An IP address is used as a host identifier and a locator to bind hosts and applications to their location in existing Internet. Several protocols are proposed to eliminate this binding. Most of these protocols use IPv6-based host identifiers to maintain compatibility with existing Internet, but these identifiers cannot be handled by standard IPv6 routers because such identifiers are unroutable. Therefore, host identifiers need to be usually converted to locators at hosts, and the standard IPv6 protocol should be modified to interoperate with these protocols. In this paper, we propose a network-based host identifier locator separating scheme in software-defined network.

The proposed scheme separates the underlyingnetwork into Host Identity and IP domains in order to directly forward unroutable identifiers. Host Identity domain includes its own routing strategy based on a distributed hash table algorithm. The proposed scheme directly handles the packets which have host identifiers as its addresses, and converts the host identifiers to the corresponding locators in the network. We compare the differences between the proposed scheme and other existing solutions.

• Next Post
• Prev Post