Mobility Management and QoS in BRAIN Access Networks

The functionality and efficiency of micro mobility schemes in wireless networks has a great impact on the overall performance of the whole network. In small cell wireless LAN environments, the network must support both frequent and numerous handovers at the same time. Together with providing fast and reliable handoffs, the micro mobility scheme must not affect the service perceived by users. This paper builds on and continues the micro mobility protocol evaluations that have been carried out in the BRAIN project. We select a small number of key design choices to classify and group existing protocol proposals. We observe that with appropriate extensions most existing protocols could be adapted for a BRAIN environment and we make suggestions on these extensions. We also show that some of the extended protocols show fundamental operational similarities when used in a BRAIN environment. As a second goal, we present a QoS architecture framework that would be as independent as possible from the micro mobility scheme used, thus leaving most room for different micro mobility decisions. Having the micro mobility and QoS mechanisms separated allows the parallel and independent evolution of the protocols. We conclude that despite the apparent diversity of existing mobility management protocols BRAIN has a limited set of applicable protocols to choose from and we outline the most important properties of alternatives. Some fundamental design decisions can however affect the QoS architecture. These issues are also identified.