A DCI alarm processing method and device, computer equipment and storage medium

Abstract

This invention relates to a DCI alarm processing method, comprising: real-time acquisition of raw alarm data from a data center interconnection network; semantic parsing and feature encoding of the raw alarm data to obtain alarm feature vectors; constructing an alarm correlation graph based on the alarm feature vectors, performing correlation reasoning on the alarm correlation graph using a graph learning model to determine the correlation between alarms and locate the root cause alarms; generating alarm processing strategies matching the root cause alarms using an intelligent decision-making algorithm based on network topology status, service priority configuration, and real-time traffic status; and distributing the alarm processing strategies to the corresponding network devices to complete automated alarm handling. By extracting features from raw alarm data, constructing an alarm correlation graph, and performing correlation reasoning, the inherent correlation between alarms can be automatically identified, the root cause alarms can be accurately located, the false alarm rate can be effectively reduced, redundant derived alarms can be eliminated, and the accuracy of alarm identification and fault location can be significantly improved.

Description

Technical Field

[0001] This invention relates to the field of communication network technology, specifically to a DCI alarm processing method, apparatus, computer equipment, and storage medium. Background Technology

[0002] With the large-scale application of cloud computing, big data, and distributed network technologies, Data Center Interconnect (DCI) networks have become the core infrastructure supporting cross-regional data interaction and business collaboration. The number of network nodes, link scale, and business traffic are growing exponentially, resulting in alarm data that is massive, sudden, and correlated, placing extremely high demands on the real-time, accuracy, and intelligence of alarm processing. Currently, the industry generally adopts three traditional technical solutions for DCI alarm processing: The first is static threshold-triggered alarms, which trigger alarm prompts by preset fixed thresholds, which cannot adapt to dynamic network changes and has extremely poor alarm adaptability; the second is expert rule matching alarms, which relies on manual sorting and maintenance of a large number of correlation rules, resulting in lagging rule updates, extremely high maintenance costs, and inability to cover unknown fault scenarios; the third is traditional machine learning alarm classification, where feature extraction relies entirely on manual experience definitions, which can only achieve simple classification of single alarms and cannot uncover deep correlations between alarms. Summary of the Invention

[0003] The purpose of this invention is to provide a DCI alarm processing method to solve the problems mentioned in the background art.

[0004] A first aspect of the present invention provides a DCI alarm processing method, comprising: Real-time acquisition of raw alarm data from the data center interconnection network; The original alarm data is semantically parsed and feature-encoded to obtain alarm feature vectors; An alarm association graph is constructed based on the alarm feature vectors. An association reasoning is performed on the alarm association graph using a graph learning model to determine the association relationship between alarms and locate the root cause alarm. Based on the network topology status, service priority configuration, and real-time traffic status, an alarm handling strategy matching the root cause alarm is generated through an intelligent decision-making algorithm. The alarm handling strategy is distributed to the corresponding network devices to complete the automated handling of alarms.

[0005] In one possible implementation, the step of performing semantic parsing and feature encoding on the original alarm data to obtain an alarm feature vector includes: Temporal and spatial features are extracted from the original alarm data. Joint encoding processing is performed on the temporal and spatial features. The correlation weights between features of different dimensions are calculated through a self-attention mechanism to generate an alarm feature vector that integrates temporal, spatial and semantic information.

[0006] In one possible implementation, the semantic parsing of the original alarm data includes: Unstructured alarm text is extracted from the original alarm data, and semantic vectorization representation processing is performed on the unstructured alarm text using a pre-trained language model to generate basic semantic features as the basis for joint encoding.

[0007] In one possible implementation, constructing the alarm association graph based on the alarm feature vector includes: Each alarm event is represented by a corresponding alarm feature vector, and the represented alarm events are mapped to graph nodes. The temporal proximity, causal correlation, and system hierarchy dependency between alarm events are constructed as edges of the alarm association graph. The edge weights are calculated based on the alarm association strength to form a weighted directed alarm association graph.

[0008] In one possible implementation, the step of performing association reasoning on the alarm association graph using a graph learning model includes: inputting the alarm feature vectors and edge weights corresponding to the graph nodes into the graph learning model, performing reasoning calculations on the alarm association relationships through the graph learning model, obtaining the association confidence between each alarm, generating an alarm association matrix, and completing root cause alarm location based on the association matrix.

[0009] In one possible implementation, the step of generating an alarm handling strategy matching the root cause alarm using an intelligent decision-making algorithm based on network topology status, service priority configuration, and real-time traffic status includes: The system acquires network topology, service level configuration, and real-time traffic distribution data, and constructs these data into a state space for reinforcement learning. It then constructs a reward function based on network operation metrics, calculates the correlation strength between network nodes through a multi-head attention mechanism, dynamically aggregates neighbor node information and generates node embeddings, predicts the probability of service quality satisfaction for each transmission path based on the node embeddings, and generates the optimal alarm handling strategy that meets service quality requirements.

[0010] In one possible implementation, a reward function is constructed based on network throughput, data transmission latency, data packet loss rate, and system resource consumption rate. The weights of each indicator in the reward function are dynamically allocated and normalized using the entropy weight method. Based on the optimized reward function, the alarm handling strategy is iteratively optimized to generate the final optimal alarm handling strategy.

[0011] A second aspect of the present invention provides a DCI alarm processing method and apparatus system, comprising an alarm acquisition module, a feature extraction module, a correlation analysis module, a processing decision module, and an execution module that are sequentially connected in communication. The alarm acquisition module is used to collect raw alarm data generated by DCI network devices in real time and transmit the raw alarm data to the feature extraction module. The feature extraction module is used to perform semantic parsing, temporal and spatial feature extraction and joint encoding on the original alarm data, generate alarm feature vectors and transmit them to the correlation analysis module; The correlation analysis module is used to construct an alarm correlation graph based on alarm feature vectors, perform correlation reasoning calculations, identify alarm correlation relationships and locate root cause alarms, and transmit root cause alarm information to the processing decision module. The processing decision module is used to obtain network topology status, service priority configuration and real-time traffic distribution data, generate alarm processing strategies that match root cause alarms and transmit them to the execution module; The execution module is used to receive alarm processing policies and distribute the alarm processing policies to the corresponding network devices to complete the automated handling of alarms.

[0012] A third aspect of the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the DCI alarm processing method as described in the first aspect of the present invention.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the DCI alarm processing method as described in the first aspect of the present invention.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. By extracting features from the original alarm data, constructing an alarm association graph, and performing association reasoning, the inherent relationship between alarms can be automatically identified, the root cause alarms of faults can be accurately located, the false alarm rate can be effectively reduced, redundant derivative alarms can be eliminated, and the accuracy of alarm identification and fault location can be greatly improved.

[0015] 2. By generating matching alarm handling strategies based on network topology, service priority, and real-time traffic status, no manual analysis or decision-making is required. At the same time, the strategies are automatically distributed to network devices to complete the entire process of automated handling, which significantly improves the real-time performance of alarm handling and greatly reduces the need for manual intervention. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the DCI alarm processing method of the present invention. Figure 2 This is a schematic diagram of the architecture of the DCI alarm processing device of the present invention; Figure 3This is a schematic diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0018] It should be noted that the serial numbers assigned to the components in the embodiments of the present invention, such as "first" and "second", are only used to distinguish the described objects and have no sequential or technical meaning.

[0019] The following is combined Figure 1 A DCI alarm processing method of the present invention is described.

[0020] A DCI alarm processing method includes: S1. Real-time acquisition of raw alarm data from the data center interconnection network; The acquisition unit, which adopts a multi-threaded parallel architecture, establishes communication connections with all devices in the DCI network, such as switches, routers, and optical transmission equipment, through a standard network interface. It receives raw alarm data reported by the devices in real time according to a preset acquisition cycle. The alarm reception rate meets the high concurrency requirement of at least 1,000 alarms per second, ensuring that no alarm data is missed.

[0021] S2. Perform semantic parsing and feature encoding on the original alarm data to obtain alarm feature vectors; The semantic parsing of the original alarm data can be performed as follows: extract unstructured text information (such as alarm description, device log, fault prompt) from the original alarm data, and perform semantic vectorization representation processing on the above text through the BERT pre-trained language model to convert the irregular text into structured basic semantic features. These features serve as the basis for subsequent joint encoding of temporal and spatial features.

[0022] This step can be performed as follows: extract the temporal and spatial features from the original alarm data, perform joint encoding processing on the temporal and spatial features, calculate the correlation weights between features of different dimensions through a self-attention mechanism, and generate an alarm feature vector that integrates temporal, spatial, and semantic information.

[0023] Among them, all time-related feature information is screened and extracted from the original alarm data, including alarm timestamp, alarm duration, alarm trigger interval, alarm repetition frequency, etc. These features can completely depict the distribution pattern and trend of alarm events in the time dimension, and are used to identify the time correlation attributes between alarms in the future.

[0024] Spatial dimension features related to network location and topology are extracted from the raw alarm data, including the device identifier to which the alarm belongs, the physical deployment location of the device, link port information, network topology level, and cross-regional node affiliation. These features can accurately characterize the spatial location and topology affiliation of alarm events in the DCI network, and are used to identify the spatial correlation attributes between alarms in the future.

[0025] The standardized temporal and spatial features are input into the feature encoding unit, and the Transformer encoder architecture is used to perform joint encoding on the two types of features, rather than encoding the temporal and spatial features independently or simply concatenating them. Through joint encoding, the temporal and spatial attributes of the alarm are deeply bound, so that the encoding result can carry the temporal distribution pattern and spatial location information of the alarm at the same time, and retain the complete correlation characteristics of the alarm in the spatiotemporal dimension.

[0026] A self-attention mechanism is introduced during the joint encoding process to automatically calculate the correlation between feature information in three dimensions: temporal, spatial, and semantic features, without the need for manual weighting. This mechanism autonomously learns the importance of different feature dimensions to alarm representation, assigning higher weights to features that reflect the essence of the alarm and have high correlation, and assigning lower weights to redundant or irrelevant features, thereby accurately uncovering the inherent dependencies between the three types of features. Based on the multi-dimensional feature correlation weights obtained from the self-attention mechanism, the temporal, spatial, and semantic features are weighted and fused for encoding, integrating the effective information of the three types of features into a single, standardized feature vector, ultimately generating an alarm feature vector that integrates temporal, spatial, and semantic information. This feature vector can comprehensively and accurately represent all the core attributes of a single alarm, and compared to traditional single-dimensional or simply concatenated feature vectors, it has stronger discriminative and representational capabilities.

[0027] S3. Construct an alarm association graph based on the alarm feature vector, and perform association reasoning on the alarm association graph through a graph learning model to determine the association relationship between alarms and locate the root cause alarm. Each alarm event is represented using a corresponding alarm feature vector. These represented alarm events are then mapped to graph nodes. The temporal proximity, causal correlation, and system hierarchy dependency among the alarm events are used to construct edges in the alarm association graph. Edge weights are calculated based on the alarm association strength, forming a weighted directed alarm association graph. The alarm feature vectors corresponding to the graph nodes and the edge weights are input into a graph learning model. The graph learning model infers and calculates the alarm association relationships, obtaining the association confidence between each alarm, generating an alarm association matrix, and using this matrix to locate the root cause of the alarm.

[0028] Each actual alarm event is digitally represented, using its corresponding alarm feature vector as its unique digital representation. This transforms alarm events from "unstructured descriptions" to "standardized vectors," ensuring that the characteristics of each alarm event are quantifiable and computable. Each alarm event represented by its feature vector is mapped to a node in an alarm association graph, with each node carrying core attributes such as the corresponding alarm event's feature vector, alarm type, and severity level.

[0029] When constructing the edges of the alarm association graph, in addition to considering the temporal proximity, causal correlation, and system hierarchy dependency among various alarm events, two additional key dimensions are introduced: temporal distance and network topology distance. Temporal distance represents the time interval between two alarm events, while network topology distance represents the physical link distance between the devices belonging to the two alarm events in the DCI network topology, further improving the accuracy of correlation determination. These five indicators—temporal proximity, causal correlation, system hierarchy dependency, temporal distance, and network topology distance—are used together as the basis for constructing the edges of the alarm association graph. The association strength weight W of each edge is then calculated using the following formula. ij : W ij =W1×T ij +W2×C ij +W3×D ij +W4×1 / 1+L tij +W5×1 / 1+L sij , T ij C is the time proximity coefficient between alarm events i and j (values ​​range from 0 to 1). ij D is the causal correlation coefficient (value 0-1). ij L is the system hierarchy dependency coefficient (value 0-1). tij L represents the time interval (in seconds) between alarm events i and j. sij The network topology distance (in hops) between alarm events i and j is given, and W1~W5 are the normalized weights of each indicator.

[0030] After the graph is constructed, the alarm feature vector corresponding to each node in the graph, and the association strength weight W corresponding to each edge are used to construct the graph. ij The data are input into a graph learning model for association inference calculations. The graph learning model aggregates node features and edge weights through hidden layers to automatically uncover implicit associations between different nodes (alarm events) and calculates the association confidence P between any two alarm events. ij (Values ​​range from 0 to 1; higher confidence indicates a stronger correlation), and all correlation confidence values ​​are integrated to generate an alarm correlation matrix M. The correlation matrix expression is as follows: Where n is the number of alarm events, P ii =1 (self-association confidence is 1). Finally, based on this association matrix, alarm association pairs with association confidence higher than a preset threshold (such as 0.8) are filtered out, redundant derived alarms are eliminated, the core association links are focused, and the root cause alarm that caused a series of alarms is finally located, realizing the accuracy and intelligence of root cause location.

[0031] S4. Based on the network topology status, service priority configuration, and real-time traffic status, generate an alarm handling strategy that matches the root cause alarm through an intelligent decision-making algorithm. Specifically, three core data types are collected: DCI network topology, service level configuration, and real-time traffic distribution. After preprocessing, these are integrated to construct a reinforcement learning state space, comprehensively covering network device connectivity, different service QoS requirements (such as latency and packet loss rate requirements for core and general services), and real-time traffic load on each link. This provides a foundational support for intelligent decision-making that aligns with real-world scenarios. A reward function is constructed around four core indicators: network throughput, data transmission latency, data packet loss rate, and system resource consumption rate. This function is used to quantitatively evaluate the merits of candidate alarm handling strategies. Unlike traditional fixed-weight decision-making methods, this step dynamically allocates the weights of each indicator using the entropy weight method. This eliminates the need for manual pre-setting and allows for adaptive adjustments based on real-time network conditions. For example, during peak periods for core services, the weights of latency and packet loss rate are prioritized, while during idle periods, the focus is on optimizing resource consumption rate. This addresses the shortcomings of traditional decision-making methods, which are highly subjective and unable to adapt to dynamic network scenarios.

[0032] Specifically, the entropy weight method is used to dynamically allocate and normalize the weights of each indicator in the reward function, ensuring the objectivity and adaptability of the weight allocation. The specific calculation process is as follows: first, the information entropy E of each indicator is calculated. i : Where n is the number of indicators, m is the number of samples, and p ij Let be the standardized value of the j-th sample for the i-th indicator, and then calculate the weight λ of each indicator based on the information entropy. iAfter normalization, it satisfies The specific formula is as follows: After weight optimization, a multi-head attention mechanism is introduced to calculate the correlation strength between network nodes. By using multiple attention heads to compute in parallel, the implicit correlations between different network nodes (devices, links) are accurately captured, and the neighbor node information of each node is dynamically aggregated to eliminate the influence of isolated node information on decision-making. Based on the aggregated node information, a node embedding vector is generated. This vector can accurately represent the operating status and correlation characteristics of network nodes, providing support for transmission path evaluation. Subsequently, the node embedding vector is input into a multilayer perceptron (MLP) to predict the service quality satisfaction probability of each transmission path, and candidate alarm handling strategies are selected for those whose service quality satisfaction probability is higher than a preset threshold.

[0033] Finally, using the optimized reward function as the evaluation criterion, all candidate strategies are iteratively optimized. Through multiple rounds of iterative calculation, the strategy with the highest reward value is selected as the final optimal alarm handling strategy. This strategy can accurately match the handling requirements of root cause alarms, and also take into account the balance between network throughput, transmission latency, packet loss rate and resource consumption, ensuring that the network can quickly recover to a stable operating state after alarm handling, while meeting the service level requirements of each service.

[0034] S5. Distribute the alarm handling strategy to the corresponding network devices to complete the automated alarm handling.

[0035] like Figure 2 As shown, the present invention also provides a DCI alarm device, including an alarm acquisition module 10, a feature extraction module 20, a correlation analysis module 30, a processing decision module 40, and an execution module 50 that are connected in sequence via communication. The alarm acquisition module 10 is used to collect raw alarm data generated by DCI network devices in real time and transmit the raw alarm data to the feature extraction module. The feature extraction module 20 is used to perform semantic parsing, temporal and spatial feature extraction and joint encoding on the original alarm data, generate alarm feature vectors and transmit them to the correlation analysis module; The correlation analysis module 30 is used to construct an alarm correlation graph based on alarm feature vectors, perform correlation reasoning calculations, identify alarm correlation relationships and locate root cause alarms, and transmit root cause alarm information to the processing decision module. The processing decision module 40 is used to acquire network topology status, service priority configuration and real-time traffic distribution data, generate alarm processing strategies that match root cause alarms and transmit them to the execution module; The execution module 50 is used to receive alarm processing policies and send the alarm processing policies to the corresponding network devices to complete the automated handling of alarms.

[0036] In one embodiment, such as Figure 3 As shown, a computer device 60 is provided, including a memory 62, a processor 61, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the processor 61 executes the computer program 63, it implements the steps in the data processing method of the above embodiments. To avoid repetition, these steps will not be repeated here. Alternatively, when the processor 61 executes the computer program 63, it implements the functions of each module in the above embodiments of the DCI alarm processing method and device system. To avoid repetition, these functions will not be repeated here.

[0037] In one embodiment, a readable storage medium is provided, which stores a computer program 63. When the computer program 63 is executed by the processor 61, it implements the steps in the data processing method of the above embodiments. To avoid repetition, these steps will not be repeated here. Alternatively, when the processor 61 executes the computer program 63, it implements the functions of each module in the above data processing device embodiments. To avoid repetition, these functions will not be repeated here.

[0038] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), direct RAM (RDRAM) on RAMbus, direct memory bus dynamic RAM (DRDRAM), and dynamic RAM on RAMbus, etc.

[0039] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional modules, sub-modules, and units as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0040] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A DCI alarm processing method, characterized in that, include: Real-time acquisition of raw alarm data from the data center interconnection network; The original alarm data is semantically parsed and feature-encoded to obtain alarm feature vectors; An alarm association graph is constructed based on the alarm feature vectors. An association reasoning is performed on the alarm association graph using a graph learning model to determine the association relationship between alarms and locate the root cause alarm. Based on the network topology status, service priority configuration, and real-time traffic status, an alarm handling strategy matching the root cause alarm is generated through an intelligent decision-making algorithm. The alarm handling strategy is distributed to the corresponding network devices to complete the automated handling of alarms.

2. The DCI alarm processing method according to claim 1, characterized in that, The step of performing semantic parsing and feature encoding on the original alarm data to obtain an alarm feature vector includes: Temporal and spatial features are extracted from the original alarm data. Joint encoding processing is performed on the temporal and spatial features. The correlation weights between features of different dimensions are calculated through a self-attention mechanism to generate an alarm feature vector that integrates temporal, spatial and semantic information.

3. The DCI alarm processing method according to claim 2, characterized in that, The semantic parsing of the original alarm data includes: Unstructured alarm text is extracted from the original alarm data, and semantic vectorization representation processing is performed on the unstructured alarm text using a pre-trained language model to generate basic semantic features as the basis for joint encoding.

4. The DCI alarm processing method according to claim 1, characterized in that, The construction of the alarm association graph based on the alarm feature vector includes: Each alarm event is represented by a corresponding alarm feature vector, and the represented alarm events are mapped to graph nodes. The temporal proximity, causal correlation, and system hierarchy dependency between alarm events are constructed as edges of the alarm association graph. The edge weights are calculated based on the alarm association strength to form a weighted directed alarm association graph.

5. The DCI alarm processing method according to claim 4, characterized in that, The step of performing association reasoning on the alarm association graph using a graph learning model includes: inputting the alarm feature vectors and edge weights corresponding to the graph nodes into the graph learning model, performing inference calculations on the alarm association relationships through the graph learning model, obtaining the association confidence between each alarm, generating an alarm association matrix, and completing root cause alarm location based on the association matrix.

6. The DCI alarm processing method according to claim 1, characterized in that, The step of generating an alarm handling strategy matching the root cause alarm based on network topology status, service priority configuration, and real-time traffic status using an intelligent decision-making algorithm includes: The system acquires network topology, service level configuration, and real-time traffic distribution data, and constructs these data into a state space for reinforcement learning. It then constructs a reward function based on network operation metrics, calculates the correlation strength between network nodes through a multi-head attention mechanism, dynamically aggregates neighbor node information and generates node embeddings, predicts the probability of service quality satisfaction for each transmission path based on the node embeddings, and generates the optimal alarm handling strategy that meets service quality requirements.

7. The DCI alarm processing method according to claim 6, characterized in that, A reward function is constructed based on network throughput, data transmission latency, data packet loss rate, and system resource consumption rate. The weights of each indicator in the reward function are dynamically allocated and normalized using the entropy weight method. Based on the optimized reward function, the alarm handling strategy is iteratively optimized to generate the final optimal alarm handling strategy.

8. A DCI alarm processing device, characterized in that, It includes an alarm acquisition module, a feature extraction module, a correlation analysis module, a processing and decision-making module, and an execution module that are connected in sequence via communication. The alarm acquisition module is used to collect raw alarm data generated by DCI network devices in real time and transmit the raw alarm data to the feature extraction module. The feature extraction module is used to perform semantic parsing, temporal and spatial feature extraction and joint encoding on the original alarm data, generate alarm feature vectors and transmit them to the correlation analysis module; The correlation analysis module is used to construct an alarm correlation graph based on alarm feature vectors, perform correlation reasoning calculations, identify alarm correlation relationships and locate root cause alarms, and transmit root cause alarm information to the processing decision module. The processing decision module is used to obtain network topology status, service priority configuration and real-time traffic distribution data, generate alarm processing strategies that match root cause alarms and transmit them to the execution module; The execution module is used to receive alarm processing policies and distribute the alarm processing policies to the corresponding network devices to complete the automated handling of alarms.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the DCI alarm processing method as described in any one of claims 1-7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the DCI alarm processing method as described in any one of claims 1-7.

Images