Internet address lookup is a challenging problem because of the increasing routing table sizes, increased traffic, higher speed links, and the migration to 128 bit IPv6 addresses. Routing lookup involves computation of best matching prefix for which existing solutions scale poorly when traffic in the router increases or when employed for IPV6 address lookup. Our paper describes a novel approach which employs multiple hashing on reduced number of hash tables on which ternary search on levels is applied in parallel. This scheme handles large number of prefixes generated by controlled prefix expansion by reducing collision and distributing load fairly in the hash buckets thus providing faster worst case and average case lookups. The approach we describe is fast, simple, scalable, parallelizable, and flexible.
展开▼
