Presentation is loading. Please wait.

Presentation is loading. Please wait.

Study of IP address lookup Schemes

Similar presentations


Presentation on theme: "Study of IP address lookup Schemes"— Presentation transcript:

1 Study of IP address lookup Schemes
Project Proposal Sana Sohail

2 Problem Statement A thorough study involving a careful comparison of IP address look up schemes in terms of: look up speed memory consumption update costs scalability.

3 Problem Background CIDR scheme solved the problem of inefficient usage of IP addresses It also made the IP look up operation more complex by converting it to longest prefix matching problem. This “problem” has two dimensions Comparison of bit patterns Finding the appropriate length CIDR (Classless Inter Domain Routing) is a bitwise, prefix-based standard for the interpretation of IP addresses. It has removed the limitations of class full addressing. Please refer to notes to learn about CIDR

4 IP look up schemes To solve this problem a lot of schemes have been proposed so far that fall in three major categories Tries Hashing Binary Search

5 Tries Schemes Binary trie Patricia Multi-bit tries LC- Tries
Lulea Algorithm 1. Binary Trie: It allows storage of IP addresses in a digital form by using bits to direct the branching. It stores 1 bit in a node. 2. Patricia: Path compresses tries, collapses one way branch nodes into one node and thus saves memory consumption. 3. Multi-bit tries: They inspect more than one bit at a time and thus increase the efficiency. 4. LC- Tries (Level Compressed tries): They use optimization techniques in addition to multi- bit tries. 5. Lulea algorithm: uses prefix expansion and thus reduce the memory size. Please refer to have an idea of key differences

6 Hashing Techniques Binary search on length and hashing on value
Multiple hashing and parallel processing Allocation of keys among memory modules based on statistics The first one was proposed by Waldvogel in which binary search is applied on hash tables organized by length. They used markers to guide their search. A second approach suggested a parallel IP address lookup architecture based on multiple hashing. Another technique introduced allocation of hash keys to different memory modules based on statistics to minimize collision. Please refer to notes for short description

7 Binary Tree techniques
Eliminated length factor - applying Binary search on disjoint address intervals Reduction of tree height by elimination of empty internal nodes The binary search is based on exact match. The keys with different lengths cannot be compared directly in IP lookup. So in order to overcome this limitation, binary search is applied on disjoint address intervals. Another approach proposed by Yazdani suggested a way to compare values with different lengths. Please refer to notes for short description


Download ppt "Study of IP address lookup Schemes"

Similar presentations


Ads by Google