Mobile Network Optimisation Through Advanced Simulation

The optimisation of large-scale mobile telecommunications networks is currently a slow and labour-intensive process. Given this, network operators tend to optimise their networks relatively infrequently, eg, every few months. However, the network traffic load and the service mix (ie, the relative proportion of different services within this traffic load) can vary dramatically from day-to-day during a week, from hour-to-hour during a day and even from minute-to-minute. Therefore, the performance of a network could be significantly improved if it were able to adapt rapidly, in real-time, to the changes in network load and traffic mix.

In Project Monotas, we are examining the feasibility of allowing a network to adapt its own network parameters (eg, pilot power levels, handover decision parameters) to the short-term changes in network load.

This research will ultimately lead to autonomous, self-optimising networks that are capable of reacting to rapid, unpredictable changes in the network traffic. Networks of this nature will require less drive-testing and optimisation effort and they will allow operators to offer services at a lower cost.