Congestion-Aware Multipath Routing in Switching Fabrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network switching technologies face challenges in optimizing bandwidth and latency by efficiently utilizing both minimal and non-minimal routes in a switching fabric, leading to potential congestion and increased power consumption.
Innovation Solution
A switching fabric that uses traffic congestion information to opportunistically select egress ports for packet forwarding, favoring minimal routes when congested and using non-minimal routes when lightly loaded, thereby balancing bandwidth and latency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If non-minimal routes are used to increase bandwidth, then bandwidth is improved, but latency increases
Solution Approach 1:
The system dynamically selects between minimal and non-minimal routes based on real-time congestion conditions. When minimal routes are congested, the system transitions to using non-minimal routes, and vice versa. This dynamic adaptation allows the network to optimize bandwidth utilization while managing latency impact through conditional route selection rather than static path assignment.
Solution Approach 2:
The system changes the routing parameter (route selection) based on congestion metrics. By monitoring congestion levels and adjusting the chosen path accordingly, the system can exploit non-minimal routes when they offer better bandwidth availability, while minimizing latency impact by selecting the optimal route combination for each traffic flow.
2Productivity
If non-minimal routes are used to utilize idle links, then bandwidth is improved, but power consumption increases
Solution Approach 1:
The system employs feedback mechanisms to monitor link utilization and congestion levels. By continuously gathering information about network conditions and feeding it back into the routing decision process, the system can identify when non-minimal routes through idle links provide beneficial bandwidth utilization, and only activate those paths when the performance benefit justifies the additional power consumption.
Solution Approach 2:
The system adjusts routing parameters based on power-performance tradeoff analysis. When non-minimal routes are selected to utilize idle links, the system does so strategically only when congestion metrics indicate that the bandwidth gain outweighs the increased power consumption, thereby optimizing the energy-efficiency-performance balance.
3Productivity
If congestion information is collected and used for routing decisions, then network performance is improved, but device complexity increases
Solution Approach 1:
The system segments the routing decision process into distinct functional modules: congestion information collection, congestion metric computation, and route selection. Each module handles a specific aspect of the routing logic, which simplifies the overall implementation complexity while enabling sophisticated congestion-based routing decisions that improve network performance.
Solution Approach 2:
The system introduces an intermediary routing layer that sits between the physical network infrastructure and the traffic flows. This intermediary component (the routing logic) collects congestion information, processes it through defined metrics, and translates it into routing decisions, thereby managing the complexity of congestion-based routing in a structured and maintainable way.
Data Source
AI summary
A switching fabric uses traffic congestion information to inform its opportunistic use of non-minimal routes. An ingress port of a network switch collects traffic congestion information from the egress ports of the network switch. The traffic congestion information includes minimal and non-minimal route congestion metrics for the egress ports. Candidate egress ports for forwarding a packet to a destination node are identified. One of the candidate egress ports is selected based on the traffic congestion information. The selection process is biased to prefer some candidate egress ports over others. Particularly, the candidate egress ports that provide non-minimal routes to the destination node and have high minimal route congestion metrics are disfavored by the selection process.