Ethernet Switch Traffic Management for Event-Triggered Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Ethernet communication standards, such as IEEE 802.1Q and IEEE 802.1Qbv, do not effectively manage high-priority real-time event-triggered traffic, leading to overprovisioning and inefficiencies in bandwidth allocation, as they are not well-suited for sporadic and bursty traffic with stringent latency requirements.
Innovation Solution
Introduction of a new traffic class, real-time event-triggered traffic (ET), with priority management and preemption mechanisms, allowing for low-latency transmission of event-triggered frames, while maintaining support for scheduled traffic (ST) and guaranteed latencies for stream-reservation traffic (SR), using modified Ethernet MAC layers and credit-based shapers to optimize bandwidth allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing Ethernet standards (IEEE 802.1Q, 802.1Qbv) are used for traffic management, then scheduled traffic and stream-reservation traffic can be supported with guaranteed latencies, but high-priority event-triggered traffic cannot be effectively managed, leading to overprovisioning and inefficiencies
Solution Approach 1:
The patent segments the traffic management system by introducing a dedicated high-priority queue specifically for event-triggered traffic, separate from the existing scheduled traffic queues. This segmentation allows event-triggered frames to be handled independently with immediate transmission priority, resolving the contradiction by enabling specialized handling for specific traffic types without disrupting the overall scheduled traffic management framework.
Solution Approach 2:
The patent implements dynamic priority assignment where event-triggered traffic can preempt scheduled traffic under certain conditions. The system dynamically adjusts transmission priorities based on traffic type and timing requirements, allowing high-priority event-triggered frames to interrupt lower-priority scheduled traffic when necessary. This dynamic behavior enables the system to adapt to varying traffic demands while maintaining efficiency.
2Reliability
If bandwidth is overprovisioned to accommodate sporadic event-triggered traffic, then latency requirements can be met, but bandwidth allocation becomes inefficient and resources are wasted
Solution Approach 1:
The patent implements a self-service mechanism where event-triggered traffic automatically gains access to the transmission medium through the dedicated high-priority queue without requiring pre-allocation of bandwidth resources. The system serves event-triggered frames immediately when they arrive, eliminating the need for overprovisioning while maintaining latency guarantees. This self-service approach allows the system to efficiently utilize bandwidth by allocating resources on-demand rather than reserving them in advance.
3Stability of the object's composition
If existing traffic classes are used, then scheduled traffic can be transmitted according to predefined temporal programming, but event-triggered traffic suffers from interference and cannot achieve low-latency transmission
Solution Approach 1:
The patent segments the transmission queues into distinct high-priority and standard-priority queues. Event-triggered traffic is directed to the high-priority queue which is transmitted immediately when ready, while scheduled traffic continues to follow its predefined temporal programming in the standard queues. This segmentation isolates event-triggered traffic from interference by scheduled traffic, enabling low-latency transmission without disrupting the temporal consistency of scheduled traffic.
Solution Approach 2:
The patent introduces a priority arbitration mechanism that acts as an intermediary between event-triggered traffic and scheduled traffic. This mediator monitors both traffic types and determines transmission order, allowing event-triggered frames to preempt scheduled traffic when necessary while maintaining the overall temporal structure of scheduled traffic. The intermediary resolves conflicts between the two traffic types, ensuring both low latency for event-triggered traffic and temporal consistency for scheduled traffic.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
A method for managing traffic in a network based upon Ethernet switches, which are, in particular, compliant with the IEEE 802.1Q and TSN communication standards, comprises receiving frames that belong to traffic classes (ST, ET, SR A, SR B, BE) classified on the basis of respective priorities, which comprise an event-triggered traffic class (ET) for high-priority real-time event-triggered received frames. The method comprises: dividing the frames received into queues (10), each queue (10) being dedicated to a respective traffic class in the traffic classes (ST, ET, SR A, SR B, BE); scheduling transmission of the frames contained in the queues (10) on the basis of the respective priorities of the traffic classes (ST, ET, SR A, SR B, BE); and, if frames are present belonging to a traffic class (ST) in the plurality of traffic classes (ST, ET, SR A, SR B, BE) that has a higher priority than frames belonging to the event-triggered traffic class (ET) and when at least one protected time window (PW) occurs, transmitting the frames that belong to the traffic class with higher priority (ST) and blocking transmission of the frames that belong to the remaining traffic classes (ET, SR A, SR B, BE); otherwise, transmitting a frame contained in the queue (10) dedicated to the event-triggered traffic class (ET), and if the queue (10) dedicated to the event-triggered traffic class (ET) is empty, transmitting, on the basis of the scheduling, a frame belonging to one of the remaining traffic classes (SR A, SR B, BE) that have frames with a priority lower than the frames belonging to the event-triggered traffic class (ET).