Dynamic Threshold Flow Control for Shared Buffer Congestion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing flow control systems in digital communication networks face challenges in managing congestion across different traffic classes, leading to unpredictable delays and underutilization of resources, as they typically pause all traffic when congestion occurs, affecting high-priority delay-sensitive traffic and failing to share buffering resources effectively.
Innovation Solution
A dynamic threshold mechanism that adjusts admission control rules based on congestion information, allowing separate accounting for high-priority and low-priority traffic, enabling dynamic flow control across shared buffer pools to prioritize high-priority traffic while leveraging unused buffer resources during non-severe congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If link level flow control (LLP) is triggered to prevent buffer overflow, then data loss is minimized, but all traffic including high-priority delay-sensitive traffic is paused causing unpredictable delays
Solution Approach 1:
The invention segments traffic into different classes (assured forwarding classes and default forwarding class) and applies flow control selectively to each class based on its buffer usage status. This allows high-priority traffic to continue flowing while lower-priority traffic is paused, resolving the contradiction between preventing data loss and maintaining timely delivery.
Solution Approach 2:
The invention applies different flow control thresholds and behaviors to different traffic classes based on their specific buffer usage conditions. Each traffic class has customized admission control rules that adapt to its current buffer state, allowing differentiated quality of service while preventing overall traffic pause.
2Ease of manufacture
If fixed thresholds are used for flow control triggering, then implementation is simple, but buffer resources are under-utilized during non-severe congestion
Solution Approach 1:
The invention transitions from static fixed thresholds to dynamic thresholds that automatically adjust based on real-time buffer usage conditions. The admission control rules adapt their thresholds according to the current buffer state of each traffic class, optimizing buffer utilization during non-severe congestion while maintaining simplicity through automated adaptation.
Solution Approach 2:
The invention implements feedback mechanisms where buffer usage information is continuously monitored and fed back to the admission control rules. This feedback loop enables the system to dynamically adjust flow control thresholds based on actual buffer conditions, preventing both over-utilization and under-utilization of buffer resources.
3Reliability
If separate buffer pools are allocated per traffic class for lossless behavior, then lossless delivery is guaranteed, but buffering resources cannot be shared leading to resource waste
Solution Approach 1:
The invention merges buffer pools across different traffic classes into a shared buffer resource pool while maintaining class-specific accounting and admission control. This allows buffer resources to be shared and leveraged across traffic classes during bursty demands, while still guaranteeing lossless delivery for each class through controlled access mechanisms.
Solution Approach 2:
The invention creates a universal buffer sharing mechanism that serves multiple traffic classes simultaneously. The shared buffer pool can be dynamically allocated to different classes based on their buffer usage status and admission control rules, enabling the same buffer resources to serve multiple functions across different traffic types.
4Ease of manufacture
If per port flow control granularity is used, then implementation is straightforward, but both high and low priority traffic are paused equally
Solution Approach 1:
The invention segments the flow control mechanism at the traffic class level rather than the port level, allowing independent control of each traffic class's flow based on its buffer usage. This segmentation enables differentiated flow control behavior for different traffic classes while maintaining implementation simplicity through standardized class management.
Data Source
AI summary
A flow control sender includes an ingress port with one or more Class Groups (CG) defined including a shared buffer pool, a shared counter per ingress port per CG tracking an amount of the shared buffer pool utilized by each CG, an ingress port utilization counter per ingress port tracking an amount of the shared buffer pool utilized by the ingress port, and a controller computing a dynamic threshold for each CG, comparing the dynamic threshold of each CG with the ingress port utilization counter, and determining a particular CG experiencing congestion when the ingress port utilization counter is greater than the dynamic threshold for the particular CG. A flow control receiver ceases transmission of data packets to the particular CG experiencing congestion and allows transmission of the data packets corresponding to other CGs.