Self Organizing Networks Coordination Function between Intercell Interference Coordination and Coverage and Capacity Optimisation using Support Vector Machine

It is expected that by 2020, billions of devices connect to internet using 5G cellular network. To provide seamless connectivity, as well as Quality of Experience and automate network functionalities like configuration, optimization and healing, Self Organized Networks (SON) have come into existence. In state of the art technology, this makes use of Machine Learning (ML) to overcome manual intervention and take appropriate decisions in given time. As there are multiple SON functions, there is a necessity of co-ordination among the functions to avoid conflicts. While detection of conflicts i.e., more than one function trying to modify the same parameter can be done using ML algorithms like anomaly detection, resolution of conflicts is implemented based on SON functions. One such problem is resolution of conflict between Inter Cell Interference Coordination (ICIC) and Coverage and Capacity Optimization (CCO). This is done by using Support Vector Machine (SVM) for generation of optimal antenna parameters using radial basis kernel and results are validated using LTE simulator ns3 LENA. From the dataset used for training and testing, which is validated using ns3, it is found that SVM is suitable algorithm for predicting antenna parameters in case of conflict between ICIC and CCO. Since SVM not only predicts multiple labels but also considers inter-relationship between the features it can be considered as most suitable algorithm for handling conflict between SON functions.