Dynamic Congestion Control Thresholds for Shared Buffers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing statistical congestion control techniques in packet communication networks, such as ECN and WRED, are inefficient when used with dynamic buffer sharing schemes, as they apply fixed thresholds that lead to unnecessary packet dropping when the buffer is empty or insufficient packet marking when it's full, disrupting network performance.
Innovation Solution
The solution involves dynamically adjusting the congestion control thresholds based on the current queue length and allocation size in the shared buffer, increasing the fraction of packets subject to control as the buffer empties or fills, using a variable threshold proportional to unused buffer space, to optimize congestion management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If fixed thresholds are used in statistical congestion control techniques, then implementation is simple, but packet dropping occurs unnecessarily when buffer is empty or marking is insufficient when buffer is full
Solution Approach 1:
The patent applies dynamics by transitioning from fixed thresholds to dynamic thresholds that adapt to buffer occupancy conditions. The threshold for each queue is calculated as a fraction of the currently available buffer space, allowing the system to automatically adjust congestion control parameters based on real-time buffer status, thereby eliminating unnecessary packet dropping and improving marking accuracy.
Solution Approach 2:
The patent changes the parameter of congestion control threshold from a static fixed value to a dynamic value that varies with buffer occupancy. By expressing the threshold as a proportion (alpha) of available buffer space, the system adapts its congestion control behavior to current conditions, resolving the contradiction between implementation simplicity and control accuracy.
2Productivity
If dynamic buffer sharing is implemented, then buffer utilization is optimized, but existing congestion control techniques become inefficient due to mismatched thresholds
Solution Approach 1:
The patent implements feedback by continuously monitoring buffer occupancy and using this information to dynamically adjust congestion control thresholds. The threshold for each queue is recalculated based on current buffer availability, creating a closed-loop system that adapts to changing buffer conditions and maintains effective congestion control alongside dynamic buffer sharing.
Solution Approach 2:
The patent changes the congestion control parameters to be proportional to available buffer space rather than fixed values. This allows the congestion control mechanism to remain effective under dynamic buffer sharing conditions, as the thresholds automatically scale with the actual buffer resources available to each queue.
3Stability of the object's composition
If congestion control is applied with fixed fractions, then control is predictable, but network performance deteriorates under varying buffer conditions
Solution Approach 1:
The patent changes the congestion control fraction from a fixed value to a dynamic value that depends on buffer occupancy and queue allocation. By expressing the control fraction as a function of current buffer conditions rather than a predetermined constant, the system maintains predictability through consistent proportional relationships while adapting to varying network conditions to preserve performance.
Data Source
AI summary
Communication apparatus includes multiple interfaces configured for connection to a packet data network. A memory, coupled to the interfaces, is configured as a shared buffer to contain packets in multiple sets of queues for transmission to the network. Each set of queues receives in the shared buffer a respective allocation having an allocation size that varies over time in response to an amount of space in the shared buffer that is unused at any given time. A controller is configured to apply congestion control to a respective fraction of the packets that are queued for transmission from each set of queues in the shared buffer to the network, such that the respective fraction is set for each set of queues at any given time in response to a relation between a length of the queues in the set and the allocation size of the respective allocation at the given time.