Improving mobility management in 5G networks

The number of Internet users is expected to grow exponentially. To keep up with the growing demand and need for 5G mobile services, small cells like pico and femto base stations are added to the network of macrocells in the same area to form Heterogenous Networks (HetNets). HetNets have the ability to greatly increase system capacity to consumers while also providing uninterrupted high-rate communication services to users more reliably. In order to provide uninterrupted service, the network must handover the user's radio connection to the new cell when they switch cells. Mobility management is one of the most difficult aspects of 5G HetNet.

When a significant number of small cells are deployed into a HetNet, cell edges and inter-cellular contact grow. This results in more handover occurrences and radio link failures. Furthermore, additional signal load arises as a result of mutual signalling that takes place during handover failure and ping-pong handovers, which consumes the resources that were allotted for data transfer in an inefficient manner while also reducing the battery life of the user equipment.

This research focuses on developing novel mobility management models for 5G HetNets that would take into consideration the speed of the user in a condensed eMBB setting in an effort to reduce handover delays and unnecessary handovers. This in turn would enable delivering better QoS for mobile user equipment (UE) in a 5G network. Furthermore, milli-metre wave frequencies are considered an enabling technology for 5G networks. However, these frequencies are notorious for their high susceptibility to the environment. Therefore, for the progression of developing 5G networks, it is imperative to establish novel mobility management mechanisms that would lower handover issues while maintaining continuous connectivity for mobile users.

In this research we propose an improved dynamic Time-to-Trigger algorithm that considers both the user's speed and signal to interference and noise ratio (SINR) difference experienced. The goal of this improved handover mechanism is to reduce the number of handovers for UEs moving at different speeds (up to 60 mph), leading to reduced packet loss and latency, and improved throughput achieved by users. Compared to the traditional dynamic algorithm, this research reports significant benefits of the proposed dynamic time-to-trigger algorithm for users moving at varying speeds. Users moving at 60 mph reported 10 percent increase in throughput when utilising the proposed algorithm, whereas users moving at 15 mph reported 5 percent increase in throughput, compared to the state of the art.