Sharma, P and Thomas, V S and Kamath, N and Tamma, Bheemarjuna Reddy
(2016)
Velocity based Dynamic Flow Mobility in Converged LTE/Wi-Fi Networks.
In: 22nd National Conference on Communications (NCC), MAR 04-06, 2016, Guwahati, INDIA.
Full text not available from this repository.
(
Request a copy)
Abstract
In recent years, the boost of mobile-phones, tablets and smart devices with data-hungry applications has caused massive growth in demand of "Mobile data services" and fulfilling this demand is a very challenging task with all the available resources of cellular networks. This has brought us to the option of offloading traffic from the cellular networks to Wi-Fi networks. Offloading on Wi-Fi network for static users is comparatively easy when compared to mobile users because switching mobile users to Wi-Fi network may cause unnecessary ping-pong effect, packet loss and delay during vertical handovers. But the flow mobility solutions like PMIPv6 and DSMIPv6 provide the mechanism to offload traffic from one network to another network seamlessly and offloading is done at the granularity of flows instead of moving the entire traffic generated by the mobile user. The idea is to intelligently move certain flows onto either network so as to achieve better throughput and thereby improve the capacity of the heterogeneous network. User mobility plays a significant role while moving a flow across different networks. Generally, offloading decisions are based on signal quality and load of the networks. In this paper, we propose an integrated offloading approach, called as Velocity based Integrated Flow Mobility (VIFM), for converged LTE/Wi-Fi networks, which considers the signal quality, network load and user velocity to maximize the network performance. We studied the performance of VIFM in different scenarios using extensive NS-3 simulations and found that it reduces unnecessary flow handovers by 50% and increases the Wi-Fi network utilization by 12% when compared to previous techniques [1].
Actions (login required)
|
View Item |