Dynamic Gate Control Lists for Network Queue Overload Relief
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network switches, routers, and gateways face issues with instantaneous network overload, leading to frame drops due to unbalanced queue utilization, particularly affecting high-priority packets and degrading Quality of Service (QoS).
Innovation Solution
Implementing a controller that dynamically adjusts the Gate Control List (GCL) based on queue utilization thresholds, reallocating buffer space from lower-priority queues to high-priority queues during overload, using adaptive and dynamic GCL strategies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If fixed depth queues and queue management algorithms per queue are implemented, then queue management is simplified, but frame drops occur during network overload due to unbalanced queue utilization
Solution Approach 1:
The patent implements dynamic GCL entries that can be modified at runtime based on actual queue utilization states. The controller monitors queue states and dynamically adjusts GCL configurations to allow high-priority queues to borrow transmission time from low-priority queues during overload conditions, transforming the static queue management into a dynamic adaptive system that prevents frame drops while maintaining manageable complexity
Solution Approach 2:
The system changes the transmission parameters of queues dynamically by modifying GCL entries. When a high-priority queue exceeds its threshold, the controller generates new GCL entries that alter the transmission gate control parameters, allowing the high-priority queue to transmit during time slots originally allocated to low-priority queues, thereby preventing frame drops without requiring increased total buffer capacity
2Reliability
If buffers are allocated to all queues, then queue capacity is ensured, but memory usage is inefficient when some ports have low activity
Solution Approach 1:
The patent creates a dynamic buffer allocation mechanism where the controller monitors queue utilization states and dynamically reallocates buffer resources through GCL modifications. When low-priority queues have low utilization, their buffer capacity is dynamically transferred to high-priority queues that are experiencing overload, ensuring capacity guarantees are met only when needed while optimizing overall memory usage
Solution Approach 2:
The system implements self-service buffer management where queues that have excess capacity automatically make their buffers available to other queues through the dynamic GCL mechanism. The controller enables queues to self-regulate their buffer usage based on actual traffic conditions, with low-utilization queues effectively donating their unused buffer space to high-utilization queues without external intervention
3Reliability
If GCL entries are modified at runtime based on queue states, then frame drops are prevented, but controller complexity increases
Solution Approach 1:
The patent implements a feedback mechanism where the controller continuously monitors queue utilization states and uses this feedback to dynamically adjust GCL entries. When a high-priority queue exceeds its threshold, the controller receives feedback about this state and automatically generates new GCL entries to reallocate transmission time, creating a closed-loop control system that prevents frame drops through adaptive response to actual network conditions
Solution Approach 2:
The controller manages complexity by focusing parameter changes on specific critical attributes of GCL entries rather than redesigning the entire queue management system. It modifies transmission gate control parameters dynamically based on queue states, changing only the necessary parameters (transmission timing and gate control) while maintaining the overall GCL structure, thereby achieving frame drop prevention with minimal increase in controller complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to controlling queue release in a network. In particular, the disclosure proposes a controller being configured to: obtain a state of each of a plurality of queues of a network node; determine, based on the states of the queues, whether the utilization of one or more queues exceeds one or more thresholds; generate one or more new entries for a gate control list of the network node that controls the plurality of queues, if one or more thresholds are exceeded; and provide the one or more new entries to the network node. Further, the disclosure proposes a network node being configured to provide a state of each of a plurality of queues to a controller, and obtain one or more new entries for a gate control list of the network node that controls the plurality of queues, from the controller.