Distributed mobility management in future networks: Tunneling or host routing?

The exponential increase of mobile Internet traffic has been forcing the current mobile network architectures to evolve towards flatten network architectures with IP backhaul support. Hence, current centralized mobility management approaches need to be adapted to cope with flat network architectures. The distributed mobility management is a promising direction to cope with the flatten network architectures and distribution of content servers. Thus, the distributed IP mobility approaches are characterized and grouped into tunneling and host routing models. Both models are evaluated regarding performance and scalability through simulation in different scenarios. The outcome demonstrates that both models are valid for mobility management of future networks, depending on the users, the network and the sessions. Overall, the host routing provides a better data delivery performance than tunneling approaches, but a worst scalability. Regarding the tunneling approach, the AR anchoring has an opposite behavior to host routing, while the CN anchoring is the approach that provides the best tradeoff between performance and scalability. Moreover, the global mobility management can be easily supported by the coexistence of these different models, when tightly-coupled.