Intelligent prediction method of network equipment port fault based on deep learning

Abstract

The intelligent prediction method of network equipment port failure based on deep learning includes the following steps: A), input data source; B), conduct intelligent analysis of big data by the intelligent analysis system of big data, match port alarm information and summarize resource usage information; C ), analyze the results, first build a deep learning training model, then perform offline training and online training, and finally perform the model effect operation output; D), output port intelligent early warning information. The invention improves the intelligent operation and maintenance capability of network equipment, realizes the development of network equipment operation and maintenance from "after the event" to "before the event", truly realizes "prevention" of network equipment operation and maintenance, and has good economic benefits for popularization and application benefits and social benefits.

Description

technical field [0001] The invention belongs to the technical field of telecommunications, relates to network equipment operation and maintenance, and in particular to an intelligent prediction method for network equipment port failures based on deep learning. Background technique [0002] The traditional operation and maintenance mode of network equipment is: after the network management system discovers the equipment alarm, it notifies the maintenance personnel to repair it. System operation and maintenance personnel spend most of their time and energy dealing with some simple and repetitive problems. The physical labor is too heavy, the work efficiency is low, and a large amount of maintenance resources need to be invested; there are the following technical defects: the network management system of traditional operation and maintenance of network equipment does not have Intelligent fault analysis function, when the business is abnormal, maintenance personnel need to spend...

AI Technical Summary

Problems solved by technology

[0002] The traditional operation and maintenance mode of network equipment is: after the network management system detects equipment alarms, it notifies the maintenance personnel to carry out maintenance, which is a manual repair after the event, and the fault response time is long, which cannot meet the high real-time business requirements

System operation and maintenance personnel spend most of their time and energy dealing with some simple and repetitive problems. The physical labor is too heavy, the work efficiency is low, and a large amount of maintenance resources need to be invested; there are the following technical defects: the network management system of traditional operation and maintenance of network equipment does not have Intelligent fault analysis function, when the business is abnormal, maintenance personnel need to spend a lot of manpower to check the fault and find the cause of the fault, the system maintenance efficiency is low, and it does not have the intelligent fault early warning function. more passive

Images