Network Quality Prediction with QoS and QoE Data for Digital Television Using WebGIS

Abstract

Digital television is part of a rapidly growing technology. Currently, Indonesia has migrated from analog television to digital television with the broadcast standard DVB-T2 (Digital Video Broadcasting-Second Generation Terrestrial). A large number of users and the development of the services provided are challenges for increasingly complex network service problems. Therefore, it is necessary to analyze the problems that occur on the network to make improvements and increase customer satisfaction by using the QoS (Quality of Services) and QoE (Quality of Experience) parameters. To make it easier to know the quality of the network, an interactive website will be developed using WebGIS to display network quality and analysis. The results of the measurement and calculation of QoS and QoE parameters are used to predict unknown network parameter values using Linear Regression. The final test results showed that the accuracy of the prediction values for each of the QoS and QoE parameters obtained was good, with the highest accuracy value being 1 and the lowest accuracy being 0.74. Predictive values were classified as good, very good, poor, or very poor to determine network quality using Naïve Bayes. Naïve Bayes accuracy for the overall classification is quite good, which is above 0.8.