Network data grading method and device, network management server and storage medium

Abstract

The invention discloses a network data grading method and device, a network management server and a storage medium, and relates to the technical field of communication. The method comprises the following steps: determining an interval to which the offline duration of a target optical network terminal belongs in a preset grading evaluation range aiming at the suspected degree of junk data according to an offline time related parameter; determining a suspected degree level of network data and junk data generated by the target optical line terminal in the optical line terminals associated with the target optical network terminal according to the interval to which the target optical network terminal belongs, wherein the generation process of the grading evaluation range comprises the following steps of: performing curve fitting on coordinate points which take off-line duration of a historical optical network terminal as an abscissa and take the off-line number as an ordinate at a historical moment; according to a beta distribution three-point estimation algorithm and the fitted curve, generating an expected value and a standard deviation of the offline duration of the historical optical network terminal; and determining a plurality of different evaluation threshold values in the grading evaluation range according to the expected value, the standard deviation and preset different grade weight sets.

Description

technical field [0001] The present application relates to the technical field of communication, and in particular to a network data classification method and device, a network management server, and a storage medium. Background technique [0002] The passive optical network system is a broadband access technology built on the passive optical distribution network (ODN). It is connected upward to CMMET, IMS, CATV and other networks, and downward to various user terminals. Services such as WLAN hotspot access. The passive optical network system includes three parts: optical line terminal, optical distribution network and optical network terminal. Among them, the optical line terminal is the core component of the passive optical network system, which is equivalent to a switch or router in a traditional communication network. Various business data are aggregated, identified and forwarded. In the communication network system, there is a fixed-bandwidth DBA bandwidth allocation m...

AI Technical Summary

Problems solved by technology

In the communication network system, there is a fixed-bandwidth DBA bandwidth allocation method. The uplink bandwidth of the optical network terminal is usually pre-allocated and cannot be shared with other optical network terminals. This will cause the optical network terminal to go offline or be removed. The upstream bandwidth allocated for it still exists

If the offline configuration data of invalid offline optical network terminals is not deleted, the uplink bandwidth will be occupied for a long time, and the uplink bandwidth will be wasted, which will inevitably affect subsequent service activation and network efficiency

[0003] In the prior art, the grading method of network data is single, including only the criteria for judging whether the network data is highly suspicious of spam data, specifically: setting a time threshold in advance according to manual experience, and then judging whether the offline duration of the optical network terminal is When the threshold is reached, if it is reached, it is judged that the optical network terminal data is highly suspicious of junk data, and if so, it is deleted

However, the accuracy of the judgment standard set by the above method is not high, which will cause mishandling of the data (for example, if the judgment standard is higher than the actual situation, a large amount of data with a high degree of suspicion of junk data will not be identified. If the judgment standard is lower than the actual situation, it will cause a large amount of normal data to be mistakenly deleted), and the classification is rough and simple, and the deletion can only be performed according to the classification or not. The operation of deletion has flaws in the supervision of network data

Images