System and method for determining equipment failure

Abstract

A system and method for determining equipment faults are provided. The method includes the following steps: performing a preprocessing operation on the monitoring data of at least one adjacent device to obtain preprocessed monitoring data of the adjacent devices, wherein the at least one adjacent device is adjacent to a target device; and calculating the information entropy corresponding to the at least one adjacent device using the preprocessed monitoring data of the adjacent devices, and using the information entropy to determine the fault determination result corresponding to the target device.

Description

Technical Field

[0001] This invention relates to a system and method for determining equipment malfunctions. Prior Technology

[0002] Currently, when network administrators / communication service providers (CSPs) use diagnostic commands to collect device data for fault detection or root cause analysis, network administrators often lack sufficient information to determine the true state of network services. Summary of the Invention

[0003] The system for determining equipment faults according to the present invention includes a monitoring data preprocessing module and a fault root cause determination module. The monitoring data preprocessing module performs preprocessing operations on the monitoring data of at least one adjacent device to obtain preprocessed adjacent device monitoring data, wherein at least one adjacent device is adjacent to the target device; the fault root cause determination module uses the preprocessed adjacent device monitoring data to calculate the information entropy corresponding to at least one adjacent device, and uses the information entropy to determine the fault determination result corresponding to the target device.

[0004] The method for determining equipment faults according to the present invention is suitable for a system including a monitoring data preprocessing module and a fault root cause determination module, wherein the method includes the following steps: the monitoring data preprocessing module performs a preprocessing operation on the monitoring data of at least one adjacent device to obtain preprocessed adjacent device monitoring data, wherein the at least one adjacent device is adjacent to a target device; and the fault root cause determination module uses the preprocessed adjacent device monitoring data to calculate the information entropy corresponding to the at least one adjacent device, and the fault root cause determination module uses the information entropy to determine the fault determination result corresponding to the target device. Simple Explanation of the Diagram

[0005] Figure 1 is a schematic diagram of a system for determining equipment faults according to an embodiment of the present invention. Figure 2 is a flowchart illustrating a method for determining equipment failure according to an embodiment of the present invention. Figure 3 is a schematic diagram of the target device and adjacent devices according to an embodiment of the present invention. Figure 4 is a schematic diagram illustrating a visualization according to an embodiment of the present invention. Implementation

[0006] Figure 1 is a schematic diagram of a system 100 for determining equipment faults according to an embodiment of the present invention. In this embodiment, the system 100 may include a monitoring data preprocessing module 110 and a fault root cause determination module 120. In one embodiment, the system 100 may include a visualization display module 130. In one embodiment, the monitoring data preprocessing module 110, the fault root cause determination module 120, and the visualization display module 130 may be software and / or firmware code executed by a processor.

[0007] Figure 2 is a flowchart illustrating a method for determining equipment failure according to an embodiment of the present invention, wherein the method can be implemented by the system 100 shown in Figure 1. Please refer to Figures 1 and 2 simultaneously.

[0008] In step S21, the monitoring data preprocessing module 110 can perform preprocessing operations on the monitoring data of at least one adjacent device to obtain preprocessed monitoring data of the adjacent devices, wherein at least one adjacent device is adjacent to the target device. In one embodiment, when the target device experiences an expired query command, the monitoring data preprocessing module 110 can obtain the monitoring data of the adjacent devices of at least one adjacent device from the monitoring data database 300. For example, network administrators can operate the fault detection system 200 to send a query command (e.g., a PING command) to the target device. When the target device experiences an expired query command (e.g., a PING command timeout), the monitoring data preprocessing module 110 can obtain the monitoring data of the adjacent devices of at least one adjacent device from the monitoring data database 300. In one embodiment, the preprocessing operations may include missing value imputation, duplicate deletion, and normalization.

[0009] In step S22, the fault root cause determination module 120 can use the preprocessed adjacent device monitoring data to calculate the information entropy corresponding to at least one adjacent device, and the fault root cause determination module 120 can use the information entropy to determine the fault determination result corresponding to the target device.

[0010] In one embodiment, neighbor device monitoring data may include port status and alarm count. In another embodiment, the target device is circuit-connected to at least one neighbor device, wherein the neighbor device monitoring data may include traffic corresponding to the circuit, and the information entropy may include the traffic information entropy corresponding to the circuit. Specifically, the magnitude of the information entropy calculated for each feature (neighbor device monitoring data) can represent the degree of change of that feature over a specific time period. Device failure typically leads to non-linear changes in network characteristics. This invention can calculate the information entropy of features as a basis for quantifying such non-linear changes, and thus as an indicator for determining whether a device is faulty.

[0011] In one embodiment, when the fault root cause determination module 120 determines that the information entropy is greater than the normal standard value, the fault root cause determination module 120 can determine that the fault determination result indicates that the target device has failed. Specifically, if the target device fails, the monitoring data of adjacent devices will change drastically within a short period of time, causing the information entropy value to be greater than the normal standard value. On the other hand, when the fault root cause determination module 120 determines that the information entropy is less than or equal to the normal standard value, the fault root cause determination module 120 can determine that the fault determination result indicates that the target device has not failed. In other words, the problem may lie in the network management path.

[0012] Figure 3 is a schematic diagram illustrating a target device and adjacent devices according to an embodiment of the present invention. Please refer to Figures 1 and 3 simultaneously. Suppose that a network administrator operates the fault detection system 200 and finds that device A in Figure 3 does not respond to the detection command / Network Management System (NMS) command, or the network administrator cannot collect the traffic of device A. The network administrator can log in to the fault detection system 200 to start fault detection. In other words, device A is the aforementioned target device. Next, the fault detection system 200 can start the monitoring data preprocessing module 110. Then, the monitoring data preprocessing module 110 can obtain the monitoring data of the adjacent devices. Specifically, devices B and C shown in Figure 3 are adjacent devices. As shown in Figure 3, the target device (device A) is connected to the adjacent devices (device B and device C) by circuitry. Furthermore, device B and device A are connected by three circuits. On the other hand, device C and device A are connected by two circuits. In one embodiment, the monitoring data preprocessing module 110 can obtain the circuit flow of devices B and C over the past 30 minutes and perform preprocessing operations on the monitoring data of adjacent devices.

[0013] Next, the fault root cause determination module 120 can calculate the flow information entropy of each circuit adjacent to device B, device C, and device A every 5 minutes to obtain the connected circuit information entropy data as shown in Table 1.

[0014] Table 1 time Adjacent equipment Flow information entropy of circuit 1 Entropy of flow information in circuit 2 Flow information entropy of circuit 3 12 / 08 11:55 Equipment B 0.1 0.2 0.1 12 / 08 11:55 Equipment C 0.1 0.1 NULL 12 / 08 12:00 Equipment B 0.5 0.4 0.6 12 / 08 12:00 Equipment C 0.4 0.5 NULL 12 / 08 12:05 Equipment B 0.8 0.6 0.9 12 / 08 12:05 Equipment C 0.6 0.7 NULL 12 / 08 12:10 Equipment B 0.8 0.6 0.9 12 / 08 12:10 Equipment C 0.6 0.7 NULL 12 / 08 12:15 Equipment B 0.8 0.6 0.9 12 / 08 12:15 Equipment C 0.6 0.7 NULL

[0015] Then, the fault cause determination module 120 can count how many circuits in devices B, C, and adjacent circuits of device A have a flow information entropy greater than the normal standard value. Assume the normal standard value is 0.5. Since the flow information entropy of all circuits adjacent to devices B, C, and A is greater than the normal standard value after 12:05 on December 8th, the fault cause determination module 120 can determine that the target device (device A) has failed.

[0016] Figure 4 is a schematic diagram illustrating a visualization display according to an embodiment of the present invention. Please refer to Figures 1, 3, and 4 simultaneously. The visualization display module 130 can use a topological map to display the target device, at least one adjacent device, and the fault determination result. Continuing with the above embodiment, the visualization display module 130 can display the result shown in Figure 4.

[0017] In summary, the system and method for determining equipment faults of the present invention can perform preprocessing operations on the monitoring data of adjacent devices, and then calculate the information entropy of the adjacent devices. Next, the information entropy can be used to determine the fault determination result corresponding to the target device. Based on this, network administrators can efficiently determine whether a device fault or a problem exists in the network management path. In particular, network administrators can determine the root cause of the fault without going to the site or requiring additional equipment, thereby saving manpower maintenance and setup costs.

[0018] 100: Systems used to diagnose equipment malfunctions 110: Monitoring data preprocessing module 120: Fault Root Cause Analysis Module 130: Visualization Module 200: Obstacle Detection System 300: Monitoring Data Database S21, S22: Steps A, B, C, D, E, F, G, H: Equipment

Claims

1. A system for determining equipment malfunctions, comprising: Monitoring data preprocessing module; The system includes a fault root cause determination module, wherein the monitoring data preprocessing module performs preprocessing operations on the monitoring data of at least one adjacent device to obtain preprocessed monitoring data of the adjacent device, wherein the at least one adjacent device is adjacent to the target device; the fault root cause determination module uses the preprocessed monitoring data of the adjacent device to calculate the information entropy corresponding to the at least one adjacent device, and the fault root cause determination module uses the information entropy to determine the fault determination result corresponding to the target device; wherein the preprocessing operations include missing value imputation, duplicate deletion, and normalization; wherein the adjacent device monitoring data includes port status and alarm count; wherein the target device is connected to the at least one adjacent device by a circuit; wherein the adjacent device monitoring data includes the traffic corresponding to the circuit; and wherein the information entropy includes the traffic information entropy corresponding to the circuit.

2. The system as described in claim 1, wherein when the target device experiences an expired detection instruction, the monitoring data preprocessing module obtains the neighboring device monitoring data of the at least one neighboring device from the monitoring data database.

3. The system as described in claim 1, wherein when the fault root cause determination module determines that the information entropy is greater than the normal standard value, the fault root cause determination module determines that the fault determination result indicates that the target device has malfunctioned.

4. The system as described in claim 1, wherein when the fault root cause determination module determines that the information entropy is less than or equal to the normal standard value, the fault root cause determination module determines that the fault determination result indicates that the target device has not experienced a fault.

5. The system as claimed in claim 1 further includes a visualization display module, wherein the visualization display module uses a topological map to display the target device, the at least one adjacent device, and the fault determination result.

6. A method for determining equipment faults, suitable for a system including a monitoring data preprocessing module and a fault root cause determination module, wherein the method includes the following steps: the monitoring data preprocessing module performs preprocessing operations on the monitoring data of at least one adjacent device to obtain preprocessed adjacent device monitoring data, wherein the at least one adjacent device is adjacent to a target device; and the fault root cause determination module uses the preprocessed adjacent device monitoring data to calculate an information entropy corresponding to the at least one adjacent device, and the fault root cause determination module uses the information entropy to determine a fault determination result corresponding to the target device, wherein the preprocessing operations include missing value imputation, duplicate deletion, and normalization, wherein the adjacent device monitoring data includes port status and alarm count, wherein the target device is circuit-connected to the at least one adjacent device, wherein the adjacent device monitoring data includes traffic corresponding to the circuit, and wherein the information entropy includes traffic information entropy corresponding to the circuit.

Images