Analysis of Mobile IPv6 handover optimizations and their impact on real-time communication

The Mobile IPv6 protocol is becoming one of the most common ways to support IPv6 mobility. Yet, despite its popularity the protocol still suffers various limitations which prevent it from being adopted in large scale commercial deployments. Among these is the protocol's poor support (or lack thereof) for rapid and seamless handovers. When moving from one subnet to another, a mobile node may experience connection and/or packet loss which may introduce noise and cuts in real-time media streams delivered to the user. Many different solutions to this problem have been presented and evaluated in various papers during the last few years but none of them have so far been selected as the standard Mobile IPv6 handover optimization. In this article we present experimental analysis of two optimization schemes and compare their performance to that of a standard Mobile IPv6 implementation. One of them, the Fast Handovers Mobile IPv6 protocol (FMIPv6) is being strongly supported by IETF working groups and is probably the one that, among all other handover optimization schemes, has most approached standardization. The GPS Handovers solution, on the other hand seems to be one of the few that have the maturity and completeness of the FMIPv6 protocol, and in the same time resolves problems that seem to be unaddressed by FMIPv6. All analysis in this paper is based upon experimentation which we believe is superior to simulation or emulation when dealing with a subject that is so heavily influenced by the implementation specifies of the accompanying technologies. Mobile nodes in all experiments are communicating via the IEEE 802.11 WLAN technology. All experiments are conducted with unmodified versions of popular conferencing and streaming applications Gnomemeeting and VLC.