A Research On Quality Prediction Of EPON Network Based On Association Analysis

Broadband access networks based on EPON technology have the advantages of high bandwidth and rapid deployment.However,with the increasing number of high-bandwidth users,the bandwidth,reliability,and timeliness of service recovery have high standards.In addition,the network based on EPON technology has the advantages of high profit,stable income,and rapid iteration and upgrade of new services in the later period.All major communication operators attach great importance to the construction of EPON networks.Therefore,the improvement of EPON network quality is the main battleground for operators to compete for.Based on the technical means in quality management,this paper collects the quality problems of communication networks,uses data association algorithms to find the correlation between the causes and causes of quality problems,and makes targeted countermeasures based on these reasons.This paper uses optimization management processes and quality management to optimize network goals.By comparing the operation and maintenance indicators after network optimization,the EPON network failure rate and complaint rate can be effectively reduced.Important tasks to achieve the purpose of predicting EPON network quality are as follows.1.Applying qualitative analysis methods to analyze quality problems,including using brainstorming method,fishbone diagram method,and expert group method to qualitatively analyze the acquired data.2.Using data association analysis method to analyze the data association through the collected data,find out the root cause of the quality,and formulate targeted countermeasures according to the reasons.Further,this paper solves the quality problem and achieves the pre-established failure rate reduction as well as the complaint rate and also the goal of improving network reliability.This quality analysis method is indicated in the future network operation and maintenance to ensure reliable and stable operation of the network.