ML Wireless Network Design Using 3D Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional wireless network design relies on manual or semi-manual methods, which are inefficient and two-dimensional, failing to optimize network performance due to reliance on static data and intuition, particularly in urban and rural areas with complex IoT device integration and varying terrain.
Innovation Solution
An intelligent machine learning (ML)-based automated system that utilizes aerial surveillance data from UAS and 3-D mapping to generate predictive models for optimal wireless network design, considering signal quality, obstacle vectors, and network configuration parameters, enabling the selection of optimal tower locations and network configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual or semi-manual network design methods are used, then network design can be implemented with existing tools, but the design efficiency is low and optimization is insufficient
Solution Approach 1:
The patent replaces manual mechanical design processes with machine learning-based automated systems. The ML models analyze network data, predict optimal configurations, and generate designs without human intervention, substituting the mechanical/manual design process with intelligent automated computation.
Solution Approach 2:
The network design system performs self-optimization through ML algorithms that automatically analyze network performance data, identify issues, and generate optimized configurations without external human input. The system serves itself by continuously learning from operational data and improving designs autonomously.
2Area of stationary object
If two-dimensional surface area coverage is used, then ground space requirements can be met, but three-dimensional signal coverage and terrain adaptation are insufficient
Solution Approach 1:
The patent transitions from two-dimensional ground space planning to three-dimensional network design by incorporating elevation data, building heights, and vertical signal propagation. The ML models analyze 3D terrain and structure data to optimize tower placements and antenna orientations for improved signal coverage in vertical dimensions.
3Loss of information
If static data and engineer intuition are used for design, then design decisions can be made with available information, but optimal network performance cannot be achieved
Solution Approach 1:
The system implements continuous feedback loops where ML models analyze real-time network performance data, compare actual results with predicted outcomes, and automatically adjust designs to improve performance. This feedback mechanism enables the system to learn from operational data and continuously optimize network configurations.
Solution Approach 2:
The patent uses ML models to perform preliminary analysis and prediction of network performance before actual deployment. The system simulates various design scenarios and predicts outcomes in advance, allowing optimization decisions to be made based on data-driven forecasts rather than waiting for post-deployment performance evaluation.
Data Source
AI summary
A system for automated ML-based design of a wireless network. The system includes a processor and a memory on which are stored machine-readable instructions that when executed by the processor, cause the processor to: receive 3-D mapping data corresponding to a target area comprising a plurality of discretized 3-D units derived from a point cloud dataset; determine a potential installation location within the 3-D mapping data for evaluating a theoretical performance of a potential network tower; generate a plurality obstacle vectors representative of signal paths; and provide the plurality of obstacle vectors to a machine learning module for generating a predicted signal quality for each obstacle vector of the plurality of obstacle vectors based on codified base relationships characterizing known profiles and behaviors of wireless signals and other electronic elements.