Fault detection method, system and apparatus, and storage medium

Abstract

The present invention relates to the technical field of artificial intelligence for information technology operations. Disclosed are a fault detection method, system and apparatus, and a storage medium. The method comprises: determining a plurality of fault events having an association relationship therebetween, and a plurality of corresponding fault network elements; on the basis of the fault events corresponding to the plurality of fault network elements, determining target protocol stacks between the plurality of fault network elements; then, on the basis of the target protocol stack of each fault network element, determining multiple protocol layers of each fault network element and a hierarchical structure of the multiple protocol layers, so as to determine internal propagation paths; next, on the basis of the corresponding target protocol stacks between the plurality of fault network elements, determining external propagation paths in the plurality of fault network elements that are connected by means of corresponding target protocol layers; and finally, on the basis of a fault event occurring in each fault network element, determining fault layers in the multiple protocol layers corresponding to each fault network element. In this way, fault propagation paths between a plurality of fault events corresponding to multiple fault layers can be determined by means of internal propagation paths and external propagation paths.

Description

Fault detection method, system, device and storage medium

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on August 11, 2023, with application number 202311018430.1 and invention name “Fault Detection Method, System, Device and Storage Medium”. The entire contents of the above patent are incorporated into this application by reference. Technical Field

[0002] The present invention relates to the field of intelligent operation and maintenance technology, and in particular to a fault detection method, system, device and storage medium. Background Art

[0003] The artificial intelligence for information technology operations (AIOps) platform applies artificial intelligence (AI) to information technology operations (IT Operations). It uses machine learning algorithms (such as deep learning (DL)) to analyze IT data such as logs, monitoring indicators, and event data to achieve operation and maintenance management such as fault detection, event management, capacity planning, and security monitoring in business systems, thereby improving the system's operation and maintenance management efficiency.

[0004] However, when a business system experiences a series of related failure events, AIOps can only determine the corresponding faulty network element based on the failure event, but cannot determine the fault propagation path between the faulty network elements. This is not conducive to the inspection and maintenance of operation and maintenance personnel, and the operation and maintenance efficiency is low.

[0005] Summary of the Invention

[0006] Embodiments of the present invention provide a fault detection method, system, device, and storage medium.

[0007] In a first aspect, the present invention provides a fault detection method for a system including multiple network elements, comprising: determining multiple fault events with an associated relationship, and multiple fault network elements corresponding to the multiple fault events; determining a target protocol stack between the multiple fault network elements based on a preset mutual communication relationship between the fault events corresponding to the multiple fault network elements; determining the multi-layer protocol layer of each fault network element based on the target protocol stack of each fault network element, and determining the internal propagation path of the corresponding fault network element according to the hierarchical structure of the multi-layer protocol layer; determining an external propagation path between the multiple fault network elements through the target protocol layer in the corresponding target protocol stack based on the preset mutual connection relationship between the corresponding target protocol stacks of the multiple fault network elements, wherein the target protocol layer for connection of each fault network element is a protocol layer in the multi-layer protocol layer corresponding to each fault network element; determining a fault layer in the multi-layer protocol layer corresponding to each fault network element based on the corresponding fault event occurring in each fault network element; and determining a fault propagation path between multiple fault events corresponding to the multi-layer fault layer based on the internal propagation path and the external propagation path.

[0008] In a possible implementation of the first aspect above, the internal propagation path of the corresponding faulty network element is determined according to the hierarchical structure of the multi-layer protocol layers, including: determining the internal topology of the corresponding faulty network element based on the multi-layer protocol layers of the faulty network element and the hierarchical structure of the multi-layer protocol layers; determining the internal propagation path of the faulty network element in the multi-layer protocol layers of the corresponding faulty network element based on the internal topology of the faulty network element.

[0009] In a possible implementation of the first aspect above, based on the preset interconnection relationship between the corresponding target protocol stacks of multiple faulty network elements, the external propagation path connecting the multiple faulty network elements through the target protocol layer in the corresponding target protocol stack is determined, including: based on the preset interconnection relationship between the corresponding target protocol stacks of the multiple faulty network elements, the multiple internal topologies corresponding to the multiple faulty network elements are connected through the target protocol layer of the corresponding target protocol stack to form a detailed topology; based on the target protocol layer in the detailed topology through which the multiple internal topologies are connected, the external propagation path of the multiple faulty network elements is determined.

[0010] In a possible implementation of the first aspect above, based on the corresponding fault events occurring in each faulty network element, the fault layer in the multi-layer protocol layer corresponding to each faulty network element is determined, including: event numbering of each fault event; based on the corresponding fault events occurring in each faulty network element and the event numbers corresponding to the fault events, the fault layer in the multi-layer protocol layer corresponding to each faulty network element and the layer number of the fault layer are determined; based on the layer number and the event number, the correspondence between the fault layer and the fault event is determined.

[0011] In a possible implementation of the first aspect above, based on the internal propagation path and the external propagation path, the fault propagation path between multiple fault events corresponding to the multi-layer fault layer is determined, including: based on the internal propagation path and the external propagation path, determining the hierarchical propagation path of the multi-layer fault layer corresponding to the multiple fault network elements in the multi-layer protocol layer of the multiple fault network elements; based on the hierarchical propagation path and the correspondence between the fault layer and the fault event, determining the fault propagation path of the corresponding multiple fault events in the multi-layer protocol layer of the multiple fault network elements.

[0012] In a possible implementation of the first aspect above, determining a target protocol stack between multiple faulty network elements includes: determining a preset set of protocol stacks for mutual communication between the multiple faulty network elements, wherein the protocol stack set includes multiple protocol stacks for mutual communication between the multiple faulty network elements; and determining a target protocol stack corresponding to the multiple faulty network elements from the multiple protocol stacks based on a preset mutual communication relationship corresponding to the fault events between the multiple faulty network elements.

[0013] In a possible implementation of the first aspect above, determining multiple fault events with an associated relationship includes: determining multiple fault occurrence times corresponding to the multiple fault events, corresponding to the multiple fault occurrence times satisfying a preset time association relationship, and determining that the multiple fault events have an associated relationship; or, determining multiple fault network elements corresponding to the multiple fault events, corresponding to the multiple fault network elements satisfying a preset network element association relationship, and determining that the multiple fault events have an associated relationship; or, determining multiple event types corresponding to the multiple fault events, corresponding to the multiple event types satisfying a preset type association relationship, and determining that the multiple fault events have an associated relationship.

[0014] In a second aspect, the present invention provides a fault detection system, comprising: an internal topology generation module, which determines the target protocol stack between multiple fault network elements from a preset set of protocol stacks for mutual communication between multiple fault network elements based on the preset mutual communication relationship of the fault events corresponding to the multiple fault network elements; determines the internal topology of the corresponding fault network element based on the multi-layer protocol layer corresponding to the target protocol stack of each fault network element and the hierarchical structure of the multi-layer protocol layer; a detailed topology generation module, which connects the multiple internal topologies corresponding to the multiple fault network elements through the target protocol layer in the corresponding target protocol stack based on the preset mutual connection relationship between the multiple fault network elements to form a detailed topology, wherein each fault network element is used to connect The target protocol layer is the protocol layer in the multi-layer protocol layer corresponding to each faulty network element; the event calibration module assigns event numbers to each fault event; based on the corresponding fault events occurring in each faulty network element and the event numbers corresponding to the fault events, the fault layer and the layer number of the fault layer in the multi-layer protocol layer corresponding to each faulty network element are determined; based on the layer number and the event number, the correspondence between the fault layer and the fault event is determined; the fault detection module determines, based on the internal topology and the detailed topology, the hierarchical propagation paths of the multi-layer fault layers corresponding to multiple faulty network elements in the multi-layer protocol layers of multiple faulty network elements; based on the hierarchical propagation paths and the correspondence between the fault layer and the fault event, the fault propagation paths of the corresponding multiple fault events in the multi-layer protocol layers of multiple faulty network elements are determined.

[0015] In a third aspect, the present invention provides a fault detection device, comprising: an internal topology generation unit, which determines the target protocol stack between multiple fault network elements from a preset set of protocol stacks for mutual communication between multiple fault network elements based on the preset mutual communication relationship of the fault events corresponding to the multiple fault network elements; determines the internal topology of the corresponding fault network element based on the multi-layer protocol layer corresponding to the target protocol stack of each fault network element and the hierarchical structure of the multi-layer protocol layer; a detailed topology generation unit, which connects the multiple internal topologies corresponding to the multiple fault network elements through the target protocol layer in the corresponding target protocol stack based on the preset mutual connection relationship between the multiple fault network elements to form a detailed topology, wherein each fault network element is used to connect The target protocol layer is a protocol layer in the multi-layer protocol layer corresponding to each faulty network element; the event calibration unit assigns an event number to each fault event; based on the corresponding fault event occurring in each faulty network element and the event number corresponding to the fault event, the fault layer and the layer number of the fault layer in the multi-layer protocol layer corresponding to each faulty network element are determined; based on the layer number and the event number, the corresponding relationship between the fault layer and the fault event is determined; the fault detection unit determines, based on the internal topology and the detailed topology, the hierarchical propagation path of the multi-layer fault layer corresponding to multiple faulty network elements in the multi-layer protocol layer of multiple faulty network elements; based on the hierarchical propagation path and the corresponding relationship between the fault layer and the fault event, the fault propagation path of the corresponding multiple fault events in the multi-layer protocol layer of multiple faulty network elements is determined.

[0016] In a fourth aspect, an embodiment of the present invention provides a readable storage medium having instructions stored thereon, which, when executed on an electronic device, enables the electronic device to implement any one of the fault detection methods provided by the first aspect and various possible implementations of the first aspect.

[0017] In a fifth aspect, an embodiment of the present invention provides a program product, which includes instructions. When the instructions are executed by an electronic device, the electronic device can implement any fault detection method provided by the first aspect and various possible implementations of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] FIG1 shows a schematic diagram of fault information according to some embodiments of the present invention;

[0019] FIG2 shows an example schematic diagram of a fault detection system according to some embodiments of the present invention;

[0020] FIG3 shows an example schematic diagram of another fault detection system according to some embodiments of the present invention;

[0021] 4A and 4B are schematic diagrams illustrating an example of an implementation process of an internal topology generation module according to some embodiments of the present invention;

[0022] FIG5 is a schematic diagram illustrating an example of an implementation process of a detail topology generation module according to some embodiments of the present invention;

[0023] FIG6 shows a schematic diagram of an example internal topology of event calibration according to some embodiments of the present invention;

[0024] FIG7 shows a schematic diagram of an example of a detailed topology of calibration according to some embodiments of the present invention;

[0025] FIG8 is a schematic diagram illustrating an example of an implementation process of a fault detection module according to some embodiments of the present application;

[0026] FIG9 shows an exemplary schematic diagram of another fault detection system according to some embodiments of the present invention;

[0027] FIG10 is another exemplary schematic diagram illustrating an implementation process of an internal topology generation module according to some embodiments of the present invention;

[0028] FIG11 is a schematic diagram showing a flow chart of a fault detection method according to some embodiments of the present invention;

[0029] FIG12 shows an example schematic diagram of a fault detection device according to some embodiments of the present invention. DETAILED DESCRIPTION

[0030] Illustrative embodiments of the present invention include, but are not limited to, a fault detection method, system, apparatus, and storage medium.

[0031] The technical solution of the present invention is described below with reference to Figures 1 to 12 .

[0032] In some embodiments, when a business system deployed with AIOps has related fault events, AIOps can usually only determine the corresponding faulty network elements (such as routers, switches, servers, base stations, etc., hereinafter referred to as network elements) in the business system based on the fault events. It is difficult to determine the fault propagation path between the faulty network elements, that is, it is impossible to display the fault propagation path between the faulty network elements to the operation and maintenance personnel, which is not conducive to the operation and maintenance personnel to inspect and maintain the related fault events.

[0033] For example, in wireless communication systems, if a DC undervoltage alarm occurs in a power environment network element (e.g., power supply equipment, hereinafter referred to as a power environment network element) in a certain area, a power shortage may occur. This power shortage can cause wireless network elements (e.g., base stations) to experience Ethernet link failure alarms, network element connection interruptions, Universal Mobile Telecommunications System (UMTS) cell unavailability, and Internet Protocol (IP) clock link anomaly alarms. Furthermore, it can cause transmission network elements (e.g., switches) to experience Ethernet loss of signal (ETH LOS) and other failure events.

[0034] However, AIOps can usually only determine that the corresponding faulty network element is the dynamic ring network element based on the fault event of the DC undervoltage alarm, determine that the corresponding faulty network element is the wireless network element based on the fault events of the Ethernet link fault alarm, network element connection interruption, UMTS cell unavailability, and IP clock link abnormality alarm, and determine that the corresponding faulty network element is the transmission network element based on the fault event of ETH LOS. In other words, AIOps cannot determine the fault propagation path between the dynamic ring network element, wireless network element, and transmission network element for the aforementioned multiple fault events. Operation and maintenance personnel may need to perform fault tests on the dynamic ring network element, wireless network element, and transmission network element one by one according to each fault event. For example, they may need to perform fault tests on the dynamic ring network element, wireless network element, and transmission network element respectively to determine whether the corresponding transmission network element, wireless network element, and dynamic ring network element have corresponding faults, and then determine the fault propagation path based on the correlation of the faults of each network element in the fault test, resulting in low inspection and maintenance efficiency.

[0035] In this way, when a related fault event occurs in the business system, AIOps can only determine the corresponding faulty network element based on the fault event, but cannot determine the propagation path of the fault between the various faulty network elements, which is not conducive to the inspection and maintenance of operation and maintenance personnel.

[0036] To this end, the present invention proposes a fault detection method. When multiple fault events with a correlation are detected, the method first determines the faulty network element corresponding to each fault event, thereby obtaining multiple faulty network elements. Then, based on the preset intercommunication relationships among the fault events corresponding to the multiple faulty network elements, the method determines the target protocol stacks between the multiple faulty network elements. Subsequently, based on the target protocol stacks of each faulty network element, the method determines the multiple protocol layers and the hierarchical structure of the multiple protocol layers. Furthermore, based on the hierarchical structure of the multiple protocol layers, the method determines the internal propagation paths of the multiple protocol layers. Secondly, based on the preset intercommunication relationships among the target protocol stacks corresponding to the multiple faulty network elements, the method determines the external propagation paths connecting the multiple faulty network elements through the corresponding protocol layers. Finally, based on the fault events occurring in each faulty network element, the method determines the protocol layer (hereinafter referred to as the fault layer) to which each fault event belongs, thereby obtaining multiple fault layers. Thus, based on the determined internal and external propagation paths, the fault propagation paths between the multiple protocol layers of the multiple faulty network elements corresponding to the multiple fault layers can be determined. Furthermore, when multiple fault events with correlation are detected, the operation and maintenance personnel can perform inspection and maintenance in sequence according to the fault propagation paths of the multiple fault events to ensure the efficiency of inspection and maintenance.

[0037] In some embodiments, based on information such as multiple fault occurrence times, multiple faulty network elements, or multiple event types corresponding to multiple fault events, it can be determined whether multiple fault events have an associated relationship. For example, when the time interval between the latest fault occurrence time and the earliest fault occurrence time among the multiple fault occurrence times corresponding to multiple fault events meets a preset time interval threshold, or when the time difference between the multiple fault occurrence times corresponding to multiple fault events meets a preset time difference threshold, it can be determined that the multiple fault occurrence times meet a preset time association relationship, and further, it can be determined that the multiple fault events have an associated relationship. For another example, when the multiple faulty network elements corresponding to multiple fault events meet a preset network element association relationship, it can be determined that the multiple fault events have an associated relationship. Or, for example, when the multiple event types corresponding to multiple fault events meet a preset type association relationship, it can be determined that the multiple fault events have an associated relationship. In other embodiments, the association relationship of multiple fault events can also be determined by other means, such as the logical association of the events, which are not specifically limited.

[0038] In some embodiments, as shown in Figure 1, when multiple fault events with a correlation are detected, the fault information of the fault events can be input into the fault detection system, and the internal propagation paths and external propagation paths of the multiple fault events in the corresponding multiple fault network elements are determined by the fault detection system, and the fault propagation paths of the multiple fault events are determined based on the internal propagation paths and the external propagation paths, so that the fault propagation paths are credible and explainable.

[0039] The following uses the fault information shown in FIG1 as an example to explain some embodiments of the present application.

[0040] In some embodiments, as shown in FIG1 , the fault information includes: multiple faulty network elements including "H00001||wireless" wireless network element B (hereinafter referred to as wireless network element B), "H00002||transmission" transmission network element C (hereinafter referred to as transmission network element C), and "H00003||dynamic ring" dynamic ring network element A (hereinafter referred to as dynamic ring network element A) having a preset network element association relationship, and fault events corresponding to each faulty network element: the fault events of the wireless network element B include: Ethernet link failure alarm, network element connection interruption, UMTS cell unavailable, IP clock link abnormality alarm; the fault events of the transmission network element C include: ETH LOS fault event; the fault events of the dynamic ring network element A include: DC undervoltage alarm. Furthermore, after the fault information is input into the fault detection system, the fault detection system can output the multi-layer protocol layers of the multiple fault network elements corresponding to the multiple fault events in the fault information, the hierarchical structure of the multi-layer protocol layers, and the connection relationship between the multiple fault network elements through the corresponding protocol layers. Then, the internal propagation path of the corresponding fault network element can be determined through the hierarchical structure of the multi-layer protocol layers, and the external propagation path of the multiple fault network elements can be determined through the connection relationship between the corresponding protocol layers between the multiple fault network elements.

[0041] In some embodiments, as shown in Figure 1, wireless network element B includes, from bottom to top, the following protocol layers: a dynamic ring layer, a network element layer, a physical layer, an Ethernet protocol layer (ETH), a media access control protocol layer (MAC), an Internet protocol layer (IP), a stream control transmission protocol layer (SCTP), and a service layer (XnApp); transmission network element C includes, from bottom to top, the following protocol layers: a dynamic ring layer, a network element layer, a physical layer, an ETH, a MAC, and an IP; and dynamic ring network element A includes the protocol layer of the dynamic ring layer. Wireless network element B is connected to the physical layer corresponding to transmission network element C via its corresponding physical layer, wireless network element B is connected to the dynamic ring layer corresponding to dynamic ring network element A via its corresponding dynamic ring layer, and transmission network element C is connected to the dynamic ring layer corresponding to dynamic ring network element A via its corresponding dynamic ring layer.

[0042] It's understandable that within multiple layers of protocol, lower layers provide services for upper layers. Lower layer failures can cause upper layer failures, but upper layer failures won't cause lower layer failures. For example, within the multiple layers of protocol in wireless network element B, the ETH protocol provides services for the MAC protocol, the MAC protocol provides services for the IP protocol, and the IP protocol provides services for the SCTP protocol. Thus, based on the multiple layers of protocol included in each faulty network element and the hierarchical structure of these layers, the internal propagation path corresponding to each faulty network element can be determined.

[0043] For example, the internal propagation path of wireless network element B is: dynamic link layer, network element layer, physical layer, ETH, MAC, IP, SCTP, and service layer; the internal propagation path of transmission network element C is: dynamic link layer, network element layer, physical layer, ETH, MAC, and IP.

[0044] Furthermore, the external propagation paths of the multiple faulty network elements are determined based on the connection relationships between the corresponding protocol layers of the multiple faulty network elements. For example, the external propagation path between wireless network element B and transmission network element C is: the physical layer of wireless network element B is connected to the corresponding physical layer of transmission network element C; the external propagation path between wireless network element B and dynamic ring network element A is: the dynamic ring layer of wireless network element B is connected to the dynamic ring layer of dynamic ring network element A; and the external propagation path between transmission network element C and dynamic ring network element A is: the dynamic ring layer of transmission network element C is connected to the dynamic ring layer of dynamic ring network element A.

[0045] In some embodiments, based on the fault events occurring in each faulty network element, the fault layers in the multi-layer protocol layers corresponding to each faulty network element are determined to obtain the multi-layer fault layers. Based on the layer positions of the multi-layer fault layers in the multi-layer protocol layers of each faulty network element, as well as the internal propagation paths and external propagation paths, the fault propagation paths of the multiple fault events corresponding to the multi-layer fault layers can be determined.

[0046] For example, based on the DC undervoltage alarm fault event occurring in the dynamic ring network element A, the dynamic ring layer of the dynamic ring network element A is regarded as the fault layer; based on the Ethernet link fault alarm, network element connection interruption, UMTS cell unavailable, and IP clock link abnormality alarm fault events occurring in the wireless network element B, the ETH, network element layer, service layer, and IP layer in the wireless network element B are regarded as the fault layers; based on the ETH LOS fault event occurring in the transmission network element C, the ETH in the transmission network element C is regarded as the fault layer.

[0047] In some embodiments, each fault event and the fault layer corresponding to each fault event can be labeled and located (hereinafter referred to as calibration). For example, the fault layer can be calibrated using the icons and numbers shown by the star-shaped mark in Figure 1. The fault layer can be determined by the icon of the star-shaped mark, and the correspondence between the fault layer and the fault event can be determined by the number of the star-shaped mark, so that the fault propagation path of each fault event corresponding to each fault layer can be determined through calibration.

[0048] For example, if the DC undervoltage alarm fault event occurring in dynamic ring network element A is labeled "Alarm No. 6", the corresponding dynamic ring layer with dynamic ring network element A as the fault layer is labeled 6; the fault events of Ethernet link failure alarm, network element connection interruption, UMTS cell unavailability, and IP clock link abnormality alarm occurring in wireless network element B are labeled "Alarm No. 3", "Alarm No. 4", "Alarm No. 2", and "Alarm No. 1" respectively, and the corresponding ETH, network element layer, service layer, and IP layer with wireless network element B as the fault layer are labeled 3, 4, 2, and 1; the ETH LOS fault event occurring in transmission network element C is labeled "Alarm No. 5", and the corresponding ETH layer in transmission network element C as the fault layer is labeled 5.

[0049] In some embodiments, as shown in FIG1 , based on the internal propagation path, the external propagation path and the corresponding fault layer determined by the fault detection system, the fault propagation path of the fault event corresponding to the fault layer in each fault network element can be determined. For example, as shown in FIG1 , the fault detection system can output the detection results in text form. The detection results output by the fault detection system may include: the cause of the failure of the service layer in the wireless network element B is: "Alarm No. 4" occurs at the network element layer of the wireless network element B, which in turn causes "Alarm No. 3" to occur at the ETH layer of the wireless network element B, "Alarm No. 1" to occur at the IP layer, and "Alarm No. 2" to occur at the service layer, that is, the corresponding fault propagation path is 4→3→1→2; the cause of the associated failure between the dynamic ring network element A, the wireless network element B and the transmission network element C is: "Alarm No. 4" occurs at the ETH layer of the wireless network element B. "Alarm No. 3" causes "Alarm No. 5" to occur at the ETH layer of transmission network element C, and the corresponding fault propagation path is 3→5; "Alarm No. 6" occurs at the dynamic ring layer of dynamic ring device A, causing "Alarm No. 4" to occur at the network element layer of wireless network element B, and the corresponding fault propagation path is 6→4. Therefore, based on the fault propagation paths 4→3→1→2, 3→5, and 6→4, the root cause network element of the fault can be determined to be: dynamic ring network element A; the root cause protocol layer is: dynamic ring layer, and the root cause fault event is "Alarm No. 6".

[0050] In other embodiments, the detection results may also be displayed in other ways, such as in the form of charts or a combination of charts and text, etc., without specific limitation.

[0051] In this way, operation and maintenance personnel can perform inspection and maintenance in sequence according to the fault propagation paths of multiple fault events to ensure inspection and maintenance efficiency.

[0052] Figure 2 shows a block diagram of a trustworthy and explainable fault detection system, according to some embodiments of the present application. As shown in Figure 2, the fault detection system includes an internal topology generation module, a detailed topology generation module, an event calibration module, and a fault detection module. The input to the fault detection system can be an external topology and device data, or an external topology, device data, and a protocol stack information library. The output of the fault detection system is a calibrated detailed topology and detection results.

[0053] In some embodiments, after detecting multiple fault events and determining the multiple faulty network elements corresponding to the multiple fault events, the fault detection system takes each faulty network element as a vertex, and obtains the communication connection relationship between the multiple faulty network elements through the asset management data of the network management device (network management asset management center) as the edge connecting the two corresponding vertices to form an external topology. It can be understood that the external topology includes vertices and edges, each vertex is used to represent a faulty network element, and the edge between two vertices is used to represent the communication connection relationship between the two faulty network elements corresponding to the two vertices. The connection relationship can carry multiple protocol stacks. Subsequently, the fault detection system obtains the device data (such as alarms, KPIs, logs, etc.) of the corresponding faulty network element based on each vertex on the external topology, so as to subsequently determine the fault events that occurred in the corresponding faulty network element based on the device data of each faulty network element. At the same time, the fault detection system determines the type of protocol stack that has failed based on the fault event on the faulty network element connected to each edge, and then obtains the protocol stack information from the protocol stack information library based on the protocol stack type. The protocol stack information includes multiple protocol layers and the hierarchical structure of multiple protocol layers.

[0054] In some embodiments, the fault detection system inputs the external topology, device data, and protocol stack information library into the internal topology generation module. Subsequently, the internal topology generation module determines the type of protocol stack that has failed on the corresponding edge based on the faulty network element connected to each edge in the external topology and the fault event on the faulty network element, and determines the target protocol stack, the multi-layer protocol layers of the target protocol stack, and the hierarchical structure of the multi-layer protocol layers corresponding to the type of protocol stack that has failed from the protocol stack information library to construct the internal topology of each faulty network element. It can be understood that the internal topology is used to represent the multi-layer protocol layers of the target protocol stack where the corresponding faulty network element has failed and the hierarchical structure of the multi-layer protocol layers. Through the multi-layer protocol layers of each faulty network element and the hierarchical structure of the multi-layer protocol layers, the internal propagation path corresponding to each faulty network element can be determined.

[0055] In other embodiments, the fault detection system may also only input the external topology and device data into the internal topology generation module. The internal topology generation module may determine the target protocol stack of each faulty network element based on the machine learning model to construct the internal topology of each faulty network element, etc., without specific limitation.

[0056] In some embodiments, after obtaining the internal topology, the internal topology generation module inputs the internal topology into the detailed topology generation module. Subsequently, the detailed topology generation module determines, based on the target protocol stacks corresponding to the multiple faulty network elements, the preset interconnection relationships between the corresponding target protocol stacks between the multiple faulty network elements from the multiple protocol stacks included in the edges corresponding to the external topology, and connects the internal topologies corresponding to each faulty network element to construct a detailed topology corresponding to the multiple faulty network elements. It will be understood that the detailed topology is used to represent the connection relationship between the multiple faulty network elements connected via the corresponding protocol layers, and the connection relationship between the multiple faulty network elements via the corresponding protocol layers can be used to determine the external propagation paths corresponding to the multiple faulty network elements.

[0057] In some embodiments, after obtaining the detailed topology, the detailed topology generation module inputs the detailed topology into the event calibration module. Subsequently, the event calibration module determines the protocol layer (i.e., the fault layer) corresponding to each fault event based on the fault event occurring in each faulty network element, and calibrates each fault event and the corresponding fault layer to obtain a calibrated detailed topology.

[0058] In some embodiments, after obtaining the calibrated detailed topology, the event calibration module inputs the calibrated detailed topology into the fault detection module. Subsequently, the fault detection module can obtain the fault propagation path for each fault event corresponding to each fault layer based on the internal topologies (internal propagation paths), the connection relationships between the internal topologies (external propagation paths), and the calibrated fault layers in the calibrated detailed topology, and output a detection result.

[0059] Some embodiments of the present application are explained below by taking the external topology, device data and protocol stack information library as examples of inputs to the fault detection system.

[0060] In conjunction with the example in Figure 1, Figure 3 shows a structural diagram of another trustworthy and explainable fault detection system, according to some embodiments of the present application, using external topology, device data, and a protocol stack information library as inputs to the fault detection system. As shown in Figure 3, the fault detection system includes an internal topology generation module, a detailed topology generation module, an event calibration module, and a fault detection module. The inputs to the fault detection system are the external topology, device data, and the protocol stack information library, and the outputs of the fault detection system are the calibrated detailed topology and detection results.

[0061] In some embodiments, as shown in Figures 1 and 3, the vertices in the external topology are used to represent the dynamic ring network element A, the wireless network element B, and the transmission network element C, respectively; the edges between the dynamic ring network element A and the wireless network element B in the external topology are used to represent multiple protocol stacks between the dynamic ring network element A and the wireless network element B, the edges between the dynamic ring network element A and the transmission network element C are used to represent multiple protocol stacks between the dynamic ring network element A and the transmission network element C, and the edges between the wireless network element B and the transmission network element C are used to represent multiple protocol stacks between the wireless network element B and the transmission network element C.

[0062] Furthermore, in the device data, alarm6@network element A is used to indicate the fault event information of the DC undervoltage alarm occurring in the dynamic ring network element A; alarm1@network element B, alarm2@network element B, alarm3@network element B, and alarm4@network element B are used to indicate the fault event information of the IP clock link abnormality alarm, UMTS cell unavailable, Ethernet link failure alarm, and network element connection interruption occurring in the wireless network element B; alarm5@network element C is used to indicate the fault event information of the ETH LOS occurring in the transmission network element C.

[0063] Furthermore, the protocol information includes the multiple protocol layers corresponding to the multiple protocol stacks between the dynamic ring network element A, the wireless network element B and the transmission network element C, and the hierarchical structure of the multiple protocol layers. Taking the wireless network element B as an example, the multiple protocol stacks between the wireless network element B and the transmission network element C or the dynamic ring network element A include the Xn protocol stack, the dynamic ring protocol stack, and the service protocol stack. Among them, the Xn protocol stack includes the protocol layers of the physical layer, ETH, MAC, IP, and SCTP, and the hierarchical structure of the physical layer, ETH, MAC, IP, and SCTP corresponding to the wireless network element B is 11, 12, 13, 14, and 17 in sequence; the dynamic ring protocol stack includes the protocol layers of the dynamic ring layer and the network element layer, and the hierarchical structure of the dynamic ring layer and the network element layer is 0 (the bottom layer) and 1 in sequence; the service protocol stack includes the protocol layers of the service layer, and the hierarchical structure of the service layer is 1000 (the highest layer).

[0064] The following is an example of the implementation process of each module in Figure 3.

[0065] 1. Internal topology generation module.

[0066] In some embodiments, the fault detection system inputs the external topology, device data, and protocol stack information library into the internal topology generation module. The internal topology generation module determines the protocol stack type of the corresponding fault event between the dynamic ring network element A, the wireless network element B, and the transmission network element C based on the fault event information corresponding to the dynamic ring network element A, the wireless network element B, and the transmission network element C. Based on the protocol stack type, the module determines the corresponding target protocol stack from the protocol stack information library and constructs the internal topology of each network element based on the target protocol stack. For example, as shown in FIG4A , based on the fault event information of the IP clock link abnormality alarm (alarm1@network element B), UMTS cell unavailable alarm (alarm2@network element B), Ethernet link failure alarm (alarm3@network element B), and network element connection interruption alarm (alarm4@network element B) corresponding to wireless network element B, and the fault event information of the ETH LOS alarm (alarm5@network element C) corresponding to transmission network element C, through word segmentation, term frequency-inverse document frequency (TF-IDF), and logistic regression (LR) processing and analysis using a classification model, it can be determined that the target protocol stack where the fault occurs between wireless network element B and transmission network element C is the Xn protocol stack of the Xn interface category. Based on the multi-layer protocol layers and hierarchical structure of the Xn protocol stack in the protocol information, it can be determined that wireless network element B includes the physical layer, ETH, MAC, IP, and SCTP, and transmission network element C includes the physical layer, ETH, MAC, and IP.

[0067] In some embodiments, the preset protocol stack of the faulty network element can also be determined based on the service system scenario where the faulty network element is located. For example, as shown in Figure 4B, in the wireless communication service system scenario, the preset protocol stack of wireless network element B may include a dynamic ring protocol stack and a service protocol stack. In this way, based on the Xn protocol stack, dynamic ring protocol stack and service protocol stack of wireless network element B, the internal topology of wireless network element B can be obtained from bottom to top as follows: dynamic ring layer, network element layer, physical layer, ETH, MAC, IP, SCTP, service layer. Correspondingly, the internal topology of dynamic ring network element A only includes the dynamic ring layer, and the internal topology of transmission network element C is from bottom to top as follows: dynamic ring layer, network element layer, physical layer, ETH, MAC, IP.

[0068] 2. Detailed topology generation module.

[0069] In some embodiments, after obtaining the internal topology, the internal topology generation module inputs the internal topology into the detailed topology generation module. The detailed topology generation module determines, based on the target protocol stacks corresponding to the dynamic ring network element A, the wireless network element B, and the transmission network element C, the preset interconnection relationships between the corresponding target protocol stacks between the network elements from the multiple protocol stacks included in the corresponding edges of the external topology, and connects the corresponding internal topologies to construct the detailed topology corresponding to the dynamic ring network element A, the wireless network element B, and the transmission network element C. For example, the wireless network element B and the transmission network element C are connected via the physical layers corresponding to the two network elements, the wireless network element B and the dynamic ring network element A are connected via the dynamic ring layers corresponding to the two network elements, and the transmission network element C and the dynamic ring network element A are connected via the dynamic ring layers corresponding to the two network elements. In this way, as shown in Figure 5, the physical layer in the internal topology of the wireless network element B is connected to the physical layer in the internal topology of the transmission network element C via the Xn interface, the dynamic ring layer in the internal topology of the wireless network element B is connected to the dynamic ring layer in the internal topology of the dynamic ring network element A, and the dynamic ring layer in the internal topology of the transmission network element C is connected to the dynamic ring layer in the internal topology of the dynamic ring network element A, to obtain the detailed topology of the dynamic ring network element A, the wireless network element B and the transmission network element C.

[0070] 3. Event calibration module.

[0071] In some embodiments, after obtaining the detailed topology, the detailed topology generation module inputs the detailed topology into the event calibration module. The event calibration module determines the fault layer in the multiple protocol layers of the dynamic ring network element A, the wireless network element B, and the transmission network element C based on the fault events that occurred in the dynamic ring network element A, the wireless network element B, and the transmission network element C, and calibrates each fault event and the fault layer corresponding to each fault event to obtain a calibrated detailed topology. For example, as shown in Figure 6, the event calibration module labels the fault events of the Ethernet link failure alarm, network element connection interruption, UMTS cell unavailable, and IP clock link abnormal alarm occurring in the wireless network element B as "Alarm No. 3", "Alarm No. 4", "Alarm No. 2", and "Alarm No. 1" respectively, and then inputs the fault events calibrated by the wireless network element B and the internal topology of the wireless network element B into a text classifier such as a recurrent neural network (RNN). It can be determined that the fault layers corresponding to the fault events of the Ethernet link failure alarm, network element connection interruption, UMTS cell unavailable, and IP clock link abnormal alarm occurring in the wireless network element B are: network element layer, ETH, IP, and service layer respectively. The ETH, network element layer, service layer, and IP layer corresponding to the wireless network element B as the fault layer are labeled 3, 4, 2, and 1. Correspondingly, if the ETH LOS fault event occurring in transmission NE C is labeled "Alarm 5," the ETH layer in transmission NE C, which is the fault layer, is labeled 5. If the DC undervoltage alarm fault event occurring in dynamic ring NE A is labeled "Alarm 6," the dynamic ring layer, which is the fault layer in dynamic ring NE A, is labeled 6. This results in the calibrated detailed topology shown in Figure 7.

[0072] 4. Fault detection module.

[0073] In some embodiments, after the event calibration module obtains the calibrated detail topology, it inputs the calibrated detail topology into the fault detection module. The fault detection module inputs the calibrated detail topology into the generative large language model (such as GPT-3, BERT, T5). The generative large language model can obtain the fault propagation path of each fault event corresponding to each fault layer based on each internal topology (internal propagation path) in the calibrated detail topology, the connection relationship of each internal topology (external propagation path) and the calibrated fault layer, and output the detection result. For example, as shown in Figure 8, the detection results may include: the cause of the failure of the service layer in the wireless network element B is: "Alarm No. 4" occurs at the network element layer of the wireless network element B, which in turn causes "Alarm No. 3" to occur at the ETH layer of the wireless network element B, "Alarm No. 1" to occur at the IP layer, and "Alarm No. 2" to occur at the service layer, that is, the corresponding fault propagation path is 4→3→1→2; the cause of the associated fault between the dynamic ring network element A, the wireless network element B and the transmission network element C is: "Alarm No. 4" occurs at the ETH layer of the wireless network element B. "Alarm No. 3" causes "Alarm No. 5" to occur at the ETH layer of transmission network element C, and the corresponding fault propagation path is 3→5; "Alarm No. 6" occurs at the dynamic ring layer of dynamic ring device A, causing "Alarm No. 4" to occur at the network element layer of wireless network element B, and the corresponding fault propagation path is 6→4. Therefore, based on the fault propagation paths 4→3→1→2, 3→5, and 6→4, the root cause network element of the fault can be determined to be: dynamic ring network element A; the root cause protocol layer is: dynamic ring layer, and the root cause fault event is "Alarm No. 6".

[0074] In this way, operation and maintenance personnel can perform inspection and maintenance in sequence according to the fault propagation paths of multiple fault events to ensure inspection and maintenance efficiency.

[0075] The following describes some other embodiments of the present application by taking external topology and device data as an example of input to the fault detection system.

[0076] In conjunction with the example of Figure 1, Figure 9 shows a structural diagram of another credible and explainable fault detection system according to some embodiments of the present application, with external topology, device data and protocol stack information library as the input of the fault detection system. As shown in Figure 9, the fault detection system includes an internal topology generation module, a detailed topology generation module, an event calibration module and a fault detection module, and the input of the fault detection system is the external topology and device data, and the output of the fault detection system is the calibrated detailed topology and detection results. It can be understood that the system shown in Figure 9 is different from the system shown in Figure 3 in that only the system input and the internal topology generation module are implemented differently. The system shown in Figure 9 is the same as the system shown in Figure 3 in that the detailed topology generation module, event calibration module, fault detection module and the output of the system are the same, so the following only explains the implementation process of the system input and internal topology generation module in Figure 9.

[0077] In some embodiments, the external topology and device data input by the system of FIG9 can refer to the relevant description of the external topology and device data in FIG3 , and will not be repeated here.

[0078] In some embodiments, as shown in Figures 9 and 10, after the fault detection system inputs the external topology and device data into the internal topology generation module, the internal topology generation module can input the external topology and device data into an internal topology generation model (e.g., DL) within the internal topology generation module. The internal topology generation model is used to determine the target protocol stack, multiple protocol layers of the target protocol stack, and the hierarchical structure of the multiple protocol layers of the target protocol stack for the corresponding fault events between the dynamic ring network element A, the wireless network element B, and the transmission network element C, and then output the internal topology corresponding to the dynamic ring network element A, the wireless network element B, and the transmission network element C. It will be understood that the internal topology generation model can be determined based on multiple network elements and different fault events corresponding to the multiple network elements as a training set.

[0079] The following describes some embodiments of the present application in combination with the description of Figures 3 to 8, taking a wireless communication scenario as an example.

[0080] Figure 11 shows a flow chart of a fault detection method according to some embodiments of the present invention. It is understood that each step of the process shown in Figure 11 is performed by the fault detection system shown in Figure 3 . To simplify the description, the following description of the execution entity of each step of the process shown in Figure 11 will not be repeated. As shown in Figure 11 , the fault detection method includes the following steps:

[0081] S111: Determine a plurality of fault events having an associated relationship and a plurality of faulty network elements corresponding to the plurality of fault events.

[0082] In some embodiments, when a fault detection system detects multiple fault events occurring in a wireless communication service system, it determines whether the multiple fault events are associated. For example, if the time difference between the multiple fault occurrence times corresponding to the multiple fault events meets a preset time difference threshold, it can be determined that the multiple fault occurrence times meet a preset time correlation, and further, the multiple fault events are associated. For another example, if the multiple faulty network elements corresponding to the multiple fault events meet a preset network element correlation, it can be determined that the multiple fault events are associated. Or, for example, if the multiple event types corresponding to the multiple fault events meet a preset type correlation, it can be determined that the multiple fault events are associated.

[0083] In some embodiments, after the fault detection system determines that multiple fault events have a correlation, it determines the multiple fault events and the fault information of the multiple fault network elements corresponding to the multiple fault events. For example, the fault information is shown in Figure 1. Subsequently, the fault detection system determines the external topology, device data, and protocol stack information library based on the determined fault information. For example, as shown in Figure 3, the external topology includes three vertices of the fault network elements of the dynamic ring network element A, the wireless network element B, and the transmission network element C, and three edges that carry multiple protocol stacks for mutual communication between the dynamic ring network element A, the wireless network element B, and the transmission network element C; the device data includes the fault event information corresponding to the fault events occurring in the dynamic ring network element A, the wireless network element B, and the transmission network element C; the protocol stack information library includes the multiple protocol layers corresponding to the multiple protocol stacks between the dynamic ring network element A, the wireless network element B, and the transmission network element C, and the hierarchical structure of the multiple protocol layers.

[0084] In other embodiments, step S111 may also be performed by a wireless communication service system. For example, after the service system detects multiple fault events within the system and determines that the multiple fault events are related, it determines the multiple faulty network elements corresponding to the multiple fault events. Subsequently, the service system sends the multiple fault events and the fault information of the multiple faulty network elements corresponding to the multiple fault events to the fault detection system. For example, the service system sends the fault information shown in Figure 1 to the fault detection system.

[0085] S112: Determine a target protocol stack between the multiple faulty network elements based on a preset mutual communication relationship between the multiple faulty network elements corresponding to the faulty events.

[0086] In some embodiments, the fault detection system inputs the external topology, device data, and protocol stack information library into the internal topology generation module. The fault detection system, through the internal topology generation module, determines the target protocol stack between multiple faulty network elements based on the preset intercommunication relationships corresponding to the fault events between the multiple faulty network elements. For example, as shown in FIG4A , the internal topology generation module can determine that the target protocol stack for the fault between wireless network element B and transmission network element C is an Xn protocol stack of the Xn interface category based on fault event information corresponding to the fault event of wireless network element B and the fault event information corresponding to the fault event of transmission network element C, by performing word segmentation, TF-IDF, and LR processing and analysis using a classification model.

[0087] S113: Based on the target protocol stack of each faulty network element, determine the multi-layer protocol layer and the hierarchical structure of each faulty network element, and determine the internal propagation path of the corresponding faulty network element based on the hierarchical structure of the multi-layer protocol layer.

[0088] In some embodiments, the fault detection system, through an internal topology generation module, determines the multiple protocol layers and hierarchical structure of each faulty network element based on the target protocol stack of each faulty network element, and determines the internal propagation path of the corresponding faulty network element based on the hierarchical structure of the multiple protocol layers. For example, as shown in FIG4B , the target protocol stack of wireless network element B includes: an Xn protocol stack, a dynamic ring protocol stack, and a service protocol stack. The internal topology generation module can obtain from the protocol stack information library the multiple protocol layers of the Xn protocol stack, the dynamic ring protocol stack, and the service protocol stack corresponding to wireless network element B, which include, from bottom to top, the dynamic ring layer, the network element layer, the physical layer, ETH, MAC, IP, SCTP, and the service layer. The internal topology of wireless network element B is determined based on the multiple protocol layers and hierarchical structure of the multiple protocol layers included in wireless network element B, and the internal propagation path of wireless network element B is determined based on the internal topology of wireless network element B. Accordingly, the internal topology generation module determines the internal propagation paths of ring network element A and transmission network element C based on the internal topologies of dynamic ring network element A and transmission network element C, respectively. For details, please refer to the relevant description of the internal topology generation module above, which is not repeated here.

[0089] S114: Based on the preset interconnection relationship between the multiple faulty network elements corresponding to the target protocol stack, determine the external propagation path connecting the multiple faulty network elements through the corresponding protocol layer.

[0090] In some embodiments, after the internal topology generation module obtains the internal topology, it inputs the internal topology into the detailed topology generation module. The fault detection system, through the detailed topology generation module, determines, based on the preset interconnected relationships between the target protocol stacks corresponding to the multiple faulty network elements, the external propagation paths connecting the multiple faulty network elements via corresponding protocol layers. For example, as shown in Figure 5, the detailed topology generation module can determine, based on the Xn protocol stack between wireless network element B and transmission network element C, that wireless network element B and transmission network element C are connected via the physical layers corresponding to the two network elements. Specifically, the physical layer in the internal topology of wireless network element B is connected to the physical layer in the internal topology of transmission network element C via the Xn interface. Correspondingly, the detailed topology generation module connects the dynamic ring layer in the internal topology of the wireless network element B with the dynamic ring layer in the internal topology of the dynamic ring network element A, and connects the dynamic ring layer in the internal topology of the transmission network element C with the dynamic ring layer in the internal topology of the dynamic ring network element A, and obtains the detailed topology of the dynamic ring network element A, the wireless network element B and the transmission network element C. The external propagation path between the dynamic ring network element A, the wireless network element B and the transmission network element C is determined through the detailed topology. For details, please refer to the relevant description of the detailed topology generation module above, which will not be repeated here.

[0091] S115: Based on the corresponding fault events occurring in the respective faulty network elements, determine the faulty layer in the multi-layer protocol layers of the respective faulty network elements.

[0092] In some embodiments, after the detailed topology generation module obtains the detailed topology, the detailed topology is input into the event calibration module. The fault detection system determines the fault layer in the multi-layer protocol layer of each corresponding fault network element based on the corresponding fault events that occurred in each fault network element through the event calibration module. For example, as shown in FIG6 , the event calibration module performs corresponding calibration on the fault events that occurred in the dynamic ring network element A, the wireless network element B, and the transmission network element C, and then inputs the corresponding calibration into the text classifier to determine the fault layer in the dynamic ring network element A, the wireless network element B, and the transmission network element C. The fault layer is calibrated by the calibration of the corresponding fault events, and the calibrated detailed topology as shown in FIG7 is obtained. For details, please refer to the relevant description of the above-mentioned event calibration module, which will not be repeated here.

[0093] S116: Determine, based on the internal propagation path and the external propagation path, a fault propagation path between multiple fault events corresponding to the multiple fault layers.

[0094] In some embodiments, after the event calibration module obtains the calibrated detailed topology, the calibrated detailed topology is input into the fault detection module. The fault detection system inputs the calibrated detailed topology into the generative large language model through the fault detection module. The generative large language model determines the fault propagation path between multiple fault events corresponding to multiple fault layers based on the internal propagation path and external propagation path in the calibrated detailed topology, and outputs the detection result. For example, as shown in Figure 8, the detection result may include: the cause of the fault in the service layer of wireless network element B, the cause of the associated fault between the dynamic ring network element A, the wireless network element B and the transmission network element C, the root cause network element, the root cause protocol layer, and the root cause fault event. For details, please refer to the relevant description of the fault detection module above, which will not be repeated here.

[0095] In this way, operation and maintenance personnel can perform inspection and maintenance in sequence according to the fault propagation paths of multiple fault events to ensure inspection and maintenance efficiency.

[0096] FIG12 shows a structural diagram of a fault detection device 10 according to some embodiments of the present invention. As shown in FIG12 , the fault detection device 110 includes: an internal topology generation unit 1101 , a detailed topology generation unit 1102 , an event calibration unit 1103 , and a fault detection unit 1104 .

[0097] In some embodiments, the internal topology generation unit 1101 is used to determine the internal topology and internal propagation path of each faulty network element corresponding to the fault event. For details, please refer to the above steps S112 and S113 and the description of the internal topology generation module, which will not be repeated here.

[0098] In some embodiments, the detailed topology generating unit 1102 is used to determine the detailed topology and external propagation path of the fault events corresponding to the multiple faulty network elements. Detailed descriptions refer to the above step S114 and the related descriptions of the detailed topology generating module, which will not be repeated here.

[0099] In some embodiments, the event calibration unit 1103 is configured to determine the fault layer in the multiple protocol layers corresponding to each faulty network element based on the fault events occurring in each faulty network element, and to calibrate the fault layer based on the fault events corresponding to the fault layer to obtain a calibrated detailed topology. For details, see the above-mentioned step S115 and the related description of the event calibration module, which will not be repeated here.

[0100] In some embodiments, the fault detection unit 1104 is configured to determine the fault propagation paths between multiple fault events corresponding to multiple fault layers based on the internal propagation paths and external propagation paths in the calibrated detailed topology, and output a detection result. For details, see step S116 and the related description of the fault detection module above, which will not be repeated here.

[0101] In some embodiments, the embodiments of the present application further provide a computer-readable medium, which stores program code. When the computer program code runs on a computer, the computer executes the methods in the above aspects.

[0102] In some embodiments, the embodiments of the present application further provide a computer program product, which includes: computer program code, which enables the computer to execute the methods in the above aspects when the computer program code is run on a computer.

[0103] In the accompanying drawings, some structural or method features may be shown in a particular arrangement and / or order. However, it should be understood that such a particular arrangement and / or order may not be required. Rather, in some embodiments, these features may be arranged in a manner and / or order different from that shown in the illustrative drawings. In addition, the inclusion of a structural or method feature in a particular figure does not imply that such feature is required in all embodiments, and in some embodiments, such features may not be included or may be combined with other features.

[0104] It should be noted that the various units / modules mentioned in the various device embodiments of the present invention are all logical units / modules. Physically, a logical unit / module can be a physical unit / module, or a part of a physical unit / module, or can be implemented as a combination of multiple physical units / modules. The physical implementation of these logical units / modules themselves is not the most important. The combination of functions implemented by these logical units / modules is the key to solving the technical problems raised by the present invention. In addition, in order to highlight the innovative part of the present invention, the above-mentioned device embodiments of the present invention do not introduce units / modules that are not closely related to solving the technical problems raised by the present invention. This does not mean that other units / modules do not exist in the above-mentioned device embodiments.

[0105] It should be noted that in the examples and description of this patent, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "including a" does not exclude the presence of other identical elements in the process, method, article or device that includes the element.

[0106] While the present invention has been shown and described with reference to certain preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention.

Claims

1. A fault detection method, characterized in that: include: Determine a plurality of fault events having an associated relationship, and a plurality of faulty network elements corresponding to the plurality of fault events; Determining a target protocol stack between the plurality of faulty network elements based on a preset mutual communication relationship between the plurality of faulty network elements corresponding to the faulty events; Determine the multi-layer protocol layer of each of the faulty network elements based on the target protocol stack of each of the faulty network elements, and determine the internal propagation path of the corresponding faulty network element according to the hierarchical structure of the multi-layer protocol layer; Based on the preset interconnection relationship between the multiple faulty network elements corresponding to the target protocol stack, determine the external propagation path for connecting the multiple faulty network elements through the target protocol layer in the target protocol stack, wherein the target protocol layer for connecting each of the faulty network elements is a protocol layer in the multiple protocol layers corresponding to each of the faulty network elements; Based on the corresponding fault events occurring in each of the faulty network elements, determining the faulty layer in the multi-layer protocol layers corresponding to each of the faulty network elements; Based on the internal propagation path and the external propagation path, a fault propagation path between multiple fault events corresponding to multiple fault layers is determined.

2. The method according to claim 1, characterized in that The determining the internal propagation path of the corresponding faulty network element according to the hierarchical structure of the multi-layer protocol layer includes: Determine the internal topology of the corresponding faulty network element based on the multiple protocol layers of the faulty network element and the hierarchical structure of the multiple protocol layers; Based on the internal topology of the faulty network element, the internal propagation path of the faulty network element in the multi-layer protocol layer corresponding to the faulty network element is determined.

3. The method according to claim 2, characterized in that The determining, based on the preset interconnection relationship between the plurality of faulty network elements corresponding to the target protocol stack, an external propagation path for connecting the plurality of faulty network elements through a target protocol layer in the target protocol stack, comprises: Based on the preset interconnection relationship between the multiple faulty network elements corresponding to the target protocol stack, the multiple internal topologies corresponding to the multiple faulty network elements are connected through the target protocol layer of the corresponding target protocol stack to form a detailed topology; Based on the target protocol layer connected with the plurality of internal topologies in the detailed topology, external propagation paths of the plurality of faulty network elements are determined.

4. The method according to claim 3, characterized in that The determining, based on the corresponding fault events occurring in the respective faulty network elements, the faulty layer in the multi-layer protocol layers corresponding to the respective faulty network elements comprises: Labeling each of the fault events; Based on the corresponding fault events occurring in each of the faulty network elements and the event numbers corresponding to the fault events, determining the faulty layers in the multi-layer protocol layers corresponding to each of the faulty network elements and the layer numbers of the faulty layers; Based on the layer label and the event label, a corresponding relationship between the fault layer and the fault event is determined.

5. The method according to claim 4, characterized in that The determining, based on the internal propagation path and the external propagation path, the fault propagation paths between the multiple fault events corresponding to the multiple fault layers includes: Based on the internal propagation path and the external propagation path, determining the hierarchical propagation paths of the multiple layers of the fault layers corresponding to the multiple faulty network elements in the multiple layers of the protocol layers of the multiple faulty network elements; Based on the hierarchical propagation paths and the corresponding relationship between the fault layers and the fault events, fault propagation paths of the corresponding multiple fault events in the multi-layer protocol layers of the multiple faulty network elements are determined.

6. The method according to any one of claims 1 to 5, characterized in that: The determining of a target protocol stack between the plurality of faulty network elements comprises: Determine a preset protocol stack set for mutual communication between the plurality of faulty network elements, wherein the protocol stack set includes a plurality of protocol stacks for mutual communication between the plurality of faulty network elements; Based on the preset mutual communication relationship between the multiple faulty network elements corresponding to the fault events, the target protocol stack corresponding to the multiple faulty network elements is determined from the multiple protocol stacks.

7. The method according to any one of claims 1 to 5, characterized in that: The determining of a plurality of fault events having an associated relationship includes: Determine a plurality of fault occurrence times corresponding to the plurality of fault events, and corresponding to the plurality of fault occurrence times satisfying a preset time association relationship, determine that the plurality of fault events have an association relationship; or, Determine the multiple faulty network elements corresponding to the multiple fault events, corresponding to the multiple faulty network elements satisfying a preset network element association relationship, and determine that the multiple fault events have an association relationship; or, A plurality of event types corresponding to the plurality of fault events are determined, and corresponding to the plurality of event types satisfying a preset type association relationship, it is determined that the plurality of fault events have an association relationship.

8. A fault detection system, characterized in that: include: An internal topology generation module, based on a preset mutual communication relationship between the multiple faulty network elements corresponding to the fault events, determines a target protocol stack between the multiple faulty network elements; Determine the multi-layer protocol layer of each of the faulty network elements based on the target protocol stack of each of the faulty network elements, and determine the internal propagation path of the corresponding faulty network element according to the hierarchical structure of the multi-layer protocol layer; A detailed topology generation module, based on a preset interconnection relationship between the plurality of faulty network elements corresponding to the target protocol stack, determines an external propagation path for connecting the plurality of faulty network elements through a target protocol layer in the target protocol stack, wherein the target protocol layer for connecting each of the faulty network elements is a protocol layer in the multi-layer protocol layer corresponding to each of the faulty network elements; An event calibration module, based on the corresponding fault events occurring in each of the faulty network elements, determines the fault layer in the multi-layer protocol layers corresponding to each of the faulty network elements; The fault detection module determines the fault propagation paths between the multiple fault events corresponding to the multiple fault layers based on the internal propagation paths and the external propagation paths.

9. The system according to claim 8, characterized in that The internal topology generation module determines the internal propagation path of the corresponding faulty network element according to the hierarchical structure of the multi-layer protocol layer, including: The internal topology generation module determines the internal topology of the corresponding faulty network element based on the multi-layer protocol layers of the faulty network element and the hierarchical structure of the multi-layer protocol layers; and determines the internal propagation path of the faulty network element in the multi-layer protocol layers of the corresponding faulty network element based on the internal topology of the faulty network element.

10. The system according to claim 9, characterized in that The internal topology generation module determines, based on a preset interconnection relationship between the multiple faulty network elements corresponding to the target protocol stack, an external propagation path for connecting the multiple faulty network elements through a target protocol layer in the target protocol stack, including: The detailed topology generation module connects the multiple internal topologies corresponding to the multiple faulty network elements through the target protocol layer of the corresponding target protocol stack based on the preset interconnection relationship between the multiple faulty network elements to form a detailed topology; and determines the external propagation paths of the multiple faulty network elements based on the target protocol layer at which the multiple internal topologies in the detailed topology are connected.

11. The system according to claim 10, characterized in that The event calibration module determines the fault layer in the multi-layer protocol layer corresponding to each of the faulty network elements based on the corresponding fault events occurring in each of the faulty network elements, including: The event labeling module labels each of the fault events; based on the corresponding fault events occurring in each of the faulty network elements and the event labels corresponding to the fault events, determines the fault layer in the multi-layer protocol layers corresponding to each of the faulty network elements and the layer label of the fault layer; Based on the layer label and the event label, a corresponding relationship between the fault layer and the fault event is determined.

12. The system according to claim 11, characterized in that The fault detection module determines the fault propagation paths between the multiple fault events corresponding to the multiple fault layers based on the internal propagation paths and the external propagation paths, including: The fault detection module determines the hierarchical propagation paths of the multiple layers of the fault layers corresponding to the multiple faulty network elements in the multi-layer protocol layers of the multiple faulty network elements based on the internal propagation path and the external propagation path; and determines the fault propagation paths of the corresponding multiple fault events in the multi-layer protocol layers of the multiple faulty network elements based on the hierarchical propagation paths and the correspondence between the fault layers and the fault events.

13. The system according to any one of claims 8 to 12, characterized in that: The internal topology generation module determines a target protocol stack between the plurality of faulty network elements, including: The internal topology generation module determines a preset set of protocol stacks for mutual communication between the multiple faulty network elements, wherein the protocol stack set includes multiple protocol stacks for mutual communication between the multiple faulty network elements; based on the preset mutual communication relationship corresponding to the fault events between the multiple faulty network elements, the target protocol stack corresponding to the multiple faulty network elements is determined from the multiple protocol stacks.

14. A computer-readable storage medium, characterized in that: The readable storage medium stores instructions, which, when executed on a device, enable the device to implement the method according to any one of claims 1 to 7.

Images