Wireless Network Optimization via Client-Access Point Benchmark Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing WLAN deployments face performance gaps due to manual optimization limitations, which fail to account for client device perspectives on radio frequency conditions, leading to skewed network views and inefficiencies.
Innovation Solution
An automated system that uses sensors to collect network benchmarks from both access points and client devices, mapping these in an n-dimensional space to identify outlier dimensions and dynamically reconfigure access points for improved alignment and performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual optimization is used to configure access points, then administrators can control network settings, but the process is time-consuming and fails to discover better configurations
Solution Approach 1:
The system enables automated self-optimization by having access points autonomously adjust their configurations based on collected benchmark data and machine learning models, eliminating the need for manual administrator intervention while continuously improving network performance
Solution Approach 2:
The system implements continuous feedback loops where benchmark data from access points and sensor devices is collected, analyzed by machine learning models, and used to generate configuration recommendations that are applied and monitored for further optimization
2Extent of automation
If metric-based optimizations are implemented using access point data, then some automation is achieved, but performance gaps remain due to not accounting for client device characteristics
Solution Approach 1:
The system segments the network perspective into two distinct views: the access point perspective (ceiling-level) and the client device perspective (floor-level), collecting and analyzing benchmark data from both sources to identify and resolve dimensional misalignments that affect performance
Solution Approach 2:
Sensor devices are introduced as intermediary elements that collect benchmark data from client devices and transmit it to the machine learning model, enabling the system to incorporate client-side characteristics without requiring direct access to client devices
3Ease of manufacture
If access points are configured from ceiling-level perspective only, then deployment is simplified, but the network view becomes skewed and does not reflect client device experience
Solution Approach 1:
The system applies local quality by treating different spatial locations (ceiling-level access points and floor-level client devices) as having different measurement characteristics, collecting and analyzing benchmarks specific to each location to create an accurate holistic network view
Data Source
AI summary
Techniques for wireless optimizations are provided. A first set of values for a set of network benchmarks is received from a sensor device, where the set of network benchmarks define network performance dimensions, and include a number of retries needed to successfully transmit data. A second set of values for the set of network benchmarks is received from a wireless access point. The first set of values is mapped against the second set of values in an n-dimensional space, and an outlier dimension is identified, based on analyzing the mapped first and second sets of values. The wireless access point is then reconfigured, based on the identified outlier dimension.