Dynamic Queueing Granularity Switching in Network Switches
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network switches face challenges in efficiently managing network traffic due to limitations in existing queuing systems, which often result in suboptimal performance and resource utilization, particularly when dealing with varying traffic conditions and quality of service requirements.
Innovation Solution
The implementation of a dual queuing system within network switches, where network traffic is initially queued according to a first system with a lower degree of granularity and dynamically switched to a second system with a higher degree of granularity based on observed traffic characteristics and switching criteria, such as bandwidth and QoS, to optimize resource allocation and traffic flow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single queuing system with fixed granularity is used, then the system structure is simple, but the network switch cannot adapt to varying traffic conditions and achieves suboptimal performance
Solution Approach 1:
The patent implements a dynamic queuing system that automatically adjusts between different queuing algorithms (WFQ, FIFO, PQ) based on real-time traffic characteristics. The system monitors traffic parameters such as bandwidth utilization, packet arrival rates, and QoS requirements, then dynamically selects the most appropriate queuing strategy, transforming a static system into an adaptive one that optimizes performance for varying network conditions
Solution Approach 2:
The system changes operational parameters by adjusting queuing algorithm selection and granularity levels based on traffic conditions. When traffic patterns change (e.g., from uniform to bursty, or when QoS requirements vary), the system modifies its queuing behavior by switching between different algorithms and granularity settings, allowing it to adapt to diverse traffic scenarios without requiring complete system redesign
2Measurement precision
If a high granularity queuing system is always used, then traffic management precision is improved, but resource utilization efficiency decreases due to excessive overhead
Solution Approach 1:
The system applies high-granularity queuing management only partially - specifically, it uses fine-grained control (individual flow-level queuing) only when traffic conditions warrant it, such as when multiple QoS classes are present or when traffic patterns require precise control. For simpler traffic scenarios, the system uses coarser-grained aggregation, avoiding the overhead of high-granularity management while maintaining adequate performance
Solution Approach 2:
The granularity level of the queuing system is dynamically adjusted based on traffic characteristics. The system monitors factors such as the number of active flows, QoS requirements, and bandwidth utilization, then adapts the queuing granularity accordingly - using fine-grained control when precision is needed and coarser-grained control when efficiency is prioritized, optimizing the trade-off between precision and overhead
3Productivity
If a low granularity queuing system is always used, then resource utilization efficiency is improved, but traffic management precision and QoS support deteriorate
Solution Approach 1:
The system applies low-granularity queuing management partially - using aggregate-level queuing for traffic flows that do not require precise control, thereby maintaining high resource utilization efficiency. However, when specific traffic conditions indicate a need for finer control (such as when QoS differentiation is required or when individual flow management is beneficial), the system selectively applies higher-granularity management to those specific flows while maintaining aggregate-level management for others
4Productivity
If multiple queuing systems are dynamically switched, then network performance is optimized for varying conditions, but system complexity and computational overhead increase
Solution Approach 1:
The patent segments the queuing management functionality into distinct, modular components - separate queuing algorithms (WFQ, FIFO, PQ), independent monitoring modules for different traffic parameters, and discrete decision-making logic for algorithm selection. This segmentation allows the system to implement multiple queuing strategies without creating a monolithic complex system, as each component can be independently managed, configured, and optimized
Data Source
AI summary
As example method includes queuing network traffic received at one or more input ports of one or more input modules of a network switch to a given output port of an output module of the network switch according to a first queuing system, subsequent to queuing the network traffic according to the first queueing system, queuing at least a portion of the network traffic according to a second queuing system instead of the first queueing system. According to the first queuing system, the network traffic is queued according to a first degree of granularity. According to the second queuing system, at least the portion of the network traffic is queued according to a second degree of granularity. The second degree of granularity is greater than the first degree of granularity.