Adaptive Wireless Client Steering via Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless networks face challenges in efficiently transitioning wireless client devices between different radios within an access point to optimize communication quality, as existing steering methods do not effectively consider client device characteristics and radio environment factors.
Innovation Solution
The implementation of a machine learning model that generates device steering rules based on client device characteristics, client state information, and radio environment characteristics, allowing for intelligent steering of client devices between radios, with the ability to update the model using roaming records for improved performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing steering methods are used without considering client device characteristics and radio environment factors, then the steering process is simple, but the quality and success rate of device transitions are insufficient
Solution Approach 1:
The patent changes the parameters considered in steering decisions from basic connection status to multiple factors including client device characteristics (device type, mobility pattern), client state information (data rate, packet loss), and radio environment characteristics (interference level, signal strength). This multi-parameter approach enables intelligent steering that adapts to varying conditions, resolving the contradiction between simple steering processes and high-quality transitions.
Solution Approach 2:
The patent implements feedback mechanisms by continuously monitoring steering outcomes and using this information to refine future steering decisions. The system learns from actual transition results and adjusts its behavior accordingly, improving the success rate of device steers while managing the complexity through data-driven optimization rather than overly complex predetermined rules.
2Adaptability or versatility
If machine learning model is used to generate device steering rules based on multiple characteristics and state information, then the adaptability and quality of steering decisions are improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent performs preliminary actions by pre-processing and organizing client device characteristics, state information, and radio environment data into structured formats before feeding them to the machine learning model. This preparation step reduces the computational burden during actual steering decisions, enabling the system to handle varied conditions adaptively while managing processing requirements through efficient data organization and feature extraction.
3Measurement precision
If real-time data is incorporated into steering decisions, then the responsiveness and accuracy of transitions are improved, but the data processing overhead and system latency increase
Solution Approach 1:
The patent applies partial action by selectively collecting and processing only the most relevant real-time data for each steering decision rather than continuously processing all available data. The system determines which characteristics and state information are most pertinent to the current situation and focuses computational resources on processing those specific elements, achieving accurate steering decisions while minimizing unnecessary processing time and overhead.
Data Source
AI summary
An example network information aggregator is disclosed. The network information aggregator includes a network interface, a memory, and processing circuitry. The processing circuitry is to receive a machine learning model trained using initialization data from a model creation device. The processing circuitry is also to generate a device steering rule to steer a client device from a first radio to a second radio using the machine learning model. The processing circuitry is also to send a steer command to the first radio.