Please use this identifier to cite or link to this item:

Title: High-Speed Route Search Mechanism for IPV6
Authors: Chen, Wen-Shyen E.
Tsai, Chung-Ting Justine
Hsu, Lu-meng
Date: 1999
Issue Date: 2017-10-18 16:52:55 (UTC+8)
Abstract: The Internet's astounding growth has begun to stress the technology that supports it, namely, the current Internet Protocol Suite, IPv4. IPv6 (Internet Protocol, Version 6) was developed to address the issues. The IPv6 resolves the Internet scaling problem, provides a flexible transition mechanism, and meets the needs of such new market as nomadic personal computing devices, networked entertainment, and device control. Nevertheless, the specification of IPv6 does not provide mechanisms to speed up the processing of the packets in the routers. As a result, the introduction of IPv6 might demand more computation from the already overloaded routers. In this paper, we propose a high speed IP lookup scheme for IPv4 and then describe how the mechanism can be scaled to IPv6 in order to speed up the packet processing in the routers. The proposed mechanisms search for the Best Matching Prefix (BMP) by using em forwarding tables consisting of Prefix Information Tables (PITs) and Lookup Tables (LTs) that provide guidelines for efficient search. The FPGA implementation of the scheme for IPv6 is also shown to validate its feasibility. For IP lookup in IPv4, the scheme needs 1 memory access in the best case, and 2 memory accesses plus one hash in the worst case to locate the BMP. It requires only 560 KBytes to 670 KBytes of memory space when about 45,000 routing tables entries in the backbone are logged for simulation. More memory (about 300 Kbytes more) can be used to reduce the worst case access time to 2 memory accesses. When 50ns DRAM is used for the forwarding tables, the scheme offers lookup speed of 10 millions packets per second. The lookup speed can be improved linearly with the speedup of the type of memory used. Further, it does not need complex compression mechanisms to reduce the memory requirement. Although the currently available traces and routing entries for IPv6 do not provide the scale similar to that of the IPv4, initial analysis shows the approach takes at most 3 hashes and 2 memory accesses to locate the BMP in IPv6. Parallelism can be explored to further improve the lookup performance.
Relation: TANet'99會議論文
Data Type: conference
Appears in Collections:[TANET 台灣網際網路研討會] 會議論文

Files in This Item:

File Description SizeFormat
090.pdf664KbAdobe PDF155View/Open

All items in 學術集成 are protected by copyright, with all rights reserved.

社群 sharing