Abstract
One of the most critical router’s functions is the IP lookup. For each incoming IP packet, IP lookup determines the output port to which the packet should be forwarded. IPv6 addresses are envisioned to replace IPv4 addresses because the IPv4 address space is exhausted. Therefore, modern IP routers need to support IPv6 lookup. Most of the existing IP lookup algorithms are adjusted for the IPv4 lookup, but not for the IPv6 lookup. Scalability represents the main problem in the existing IP lookup algorithms because the IPv6 address space is much larger than the IPv4 address space due to longer IPv6 addresses. In this paper, we propose a novel IPv6 lookup algorithm that supports very large IPv6 lookup tables and achieves high IP lookup throughput.
Highlights
The Internet constantly grows in every aspect: number of hosts and network devices, number of links, link capacities, variety of services and applications, QoS demands
Internet routers are the main components of the Internet infrastructure because the routers are responsible for the forwarding of IP packets to their proper destinations
Lookup table contains forwarding information for all the networks known to the router
Summary
The Internet constantly grows in every aspect: number of hosts and network devices, number of links, link capacities, variety of services and applications, QoS demands. The Internet routers must implement efficient high-speed packet processing to keep up with the Internet’s growth. Lookup table contains forwarding information for all the networks known to the router. We propose a novel IPv6 lookup algorithm Balanced Pipelined Lookup (BPL). BPL uses pipeline technique to achieve high lookup throughput and to support. Each stage contains similar number of nodes In this way, each pipeline stage uses the same type of memory making hardware implementation very efficient. The main contribution of the paper is the proposed scalable BPL scheme suitable for fast and efficient IPv6 lookups. BPL has lower memory requirements compared to other similar tree based lookup schemes. BPL supports efficient lookup table updates that do not decrease the lookup performance.
Published Version (Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have