Application-Aware Queue Management for Packet Latency Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing queue management systems fail to effectively manage packet priorities to meet the timeliness constraints of various data types, leading to network congestion and latency issues in wireless communication networks.

Innovation Solution

Implementing a latency controller that classifies packets based on their application type, assigns arrival and departure time stamps, and uses separate queues with latency budgets to prioritize high-priority packets, ensuring timely transmission.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If packets are enqueued for transmission using simple priority queues, then the queue management is simple, but the timeliness constraints of different data types cannot be met

Engineering Contradiction:
Improvequeue management simplicityVSAvoidtimeliness constraint satisfaction
Core Design Contradiction:
Ease of operationVSReliability

Solution Approach 1:

The patent segments the single queue into multiple classification queues (first classification queue, second classification queue, third classification queue) based on different timeliness constraints. Each queue handles packets with specific latency requirements, allowing differentiated service while maintaining manageable structure. This resolves the contradiction by segmenting the queue management function to satisfy diverse timeliness constraints without overwhelming complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic queue management where the controller selectively activates different queues based on packet characteristics and current network conditions. The system dynamically adjusts which queue a packet enters based on its timeliness constraints, enabling flexible adaptation to varying network demands while maintaining operational simplicity through automated decision-making.

Inventive Principle:
Principle #15Dynamics

2Reliability

If multiple classification queues are implemented to meet different timeliness constraints, then the timeliness constraints are satisfied, but the device complexity increases

Engineering Contradiction:
Improvetimeliness constraint satisfactionVSAvoidqueue management complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The packet classification and queue selection process is automated through the controller, which makes self-service decisions about packet routing based on pre-configured timeliness constraints. The system self-manages packet distribution across queues without requiring complex external control mechanisms, reducing operational complexity while maintaining multiple classification queues for different data types.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The controller serves multiple functions: it classifies packets, determines queue assignments, manages transmission timing, and coordinates with the transmission scheduler. This multi-functionality consolidates what would otherwise require separate complex components into a single controller, managing the complexity of multiple queues through integrated control rather than separate management systems.

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Measurement precision

If arrival time stamps and departure time stamps are tracked for each packet, then precise latency control is achieved, but the processing overhead increases

Engineering Contradiction:
Improvelatency measurement precisionVSAvoidpacket processing throughput
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The system sets departure time stamps in advance based on latency budgets and timeliness constraints, rather than calculating them dynamically during transmission. This preliminary action allows the controller to prepare timing information ahead of time, reducing real-time processing overhead while maintaining precise latency control. The transmission scheduler then simply needs to execute these pre-calculated time stamps.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements efficient timestamp generation where the controller rapidly assigns departure time stamps based on pre-determined latency budgets without extensive calculation for each packet. The transmission scheduler then rushes through the packet transmission process using these pre-computed time stamps, minimizing additional processing overhead while achieving precise latency measurement and control.

Inventive Principle:
Principle #21Skipping (Rushing through)

Data Source

PatentEP4645818A1Systems for and methods for queue management for latency control
Publication Date: 2025.11.05 AVAGO TECHNOLOGIES INTERNATIONAL SALES PTE LTD
  • EP4645818A1 patent drawingFigure 1A
  • EP4645818A1 patent drawingFigure 1B
  • EP4645818A1 patent drawingFigure 1C

AI summary

Queue management for latency control is provided. A method includes receiving packets of multiple flows, each of the packets including an arrival time stamp. The method can include classifying, based on a type of application, the flows for transmission as higher priority via a higher priority channel, or lower priority via a lower priority channel. The method can include determining, for packets for the flows classified as higher priority, a queue of one or more second queues (325) to use for transmission and a departure time stamp for each of the packets according to a latency budget for the type of application. The method can include communicating the higher priority packets with the departure time stamp to a corresponding queue of the one or more second queues (325) based on the type of application. A transmission scheduler (340) can prioritize transmission of packets from the second queues (325) based on the departure time stamp.