Network Switch Circuit Priority Queue Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing network switches struggle with efficient prioritization of data packets, particularly in vehicle communication systems, where high-priority control-data traffic is sensitive to latency, and current solutions often require complex protocols and additional circuitry, leading to susceptibility to failure and increased costs.

Innovation Solution

A two-stage prioritization scheme is implemented using multiple queues, where data packets are assigned to high-priority or low-priority groups based on their traffic class and MAC address, allowing for granular prioritization and sorting to minimize latency and reduce processor busy waits.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a FIFO methodology is used to allocate data packets, then the allocation is simple and follows first-come-first-served order, but high-priority control-data traffic experiences increased latency and cannot be prioritized over low-priority traffic

Engineering Contradiction:
Improvepriority allocation reliabilityVSAvoiddata packet latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the single FIFO queue into multiple priority-based queues (e.g., high-priority queue and low-priority queue). This segmentation allows the system to differentiate between critical control-data traffic and non-critical traffic, ensuring that high-priority packets are transmitted first without being blocked by low-priority packets, thereby reducing latency for time-sensitive communications.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the allocation parameter from a single FIFO order to multiple priority levels. By introducing priority as a new parameter for queue selection, the system can dynamically select which queue to service based on the urgency of the data, transforming the static FIFO allocation into a dynamic priority-based allocation mechanism that reduces latency for critical traffic.

Inventive Principle:
Principle #35Parameter changes

2Reliability

If complex protocols and additional communications circuits are used to achieve priority allocation, then data packet prioritization can be implemented, but the device complexity increases and susceptibility to failure increases

Engineering Contradiction:
Improvedata packet prioritizationVSAvoidcircuitry complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Instead of implementing complex priority protocols, the patent segments the transmission medium into multiple logical queues based on priority levels. This segmentation approach simplifies the control logic by using straightforward queue management rather than complex protocol handshaking, reducing device complexity while maintaining reliable priority allocation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent enables packets to self-identify their priority level through inherent packet headers or classification mechanisms, eliminating the need for complex external protocol negotiations. Each packet carries its own priority information, allowing the system to automatically route it to the appropriate queue without requiring complex control circuits or protocol processing.

Inventive Principle:
Principle #25Self-service

3Loss of time

If multiple queues are used for priority allocation, then high-priority traffic can be transmitted first, but the queue management complexity and processor busy waits increase

Engineering Contradiction:
Improvecontrol-data latencyVSAvoidqueue management complexity
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

The patent segments queues into a limited number of priority levels (e.g., 2-4 queues rather than numerous fine-grained queues). This segmentation strikes a balance between providing sufficient priority differentiation for control-data traffic while keeping queue management complexity manageable. The segmented approach reduces processor busy waits by limiting the number of queues that need to be monitored and serviced.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the queue management approach by using priority-based selection criteria rather than complex round-robin or weight-based scheduling. This parameter change simplifies the queue management logic to a straightforward priority comparison, reducing processor overhead and busy waits while still achieving low latency for high-priority traffic.

Inventive Principle:
Principle #35Parameter changes

4Productivity

If priority allocation is implemented without granular sorting, then the system is simpler, but the prioritization granularity is insufficient for fine-tuned control of different data packets

Engineering Contradiction:
Improveprioritization granularityVSAvoidsorting circuitry complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the prioritization process into two stages: first segmenting packets into broad priority queues, then applying granular sorting within each queue based on secondary criteria (such as destination address or packet type). This two-stage segmentation provides fine-tuned prioritization granularity without requiring complex single-stage sorting circuitry, as each stage handles a specific aspect of priority differentiation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds another dimension to priority allocation by implementing sorting within queues based on secondary criteria (such as destination MAC address). This dimensional approach transforms single-dimensional priority allocation into multi-dimensional prioritization, enabling fine-grained control over packet transmission order while keeping the base queue structure simple and manageable.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP3079316B1Network switch circuit, system and method
Publication Date: 2018.07.04 NXP BV
  • EP3079316B1 patent drawingFigure 1
  • EP3079316B1 patent drawingFigure 2
  • EP3079316B1 patent drawingFigure 3A

AI summary

Data packets are received from a plurality of communication nodes within a local Ethernet network, the data packets including a media access control address ("MAC address") are indicative of a destination and a data packet traffic class. Based upon the traffic class of each data packet, each of the received data packets are assigned in to one of the plurality of queues of a memory circuit. Based upon the MAC address of each data packet, the data packets within at least one of the queues are sorted. Each queue is then serviced and the data packets within are transmitted based upon the sorting.