No abstract

No abstract

The anticipation time has a significant impact on the performance of fast handover, especially predictive fast handover. Predictive fast handover which generally outperforms reactive fast handover requires a sufficient anticipation time to complete necessary operations before the current network is broken. However, the anticipation time may spur packets for the Mobile Node (MN) to be forwarded to the New Access Router (NAR) while the MN is still connected to the Previous Access Router (PAR). These packets are either delayed in the NAR's buffer or lost. To address such problem, in this paper, we propose a Link Triggered Buffer mechanism (LT-Buffer). The basic idea behind the LT-Buffer is to utilize the advantage of cross-layer interactions between the link layer and the network layer so that the delay time for packets in the buffer can be reduced. Moreover, the LT-Buffer mechanism can trigger buffering on the PAR to reduce packet loss in reactive fast handover. Finally, we evaluate the effect of the LT-Buffer.Previous Access Router (PAR)'s link. Then packets are forwarded to the NAR through the tunnel. The NAR can buffer these packets until the MN arrives. MN PA Link_Goin RtSolPr g Down C PrRtAdv FBU anticipation time FBack L Link Down< -disconnect Link deliver packets (a) Predictive Fast Handover

In Hierarchical Mobile IPv6 (HMIPv6), most of existing MAP selection schemes is Mobile Node (MN) based, which implements MAP selection on MN side and supposes all MNs use the same selection algorithm. But in practical environment, MNs usually adopt different MAP selection algorithms, which will disorder the MAP selection and make the performance of entire system hard to achieve as expected. Thus, a Dynamic Host Configuration Protocol for IPv6 (DHCPv6) based framework for MAP selection is proposed, where a DHCPv6 server is deployed to control the MAP selection by allocating proper Regional Care-of-Address (RCoA) for MNs. Moreover, a statistical prediction based MAP selection algorithm is introduced in the framework to reduce the Inter-MAP handoffs and achieve better load control. The algorithm uses Poisson Model to predict the number of MNs which will use MAPs in the future, and then selects a proper MAP based on these predicted quantities. Experimental results show that our proposed scheme significantly reduces the number of Inter-MAP handoffs while a better load control is achieved compared with existing distance-based and mobility-based schemes.