Handover optimisation using neural networks within LTE

Mobile communication infrastructures are getting more complex with the addition of femtocells into the network architecture. Allied with this, the increased use of smart phones add strain onto the network because of higher data requirements. Femtocells are a useful resource to reduce the demand on the macrocell layer and effective handover management is needed to transfer services to and from each base station. The importance of handover management is high within LTE and is included within a use case of Self Organizing Networks. Base stations can autonomously decide whether handover should take place and assign the values of relevant parameters. Setting relevant parameters effectively requires more delicate attention with femtocells to allow for effective and seamless handover to the macrocell. Novel approaches with small amounts of additional signal processing can be utilised to improve handover efficiency. In this thesis, variations of Self Organising Maps have been implemented. Self Organising Maps can be used to learn the locations of the indoor environment from where handover requests have occurred and, based on previous experience, decide whether to permit or prohibit these handovers. Once the neural network has adapted to the indoor environment, handover can be optimised in different regions independently while still permitting necessary handover. The results of the investigations described within this thesis show that utilising location within the handover process is en effective way to improve handover performance within an indoor environment using an LTE femtocell.Mobile communication infrastructures are getting more complex with the addition of femtocells into the network architecture. Allied with this, the increased use of smart phones add strain onto the network because of higher data requirements. Femtocells are a useful resource to reduce the demand on the macrocell layer and effective handover management is needed to transfer services to and from each base station. The importance of handover management is high within LTE and is included within a use case of Self Organizing Networks. Base stations can autonomously decide whether handover should take place and assign the values of relevant parameters. Setting relevant parameters effectively requires more delicate attention with femtocells to allow for effective and seamless handover to the macrocell. Novel approaches with small amounts of additional signal processing can be utilised to improve handover efficiency. In this thesis, variations of Self Organising Maps have been implemented. Self Organising Maps can be used to learn the locations of the indoor environment from where handover requests have occurred and, based on previous experience, decide whether to permit or prohibit these handovers. Once the neural network has adapted to the indoor environment, handover can be optimised in different regions independently while still permitting necessary handover. The results of the investigations described within this thesis show that utilising location within the handover process is en effective way to improve handover performance within an indoor environment using an LTE femtocell