Distributed IPv6 multihoming support

The typical motivation for multihoming [Black, B. et al., November 2001] is to improve reliability and performance. However, being multihomed is intrinsically complicated. Specially, the fundamental to IPv6 is based on the concept of firm route aggregation, by which that TLA (top level aggregator) would need to announce the aggregate of their space only to other TLA peer [Wegner, J.D. et al., 2000]. Thus, current IPv4 multihoming schemes [Stewart, J. W. III, 1999] that announce the IP address space to all upstream providers violate fundamental principle regarding aggregation. Due to above reason, it required a fair amount of thoughts and careful configuration to deploy multihoming on IPv6 network. In this paper, we propose a new efficient and deployable multihoming scheme, which is to target for solving the technical and potential problem of multihoming support at site exit router (RFC 3178) [Hagino, J. L. et al., October 2001]. Moreover, it can also provide the limited load sharing functionality. Such advantages are evaluated specifically through the results of the simulation. For all case, we can see performance of the proposed scheme improved with defined comparison factors.