A context-aware seamless handover mechanism for mass rapid transit system

Internet users are now able to connect to the Internet anywhere at any time for the provision of ubiquitous wireless network. Furthermore, as IEEE 802.11 wireless networks have deployed widely, passengers of Mass Rapid Transit (MRT), one of the most popular transportation systems in modern cities nowadays, can access to the Internet through their mobile devices easily. However, MRT passengers bring massive simultaneous handovers to the system while they are getting on and off MRT coaches. Hence, mobility management becomes a challenging problem for ubiquitous Internet service in a MRT system. Although Mobile IPv6 (MIPv6) is designed to support IP mobility, several drawbacks of MIPv6 are reported and result in unacceptable handover latency. As a consequence, many proposals, such as Fast handovers for Mobile IPv6, Hierarchical Mobile IPv6 (HMIPv6), Fast Handover for Hierarchical MIPv6 (F-HMIPv6), and Proxy Mobile IPv6, have been proposed to tackle these drawbacks. Nevertheless, none of these proposals are adequate to cope with the large-number-simultaneous-handovers challenge. In this paper, we propose a context-aware seamless handover mechanism (C-HMIPv6) which solves the massive simultaneous handover problem based on the concept of context-awareness. C-HMIPv6 is based on HMIPv6 with following special designs. Firstly, distributed mobility anchor points (MAPs) are deployed to separate the loading of forwarding traffic. Secondly, every access router (AR) periodically exchanges mobile nodes (MNs') context with adjacent ARs and periodically broadcasts the network configuration of adjacent ARs to its MNs. Thus, all MNs and ARs are fully context-awareness in the MRT system. The MN is able to generate its new CoA prior to the actual handover and skip IEEE 802.11 channel scanning, which alleviate the majority of the handover latency. The old AR can notify the MN's MAP to take care of the MN's packets during the handover procedure while the new AR can perform binding update on behalf of the MNs. In C-HMIPv6, MNs do not need to participate in sending any related IP mobility signaling. As a result, seamless handover can be achieved even when a large number of MNs perform handover simultaneously. The performance of C-HMIPv6 and F-HMIPv6 is evaluated via simulations. The simulation results show that C-HMIPv6 is able to provide better performance in terms of handoff delay, packet delay and packet loss rate than F-HMIPv6.