Data collection in future mobile networks

Future mobile networks will be increasingly heterogeneous. Already today, wireless LAN is used by many mobile network operators as an addition to traditional technologies like GSM and UMTS; WiMax and 3GPP Long Term Evolution (LTE) will be added. Having heterogeneous wireless networks, one challenging research question needs to be answered: Which user should be served by which access network and when to conduct a handover? For such decisions, information on the state of networks and terminals is required. In this publication, we simulate mobile networks in which a central entity called Network Resource Management (N-RM) gives handover recommendations to mobile terminals. Based on these recommendations and local knowledge on link qualities, the terminals choose the cell to switch to. The N-RM should have a global view on the networks to give best recommendations. We designed the Generic Metering Infrastructure (GMI), a publish/subscribe system to collect information about access networks and terminals efficiently. We investigate the tradeoff between the signalling overhead caused by data collection and the quality of the handover decisions and show, how smart monitoring can reduce the amount of measurement data while ensuring the efficient use of heterogeneous networks. In addition, our simulation results show that combined local and central handover decisions significantly increase the capacity of the networks as compared to only local decisions.