The introduction of measurement based optimisation in the context of self-organising networks enables the optimisation of the handover decision. The key enabler for the optimisation are a cost function that incorporates the relevant handover performance indicators, a reasonable observation time to evaluate the performance and an optimisation algorithm that reliably improves the handover performance in various, ever-changing network conditions. In the recent years several handover optimisation algorithms have been investigated. Nevertheless, the influence of the target function on the optimisation, the dimensioning of the observation window and the impact of network condition changes have not been investigated so far. In this dissertation a detailed analysis of the handover performance indicators is presented. Beyond that, additional system information or measurements are valued as potential candidates to allow further improvement of the handover performance. Particular attention is paid to the ability to adapt to changing network conditions since the introduction of new cell layers (small cells), new techniques like adaptive antenna systems or spectrum sharing or the introduction of new communication technologies like LTE-Advanced increases the complexity of future mobile communication networks. Finally, we develop an optimisation algorithm that reliably and quickly optimises the handover performance in various and fast-changing network conditions.