Our site saves small pieces of text information (cookies) on your device in order to deliver better content and for statistical purposes. You can disable the usage of cookies by changing the settings of your browser. By browsing our site without changing the browser settings you grant us permission to store that information on your device.
Gigabit passive optical network (GPON) has evolved into large data provider technique. Analysis of network-generated data is critical for efficient fault diagnosis and self-configuration in GPON. Machine learning-based data analytics technologies could be significant in analyzing the performance of these kinds of optical networks. So, a machine learning approach for estimating the performance of GPON is proposed in this paper. A dataset containing fiber length, transmission power, the number of power splitters, line width, and extinction ratio factors has been created in this work to assess the Q factor value for a specific optical network. Then, the relief attribute evaluation technique is used to pick fiber length, transmission power, and the number of power splitters from among given parameters. For predicting different levels of Q factor, these specific parameters are supplied into a regression-based tree classification model. This paper considers logistic regression, decision tree, decision table, PART, and random forest algorithms for estimating the performance of GPON. As per the simulation findings of the present work, the proposed regression-based tree classification technique gives an effectual approximation of the Q factor with the accuracies of 93.8 % and 96.41% for seven-class and three-class cases respectively. As a result, the proposed approach appears to be a good fit for accurately estimating the performance of GPON.
Gigabit passive optical network, Machine Learning, Classification, Q factor, Performance estimation.
RAJANDEEP SINGH, P. RAVI KRUPA VARMA, KULDEEP SINGH, RAMANDEEP KAUR, GURPREET KAUR, Performance estimation method for gigabit passive optical networks using machine learning, Optoelectronics and Advanced Materials - Rapid Communications, 17, 1-2, January-February 2023, pp.44-50 (2023).
Submitted at: July 24, 2022
Accepted at: Feb. 6, 2023