Beamforming Model Training for Obstruction-Aware Antenna Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing beamforming technologies face challenges in accurately controlling interference and radiation patterns due to obstructions in real spaces, leading to calculation errors and difficulty in setting appropriate antenna element configurations for desired radiation directions.
Innovation Solution
A training method utilizing neural networks to calculate and optimize beamforming settings by simulating radiation patterns, interference, and signal-to-noise ratios, adjusting phases and gains of antenna elements to minimize interference and enhance desired radiation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If beamforming is performed by adjusting phases of radio waves from multiple antenna elements, then signal-to-noise ratio and throughput are improved, but interference control accuracy deteriorates due to obstructions in real space
Solution Approach 1:
The patent creates a virtual replica of the real space environment using point cloud data generated from images. This virtual space model copies the spatial relationships, obstructions, and geometric features of the actual environment, allowing beamforming calculations to be performed in a simplified digital representation rather than directly in the complex physical space with its obstructions
Solution Approach 2:
The patent introduces point cloud data as an intermediary representation between the physical environment and the beamforming calculation system. Instead of directly processing the complex real-space geometry with obstructions, the system uses this intermediate point cloud model to calculate optimal beamforming parameters, effectively mediating between the physical constraints and the computational requirements
2Ease of operation
If antenna element settings are adjusted to control radiation patterns, then desired radiation direction is enhanced, but calculation errors increase due to real space obstructions
Solution Approach 1:
The patent copies the real space environment into a virtual point cloud model, preserving spatial relationships and geometric features while eliminating physical obstructions from the calculation process. This allows accurate calculation of radiation patterns and antenna element settings without the interference of real-world obstacles
Solution Approach 2:
The patent performs preliminary generation of point cloud data from images before conducting beamforming calculations. By pre-processing the environmental representation into a simplified point cloud format, the system prepares the calculation input in advance, avoiding calculation errors that would arise from processing real-space obstructions during the actual beamforming optimization
3Measurement precision
If neural network training is performed using real space data, then model accuracy improves, but processing time increases due to complex calculations
Solution Approach 1:
The patent uses point cloud data as a simplified copy of the real environment for training the neural network. This virtual representation maintains the essential spatial and geometric information needed for accurate beamforming while significantly reducing the computational complexity compared to processing full real-space data, thereby decreasing processing time while preserving model accuracy
Data Source
AI summary
A method includes: executing, for each of beams, first calculation processing of outputting setting information set for an antenna element that forms a beam, from first information for identifying a radiation shape of the beam, second calculation processing of outputting radio wave radiation shape information on a radio wave radiation shape from the setting information, an arrangement of the antenna element, and coordinates of an installation position of the antenna element, and third calculation processing of outputting received power information on a received power of each terminal apparatus from the radio wave radiation shape information of the plurality of beams and transmission path characteristics; and executing fourth calculation processing of outputting reception state information on a reception state of the beam from the received power information of each terminal apparatus, and training processing of training a first model that executes the first calculation processing by using the first information and the reception state information.