Dynamic Scheduling Priority for Network Traffic Oversubscription
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing traffic management systems in communication networks fail to accurately handle cases where the sum of minimum rates demanded by child nodes exceeds the sum of allocated minimum rates, leading to oversubscription issues and inadequate bandwidth allocation, particularly in scenarios involving aggregation services.
Innovation Solution
A traffic manager with a scheduler and hierarchically arranged scheduling nodes that determine scheduling priority based on configured queue priorities, meter values, and shaper values, allowing for dynamic priority adjustments and efficient bandwidth allocation across multiple scheduling levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If statistical multiplexing is used to allow sum of allocated minimum rates to exceed total available rate, then bandwidth utilization is improved, but quality of service reliability deteriorates when demand exceeds allocation
Solution Approach 1:
The patent implements dynamic scheduling priorities that change based on traffic conditions and meter values. When minimum rate oversubscription occurs, the system dynamically adjusts priority levels to ensure that services experiencing congestion receive appropriate bandwidth allocation, thereby maintaining QoS reliability while allowing statistical multiplexing to improve overall bandwidth utilization.
Solution Approach 2:
The system changes the parameter of scheduling priority dynamically based on meter values and traffic conditions. By adjusting priority parameters in response to oversubscription events, the system maintains service quality guarantees even when the sum of allocated minimum rates exceeds available bandwidth.
2Reliability
If minimum rate propagation is implemented to allow child nodes to have minimum rate guarantees exceeding parent node allocation, then service delivery reliability is improved, but bandwidth allocation accuracy deteriorates in aggregation scenarios
Solution Approach 1:
The patent applies local quality by implementing different scheduling priority rules at different levels of the hierarchy. Child nodes can have minimum rate guarantees propagated to them, but the parent node implements dynamic priority adjustment based on actual traffic conditions and meter values, allowing local optimization at each level while maintaining overall allocation accuracy.
Solution Approach 2:
The system dynamically adjusts scheduling priorities at parent nodes based on real-time meter values and traffic conditions, rather than using static minimum rate propagation. This dynamic approach maintains service delivery reliability for child nodes while improving bandwidth allocation accuracy in aggregation scenarios by responding to actual traffic demands.
3Ease of operation
If priority is associated with a single user-defined bandwidth threshold, then ease of configuration is improved, but bandwidth allocation flexibility deteriorates when multiple priority levels are needed
Solution Approach 1:
The patent segments the priority system into multiple hierarchical levels (configured priority and dynamic priority). This segmentation allows the system to maintain simple configuration at the base level while providing flexible multi-level priority management through the dynamic scheduling mechanism, thereby improving bandwidth allocation flexibility without significantly complicating configuration.
Solution Approach 2:
The dynamic scheduling priority mechanism serves multiple functions: it maintains simple configuration through inheritance from parent nodes, provides flexible bandwidth allocation through dynamic adjustment, and supports multiple priority levels simultaneously. This multi-functionality resolves the contradiction between configuration simplicity and allocation flexibility.
Data Source
AI summary
A system for managing traffic in a communication network. The system includes a plurality of queues each configured to store data packets and a plurality of scheduling nodes each configured to process data packets from one or more of the plurality of queues. A scheduler is configured to schedule, using the plurality of scheduling nodes, respective transfers of the data packets from the plurality of queues. Each of the plurality of scheduling nodes is assigned to one or more of the plurality of queues. Each of the plurality of scheduling nodes and each of the plurality of queues is assigned a respective scheduling priority. The respective scheduling priorities are selectively changeable between a predetermined scheduling priority and a dynamic scheduling priority, wherein the dynamic scheduling priority corresponds to a priority propagated from the one or more of the plurality of queues.