Hand-off analysis for 5G networks using poisson point process in homogeneous and non-homogeneous environments

Abstract

Due to the high bandwidth of the 5G spectrum, the cell coverage area in 5G gets reduced promoting the Ultra-Dense Network (UDN) deployment. To keep user Quality of Service (QoS) intact, 5G needs to maintain an enormous number of active connections from Base Stations (BS) to mobile users during movement. Thus, in this thesis, we primarily show a process for the enhancement of network mobility in 5G by increasing the successful hand-off rate against one of the primary QoS parameters, Signal to Interference plus Noise Ratio (SINR). To construct mobility in simulation, we adopt user distribution as a spatially distributed Poisson Point Process (PPP) and iteratively deployed the same PPP to impose randomness in motion. The network models where PPP has been incorporated comprise three different case studies for analyzing the gradual enhancement of successful hand-off (in terms of successful Poisson arrival metric). Primary observations reveal that the successful handoff rate increases for higher SINR if we choose the interferer BSs layer as a group of selected cells forming variable cluster radius. Moreover, we found that the successful handoff percentage in higher SINR increases further by applying the Heterogeneous Network (HetNet) of two different network tiers. The first one is the traditional Femto Base Stations (FBSs) and the second is the Macro Base Stations (MBSs). Additional work has been made to establish a direct relationship between the average user rate and the Poisson arrival parameter 𝜆 and depicts how 𝜆 can be suitably varied to have control over the network load. Our work finally extends to the complex network environment where we have considered networks to be of anisotropic type. It is called a fractal coverage network where different paths in the network undergo different path loss exponents. It resembles a real-time network and performance analysis shows the behavior of coverage and association probability based on the SINR threshold in this multi-directional path loss model. It could be concluded that although the QoS of the 5G network can be taken care of through multiple techniques, it will still face a heavy network overload which is the key challenge in 5G UDN deployment.