order byte of this instruction carries the logical channel identifier, and the low-order byte carries the suspension trigger operator. The bitmask instruction is directly written into the resource reservation table of the physical layer scheduling controller, causing the physical layer scheduling controller to skip the physical resource block search space for this logical channel identifier in the following 5 subframes. To overcome the logical isolation of external instructions by the closed black-box module of the underlying physical layer, the media access control layer calls the standardized fronthaul management plane data channel to perform cross-layer penetration. The system encapsulates the generated bitmask instruction into a direct-write message packet structure specific to the baseband vendor extension, and transmits the message stream directly to the direct memory access control of the baseband digital signal processing cluster using a bypass delivery mechanism. The device forcibly overwrites the underlying available status bits in the target channel resource mapping register. This physical-level direct hardware path ensures that control plane signaling avoids the layer-by-layer unpacking loss of firmware. It implements the link continuity of the reserved queue suspension action from the micro-physical level. The negative acknowledgment density detection unit of the Hybrid Automatic Repeat Request (HARQ) monitors the number of negative acknowledgments in a continuous scheduling cycle. When the number exceeds 8 in 10 consecutive transmission block allocations, it triggers the collision trigger line and locks the air interface transmission anomaly command. This embodiment establishes a closed-loop mapping from underlying physical characteristics to system execution mechanism by disclosing the format degradation weight calibration, dynamic threshold generation, and bitmask execution path. This enables the technical solution to have repeatable calibration basis and definite execution boundary when facing non-stationary interference environment, and improves the resource self-healing capability of the access network when processing high-frequency concurrent messages.

[0038] Example 5: In a mobile communication test network deployment scenario including distributed units and terminal devices, the system performs initial calibration of the media access control layer scheduling interface, aligns the sampling clock of the physical downlink control channel log recording module with the transmission time interval boundary of the media access control layer, and sets a synchronization pulse. For 0.1ms, under controlled conditions with a reference signal received power of -90dBm, the modulation and coding strategy level distribution of the logic channel in a non-congested state was measured. A reference bias sequence reflecting the inherent measurement jitter of the hardware was established. Based on the independent physical laws of multipath effect and path loss in the radio wave propagation fading model, the system executed a pre-calibration procedure to separate physical signal distortion and logic resource congestion. The test platform controlled the base station RF unit to send a probe reference signal to a stationary, unobstructed reference terminal. The measured values ​​of the reference signal received power within 100 consecutive measurement cycles were extracted. The average absolute value of the difference between the measured values ​​at adjacent time nodes was calculated and a tolerance coefficient of 1.2 was added. The final calculation result was solidified as the first preset threshold value. This specific coefficient value was locked based on the physical measurement conclusion of the microwave anechoic chamber of the hardware front-end background thermal noise disturbance width. A full-band broadband clutter injection was performed using a Gaussian additive noise source. Multiple tests showed that the reference signal received power was limited by the temperature drift of the RF device. The physical limit jitter presented by the control is 1.2 times the aforementioned differential mean. If this compensation step is lacking and the scale-free mean is directly applied, the state assessment module will frequently generate false alarms due to over-preparation when there is no severe channel fading. Simultaneously extract the signal-to-interference-plus-noise ratio data array within the same measurement window, calculate the deviation from the mathematical expectation second-order central moment and establish it as the second preset threshold value. Through hardware-level calibration, the first preset threshold value is locked at 1.5dB / ms and the second preset threshold value is locked at 2.8 in this environment. After completing the basic parameter locking, the downlink signal-to-interference-plus-noise ratio is adjusted in steps in the physical layer simulation unit to measure the fluctuation envelope of the channel quality indicator under additive white Gaussian noise interference. One standard deviation of the fluctuation envelope is defined as the lower limit boundary of the non-fading fluctuation range. Calculate the probability statistics of signaling decoding failure within 10 consecutive scheduling cycles and determine the initial value of the collision trigger line used to determine air interface resource conflicts.

[0039] When the system faces a new air interface configuration with a subcarrier spacing of 30kHz, the media access control layer scheduling module initiates format degradation weighting. The pre-filling process iterates through the demodulation success rate curves of downlink control information at aggregation levels 1 to 8, calculates the transmit power difference between DCI format 1_1 and DCI format 1_0 at a target block error rate of 1%, and stores this difference as a dimensionless mapping coefficient in the format degradation weight and writes it to the local configuration register on the base station side. This allows the scheduler to call the corresponding weight based on the actual DCI format captured in real time. After parameter calibration and data loading, the distributed unit enters the service bearer monitoring state and uses the locked benchmark parameters to monitor the logical channel identifier of the 5G message service flow image. When the scheduling degradation index is detected to exceed the dynamically corrected judgment threshold, the writing path of the bitmask instruction to the physical layer resource allocation register is maintained, and the system is in a ready state to intercept air interface transmission anomalies.

[0040] Example 6: In a deployment scenario where distributed units carry ultra-reliable low-latency communication service flows and face preemption of air interface control channel resources by enhanced mobile broadband services, the system, in order to adapt the transmission anomaly detection logic to the physical characteristics of different network slices, executes a modulation and coding strategy baseline level. The pre-calibration procedure involves the wireless resource management module retrieving local configuration entries to set the target block error rate for the ultra-reliable low-latency communication service to a certain value. Furthermore, the target block error rate for enhanced mobile broadband services is set at [value missing]. In a controlled environment where the reference signal received power is -90dBm and meets a preset Gaussian noise distribution, the control physical layer traverses the modulation and coding strategy levels from 0 to 28 in steps of 1 level, and determines the baseline level by searching for the level with the highest throughput under a given block error rate constraint. The value is then stored as a parameter in the scheduler's slice index address space.

[0041] When the access network system enters the service flow monitoring state, the distributed unit initiates a sampling alignment step for the physical downlink control channel log. The sampling clock generator captures the subframe start boundary of the system frame number SFN and adjusts the phase offset of the data extraction window to 10. Within a range of s, the media access control layer scheduler reads the control channel unit occupancy rate register of the current cell in real time and determines the occupancy rate based on the obtained real-time occupancy rate. Update collision trigger lines The specific calculation logic follows the following relationship: ,in, For collision trigger line, The total number of hybrid automatic retransmission request processes configured for the system. The system calculates the control channel unit occupancy rate for the current scheduling period and stores the resulting discretized value in the comparison port of the collision determination unit. This value is used in the scheduling degradation index. When fluctuations occur, a feedback judgment benchmark based on the current load density is provided.

[0042] After the system completes the allocation procedure for dedicated scheduling request resources, the media access control layer monitoring unit initiates closed-loop detection of the uplink synchronization status of the terminal device. The system counts the number of energy pulses on the dedicated scheduling request channel in real time within a preset observation period, uses the number of energy pulses as a physical quantity input parameter of terminal activity, and determines whether the reporting frequency is constant at 0. In response to the reporting frequency being constant at 0 for 50ms and the number of retransmissions by the radio link control layer (RLC) reaching the maximum threshold, the system sends a radio access bearer release request containing the logical channel identifier to the core network user plane function element and clears the physical resource block reservation queue information in the local cache of the media access control layer.

[0043] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0044] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for intelligent detection of 5G message transmission anomalies, characterized in that, Includes the following steps: Step 101: Obtain the downlink resource allocation characteristics of the radio access network media access control layer. The downlink resource allocation characteristics include the downlink control information format and the modulation and coding scheme level. Step 102: Monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, and determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and determine the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. Step 103: Combine the format degradation weights and scheduling response delay characteristics corresponding to the handover action to calculate the scheduling degradation index that characterizes the air interface resource congestion state. Step 104: Obtain the channel quality indication fed back by the terminal device, and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range; Step 105: Under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, if the scheduling degradation index reaches the abnormal judgment threshold dynamically determined according to the service priority identifier of 5G message, an air interface transmission abnormal command for the target logical channel is output, and the radio access network media access control layer is triggered to suspend the physical resource block reservation queue corresponding to the target logical channel.

2. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, Before determining that a 5G message transmission anomaly has occurred, the method further includes the following steps: Step 201, obtaining the scheduling request collision density and the negative response density of the hybrid automatic repeat request of the terminal device; Step 202, verifying the scheduling request collision density and the negative response density by correlation to determine the logical link blocking state of the terminal device in the physical resource block allocation process; Step 203, generating a correction coefficient based on the logical link blocking state, and using the correction coefficient to correct the scheduling degradation index, so that Step 105 executes the determination logic based on the corrected scheduling degradation index.

3. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, After outputting the air interface transmission anomaly command for the target logical channel, the method further includes the following steps: Step 301, extracting the service priority identifier corresponding to the 5G message; Step 302, based on the service priority identifier, allocating dedicated scheduling request resources that are in an orthogonal state and do not generate spectrum overlap to the terminal device on the base station side.

4. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, Step 102 determines the scheduling response delay characteristics of the downlink by the step change of the modulation and coding strategy level within adjacent transmission time intervals, including the following sub-steps: Step 1021, count the total number of continuous transmission time intervals in which the modulation and coding strategy level is within a preset low-order value range; Step 1022, when the total number of continuous transmission time intervals exceeds a preset time threshold and the slope of the first block error rate of the corresponding downlink physical shared channel reaches a preset change rate limit, determine that the downlink has entered the scheduling response delay state.

5. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, Step 104 determines whether the drop in the channel quality indicator is within the preset non-fading fluctuation range, including the following sub-steps: Step 1041, synchronously acquire the reference signal received power and the signal-to-interference-plus-noise ratio of the terminal device; Step 1042: When the rate of change of the received power of the reference signal is below the first preset threshold and the variance of the fluctuation of the signal-to-interference-plus-noise ratio is below the second preset threshold, it is determined that the drop in the channel quality indication is within the non-fading fluctuation range.

6. The intelligent detection method for 5G message transmission anomalies according to claim 3, characterized in that, After allocating dedicated scheduling request resources to the terminal device, the method further includes the following steps: Step 601, monitoring the scheduling request reporting frequency of the terminal device on the dedicated scheduling request resources; Step 602, when the scheduling request reporting frequency is always 0 within a preset observation period, sending a radio access bearer release request to the core network user plane function network element to reclaim the physical resources corresponding to the physical resource block reserved queue.

7. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, Step 105 triggers the suspending of the physical resource block reservation queue by the radio access network media access control layer, including the following sub-steps: Step 1051, locate the hybrid automatic repeat request process identifier corresponding to the target logical channel; Step 1052, clear the soft repeat buffer corresponding to the hybrid automatic repeat request process identifier, and restrict the downlink resource allocation authority of the terminal device in the subsequent 3 scheduling cycles.

8. The intelligent detection method for 5G message transmission anomalies according to claim 1, characterized in that, The method is applied to the distributed unit of 5G base stations to determine the abnormal state of air interface transmission based on the abnormal quantization characteristics of radio access network control plane signaling without parsing the 5G message application layer payload carrier.

9. A 5G message transmission anomaly intelligent detection system, used to implement the 5G message transmission anomaly intelligent detection method as described in claim 1, characterized in that, include: The feature acquisition module is used to acquire the downlink resource allocation features of the media access control layer of the radio access network. The downlink resource allocation features include the downlink control information format and the modulation and coding scheme level. The delay detection module, connected to the feature acquisition module, is used to monitor the downlink control information format, which is the downlink scheduling signaling format, in real time, determine whether there is a switching action from DCI format 1_1 to DCI format 1_0 in the downlink control information format, and associate the step change of the modulation and coding strategy level in adjacent transmission time intervals to determine the scheduling response delay characteristics of the downlink. The degradation assessment module, connected to the latency detection module, is used to fuse the format degradation weights corresponding to the handover action with the scheduling response latency characteristics to calculate the scheduling degradation index that characterizes the air interface resource congestion state. The channel monitoring module is used to acquire the channel quality indication fed back by the terminal device and determine whether the drop in the channel quality indication is within the preset non-fading fluctuation range. The anomaly determination module, connected to the degradation assessment module and the channel monitoring module respectively, is used to output an air interface transmission anomaly command for the target logical channel if the scheduling degradation index reaches the anomaly determination threshold dynamically determined according to the service priority identifier of the 5G message, under the technical premise that the drop in channel quality indication is within the non-fading fluctuation range, and triggers the radio access network media access control layer to suspend the physical resource block reservation queue corresponding to the target logical channel.

Images