Abstract

Proposed herein are techniques to determine whether a 4G/5G base station, or a Wireless Local Area Network (WLAN) Access Point (AP), can meet/satisfy performance requirements of newly added (or handed over) applications or user equipment (UE) in a 4G/5G/WLAN system. The techniques presented herein collect a variety of parameters from different base stations, perform pre-processing, and analyze these parameters at a Performance Predictor Module. The Performance Predictor Module uses Machine Learning (ML) methods, such as Support Vector Machines (SVM) or Support Vector Regressions (SVR) with the feature set and target variables, to learn to predict whether or not a base station can meet performance requirements of, for example, added UE. The learning process is accelerated since the Performance Predictor Module learns from multiple base stations. The Performance Predictor Module also predicts degraded Quality of Service (QoS) classes that can be supported by the originating or neighboring base stations if the required/desired QoS class cannot be supported by the originating base station. This feedback is provided to the originating base station which takes policy based decisions using the feedback and the policies configured at that base station and the UE. The Performance Predictor Module may also provide feedback for various what-if scenarios.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS