The eNodeB Selection Using Channel Outcome with Machine Learning in Dense 5G Networks

This paper deals with the problem of eNodeB selection and channel estimation during the random access process of dense 5G networks and proposes a machine learning (ML) based solution to the problem. In a dense 5G network, a device may initiate the random access procedure with one of the many available eNodeBs, since the device can get coverage from many surrounding eNodeBs. Selection of the eNodeB such that the probability of success is maximum is the task of primary importance. The selection depends on the channel condition whether idle or busy, and estimation of the channel condition is also very important. We propose the use of ML to solve the problems of eNodeB selection and channel estimation. The simulation results validate our mentioned propositions.