|
|
Location Aware Analytics in the Context of Mobile Network Performance Optimization
Urbanová, Lucie ; Miloš, Jiří (oponent) ; Slanina, Martin (vedoucí práce)
This thesis deals with the location aware analytics in the context of mobile network performance optimization. A tool which estimates initial network parameters in the location with unknown network performance based on RTR NetTest measurements database is presented. The thesis briefly introduces the topic of big data and machine learning and gives an overview of NetTest application concept and functionality. A set of regression methods is presented and their complexity and suitability for the purposes of coverage maps creation is compared. After their thorough 1D analysis, IDW and GPR are analysed in 2D and used to create a set of estimation maps of network parameters. Evaluation of their accuracy is made based on reference measurements using NetTest application.
|
|
|
Location Aware Analytics In The Context Of Mobile Network Performance Optimization
Urbanová, Lucie
The goal of this paper is to develop an estimation tool capable of predicting location aware network parameters, based on their previous field measurements and to perform a set of additional measurements to verify the accuracy of the tool. Additionally, the paper evaluates a number of regression methods in terms of their prediction accuracy, complexity and the amount of input data needed to create a prediction map of valid results.
|
|
|
Location Aware Analytics in the Context of Mobile Network Performance Optimization
Urbanová, Lucie ; Miloš, Jiří (oponent) ; Slanina, Martin (vedoucí práce)
This thesis deals with the location aware analytics in the context of mobile network performance optimization. A tool which estimates initial network parameters in the location with unknown network performance based on RTR NetTest measurements database is presented. The thesis briefly introduces the topic of big data and machine learning and gives an overview of NetTest application concept and functionality. A set of regression methods is presented and their complexity and suitability for the purposes of coverage maps creation is compared. After their thorough 1D analysis, IDW and GPR are analysed in 2D and used to create a set of estimation maps of network parameters. Evaluation of their accuracy is made based on reference measurements using NetTest application.
|
|
| |
host ::
RSS.