Twin-bit Based Fast IP Lookup and Update Algorithm

Abstract

Success of the Internet and the increased use of broadband in homes have caused a gradual shift in traffic on the Internet from data to multimedia communications. Traffic on the Internet traffic is increasing daily, while advances in communication technologies have allowed the Ethernet speeds to rise from 10Mbps to 100Mbps, and now to 10Gbps. The IP address lookup time in gigabit networks is a bottleneck for a router, which needs to find the longest prefix matching for the address. This study proposes a Twin-bit based IP address lookup and update algorithm, based on tree structures, called Fast Twin-bit Tree (FTBT). FTBT can effectively reduce the number of memory access times, and provide fast routing table update. Performance evaluation results reveal that the proposed algorithm can lookup an address among 78504 routing entries in six memory accesses on average.