Radio Resources Dimensioning Using Machine Learning with case study in Live Network
Ashraf Roshdy, Ayman Gaber, Mohamed M. Khairy · 2017
The authors propose a new tool to extensively analyze and plan the radio resources based on machine learning technique using available data analytics methods like clustering, correlation and regression. This will be done through; Classifying the cells according to their Priority and required quality of service, detecting resources utilization that cause throughput limitation, efficient dimensioning of the sites' capacity according to their load, advising by the capacity off-loading solutions, and offering potential cost saving by redeployment the resources from low to high utilized cells. Those features will achieve optimum resources utilization across the network. Moreover, detected throughput bottlenecks are relieved and overall system performance is enhanced. This will reflect on the customer satisfaction with a minimum spending cost, and minimum human intervention.