Local Network Traffic Prioritization Using Device-Aware ML

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing local networks face challenges in efficiently prioritizing device traffic due to user inability to manage network linking devices, leading to inefficient bandwidth utilization and potential device processing errors, especially with the addition of new devices requiring re-prioritization.

Innovation Solution

A machine-learning based device prioritization model that collects traffic data from multiple local networks to determine device types and priority values, training a model to automatically prioritize network traffic based on device types and usage patterns, reducing the need for manual rule-based approaches.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual traffic prioritization is implemented by users, then device traffic priority can be controlled, but user complexity and difficulty of operation increase significantly

Engineering Contradiction:
Improvetraffic prioritization controlVSAvoiduser management complexity
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The system performs self-service by automatically discovering devices on the network, collecting traffic data, determining device types, and generating prioritization schemes without user intervention. The router autonomously monitors network traffic patterns and adjusts QoS parameters based on learned device behaviors and usage patterns.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical configuration operations with automated electronic/machine learning systems. Instead of users manually configuring QoS rules through complex interfaces, a machine learning model automatically analyzes traffic data and generates prioritization schemes, substituting human操作 with automated intelligent systems.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If multiple prioritization schemes are implemented through trial and error, then optimized QoS can be achieved for particular devices, but time consumption and network inefficiency increase

Engineering Contradiction:
Improvequality of service optimizationVSAvoidtrial and error time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by proactively collecting and analyzing traffic data from all devices, pre-determining device types and usage patterns before QoS optimization is needed. The machine learning model continuously learns network behavior patterns in advance, so when prioritization is required, optimized schemes are already prepared based on pre-analyzed data.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system implements continuous feedback loops where traffic data is constantly monitored, analyzed, and used to refine prioritization schemes. QoS parameters are adjusted based on feedback from actual network performance and traffic patterns, enabling dynamic optimization without manual trial and error iterations.

Inventive Principle:
Principle #23Feedback

3Adaptability or versatility

If network devices are added to the local network, then network functionality is enhanced, but re-prioritization is required leading to bandwidth utilization inefficiency

Engineering Contradiction:
Improvenetwork device additionVSAvoidbandwidth utilization efficiency
Core Design Contradiction:
Adaptability or versatilityVSProductivity

Solution Approach 1:

The patent implements dynamic prioritization where QoS parameters are not static but continuously adapt to changing network conditions and device additions. When new devices join the network, the system dynamically discovers them, analyzes their traffic patterns, and automatically integrates them into the prioritization scheme, maintaining optimal bandwidth utilization without manual reconfiguration.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The machine learning model serves multiple functions: device discovery, device type determination, traffic pattern analysis, prioritization scheme generation, and continuous optimization. This multi-functional system handles device additions, removals, and priority changes universally through a single automated framework, eliminating the need for separate re-prioritization processes.

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

4Productivity

If machine learning model is used for auto-prioritization, then bandwidth efficiency is improved, but device complexity and training requirements increase

Engineering Contradiction:
Improvebandwidth efficiencyVSAvoidmodel training complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary layer between raw network traffic and QoS control decisions. The machine learning model acts as a mediator that automatically processes traffic data, determines device types, and translates complex traffic patterns into actionable prioritization rules, shielding users from the complexity of model training while enabling sophisticated bandwidth management.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3910885B1Auto-prioritization of device traffic across local network
Publication Date: 2026.02.04 GOOGLE LLC
  • EP3910885B1 patent drawingFigure 1
  • EP3910885B1 patent drawingFigure 2
  • EP3910885B1 patent drawingFigure 3

AI summary

A device prioritization model (130) receives information about a set of traffic data for devices (204) in a local network as input and outputs at least one prioritization scheme (210). Each prioritization scheme (210) includes optimized device bandwidth allocation parameters, comprising at least one or more traffic priority values PN for each network connected device DN. The prioritization scheme is applied to a local network with the following devices: a coffeemaker, a credit card reader, and a security system. The prioritization scheme could include multiple traffic priority values for each of the devices. For a security system having two types of alerts, one "normal operation" and one "emergency situation," the traffic from the "normal operation" may be assigned a lower traffic priority value by the prioritization scheme, while the "emergency situation" may be assigned a higher traffic priority value by the prioritization scheme. The lower and higher traffic priority values assigned to the security system are relative to the traffic priority values of the other devices (e.g., the coffee maker and the credit card reader) which each have at least one traffic priority value assigned by the prioritization scheme. Training of the device prioritization model (130) may incorporate one or more features (206) for producing a prioritization scheme function, PS (208).