Neural Line-of-Sight Detection for 5G Positioning Accuracy

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing positioning methods in wireless communication systems, particularly in 5G networks, face challenges in accurately estimating line-of-sight delays and angles of arrival due to weak signals and dense multipath conditions, which affect the precision of location estimation.

Innovation Solution

Utilizing deep neural networks to process channel impulse responses and determine line-of-sight delays and angles of arrival by integrating receiver configuration information and channel state information, enabling improved positioning accuracy even in challenging scenarios.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional signal processing algorithms are used for positioning in 5G networks, then device compatibility and implementation simplicity are maintained, but positioning accuracy deteriorates in weak signal and dense multipath environments

Engineering Contradiction:
Improvepositioning accuracyVSAvoidalgorithm complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces neural networks as an intermediary component between the received signal and the positioning calculation. The neural network processes channel impulse responses and extracts line-of-sight delay and angle of arrival information, serving as a mediator that transforms complex signal processing into more manageable tasks for traditional positioning algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent segments the positioning problem into distinct processing stages: first using neural networks to extract specific parameters (line-of-sight delay and angle of arrival) from channel impulse responses, then feeding these extracted parameters into traditional positioning algorithms. This segmentation allows each component to specialize in what it does best.

Inventive Principle:
Principle #1Segmentation

2Measurement precision

If neural networks are introduced to improve positioning accuracy in challenging environments, then measurement precision improves, but device complexity and computational requirements increase

Engineering Contradiction:
Improveline-of-sight delay estimation accuracyVSAvoidneural network integration complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent designs the neural network to perform multiple functions simultaneously: it processes channel impulse responses, identifies line-of-sight paths, estimates delay values, and determines angle of arrival information. This multi-functionality reduces the need for separate processing modules and simplifies the overall system architecture.

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

Solution Approach 2:

The neural network performs preliminary processing of channel impulse responses before the main positioning calculation. By pre-extracting line-of-sight delay and angle of arrival information, the neural network prepares the data in a form that is more suitable for subsequent positioning algorithms, reducing their computational burden.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If more sophisticated processing of channel impulse responses is applied, then positioning accuracy in multipath environments improves, but processing time and computational energy increase

Engineering Contradiction:
Improveangle of arrival estimation accuracyVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces traditional mechanical signal processing methods (such as peak detection and correlation techniques) with neural network-based processing. This substitution allows for more sophisticated analysis of channel impulse responses without proportionally increasing processing time, as neural networks can parallelize computations efficiently.

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

Data Source

PatentEP4136583B1Neural network based line of sight detection for positioning
Publication Date: 2026.02.18 QUALCOMM INC
  • EP4136583B1 patent drawingFigure 1
  • EP4136583B1 patent drawingFigure 2
  • EP4136583B1 patent drawingFigure 3

AI summary

Techniques are provide for neural network based positioning of a mobile device. An example method for determining a line of sight delay, an angle of arrival, or an angle of departure value, according to the disclosure includes receiving reference signal information, determining a channel frequency response or a channel impulse response based on the reference signal information, processing the channel frequency response or the channel impulse response with a neural network, and determining the line of sight delay, the angle of arrival, or the angle of departure value based on an output of the neural network.