FIB Aggregation Algorithm Reducing Router Memory and Update Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The rapid growth of the Forwarding Information Base (FIB) due to increased network prefixes and address fragmentation leads to scalability issues, high processing costs for route updates, and memory constraints in routers, necessitating efficient FIB aggregation techniques that balance compression ratio, running time, and update handling efficiency.
Innovation Solution
The FIB Aggregation with Quick Selections (FAQS) algorithm employs a single tree traversal for incremental updates, uses a PATRICIA trie data structure, and applies three optimization techniques to reduce memory consumption and processing time, achieving high compression ratios and maintaining forwarding correctness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If FIB aggregation is applied to reduce the number of FIB entries, then the size of FIB is reduced and router lifetime is prolonged, but the processing time for route updates increases due to subtree traversals
Solution Approach 1:
The patent applies preliminary action by pre-computing and storing aggregation information in the PATRICIA trie structure during initial FIB aggregation. This allows the system to quickly determine next-hop selections without performing extensive subtree traversals when route updates occur, thus reducing update processing time while maintaining aggregation benefits
Solution Approach 2:
The patent replaces the mechanical subtree traversal mechanism with a more efficient data structure-based approach. By using PATRICIA trie with pre-computed aggregation information, the system substitutes the traditional mechanical traversal process with a faster lookup and selection mechanism that achieves similar aggregation goals without the time penalty
2Quantity of substance
If optimal FIB aggregation is achieved using existing algorithms, then the compression ratio is maximized, but the memory consumption increases due to maintaining candidate lists
Solution Approach 1:
The patent extracts only the essential aggregation information needed for next-hop selection and stores it compactly in the PATRICIA trie structure. By taking out only the necessary data elements rather than maintaining complete candidate lists, the system achieves high compression ratios while consuming less memory
Solution Approach 2:
The patent changes the data structure parameters from maintaining full candidate lists to storing compact aggregation metadata in the PATRICIA trie. This parameter change enables the system to achieve optimal compression ratios with reduced memory footprint by storing only the essential selection criteria rather than complete candidate sets
3Ease of manufacture
If FIB aggregation is implemented to reduce operational costs, then hardware upgrades are delayed, but the complexity of the aggregation algorithm increases
Solution Approach 1:
The patent implements a universal PATRICIA trie structure that serves multiple functions: it stores the FIB entries, maintains aggregation information, and enables efficient next-hop selection. This multi-functional approach consolidates what would otherwise require separate complex aggregation algorithms and data structures, reducing overall system complexity while achieving the desired FIB compression
Data Source
AI summary
A system and method for FIB aggregation. FIB Aggregation with Quick Selections (FAQS) is a FIB aggregation algorithm that leverages compact data structures and three unique optimization techniques to quickly and incrementally select next hops when handling route updates. As a result, FAQS can run up to 2.53 and 1.75 times faster for IPv4 and IPv6, respectively, than the optimal FIB aggregation algorithm while achieving a near-optimal aggregation ratio. Meanwhile, it consumes much less memory and generates much smaller number of FIB changes when carrying out frequent updates. The performance enhancement of the new algorithm addresses many concerns from ISPs regarding performance issues, and enhances the probability to push FIB aggregation techniques further to the level of production adoption by the industry.