A kernel methods approach to reducing handover occurrences within LTE

This paper presents a kernel-based approach to indoor-outdoor handover management for 4G femtocells. It is a necessary but difficult task to perform seamless handover from indoor femtocells to outdoor macrocells whilst maintaining call continuity. This paper describes a machine learning algorithm to operate in conjunction with 4G handover triggering mechanisms to reduce the rate of unnecessary handovers between femtocells and macrocells. The results of this algorithm show that handovers can be reduced by 65% by detecting where unnecessary handovers are likely to occur and minimising them. By reducing the number of unnecessary handovers, the system resources efficiency may be improved as a result of the potential reduction in signalling exchange taking place which in turn reduces bandwidth usage, the power used by both the femtocell and the mobile terminal and, the level of signal processing necessary.