Effective Resource Segmentation for Centralized-RAN in 5G Networks

Abstract

The technology used in tele-communication system is evolving rapidly to meet the growing demand of clients. The 5G cellular network, intended to house enormous number of IoT and other internet enabled end user devices with low latency. As the number of users increase in the system it becomes more difficult to manage resource allocation. Therefore, the implementation of 5G network requires an efficient approach towards management and segmentation of resources. Use of Centralized Radio Access Network (CRAN) is proposed because of its efficiency to handle highly dense networks. Coordination among the resources as well as huge number of user devices is quite a problem due increased complexities. In this paper we propose a solution which could efficiently handle the resource allocation problem with a learning based RA technique that uses Random Forest Algorithm. This technique takes various parameters as input, then predicts the Code and Modulation Schema (MCS) which should be used for establishing connection between end user device and the Remote Radio Head (RRH). This technique is efficient, compared to the conventional technique which uses Channel State Information (CSI).