Packet Routing Engine for Adaptive Multipath Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing multipath routing in networks to minimize latency and improve fairness between destinations poses challenges, particularly in optimizing traffic distribution across multiple paths.
Innovation Solution
A routing engine that dynamically selects from multiple algorithms (HBF, weighted HBF, random, weighted random, pseudorandom, round robin) based on network conditions, implemented in hardware to enhance load balancing and traffic management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single routing algorithm is used, then device complexity is reduced, but adaptability to different network conditions deteriorates
Solution Approach 1:
The routing engine is designed to perform multiple routing algorithms (hash-based forwarding, weighted hash-based forwarding, random, weighted random, pseudorandom, round robin, and weighted round robin) within a single hardware unit. This multi-functional design allows the engine to adapt to different network conditions by selecting the appropriate algorithm while maintaining a unified hardware structure, thus resolving the contradiction between device complexity and adaptability.
2Ease of operation
If traffic is concentrated on fewer paths, then routing simplicity is improved, but network performance and fairness deteriorate
Solution Approach 1:
The routing engine dynamically selects among multiple routing algorithms based on current network conditions such as queue occupancy and latency. This dynamic adaptation allows the system to distribute traffic across multiple paths when beneficial for performance while maintaining algorithmic simplicity through a unified hardware structure, resolving the contradiction between routing simplicity and network performance.
3Adaptability or versatility
If multiple routing algorithms are implemented, then adaptability to network conditions is improved, but device complexity increases
Solution Approach 1:
Multiple routing algorithms that would traditionally require separate hardware implementations are merged into a single unified routing engine. The engine uses a common hardware structure with configurable parameters to execute different algorithms (hash-based, weighted hash-based, random, weighted random, pseudorandom, round robin, and weighted round robin), thereby achieving high adaptability without proportionally increasing device complexity.
4Productivity
If load balancing is optimized, then network performance is improved, but queue occupancy management complexity increases
Solution Approach 1:
The routing engine incorporates feedback mechanisms that monitor network conditions such as queue occupancy and latency for each egress port. Based on this feedback, the engine dynamically selects the most appropriate routing algorithm to optimize load balancing and network performance. This feedback-driven approach automates queue management, reducing the complexity of manual optimization while maintaining high network performance.
Data Source
AI summary
A switch, device, and method of routing packets are provided. In one example, a switch includes one or more circuits to generate an inflated transmission queue (TQ) vector based on a subset of elements, use the inflated TQ vector to select an element from the subset of elements, and route a packet via an egress port based on the element selected from the subset of elements.