Wireless Positioning via Deep Learning Signal Models

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional wireless communication positioning methods, such as those based on global satellite navigation systems and triangulation, suffer from low accuracy and high power consumption in LOS scenarios and are ineffective in NLOS scenarios.

Innovation Solution

The use of multiple positioning models associated with each base station, obtained through deep learning, to extract positioning information from received signals, and then fusing this information to obtain a higher-precision positioning result for user equipment.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional positioning methods (satellite navigation or triangulation) are used, then positioning can be achieved, but positioning accuracy is low and power consumption is high

Engineering Contradiction:

Improvepositioning accuracyVSAvoidpower consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The patent replaces traditional mechanical positioning methods (satellite navigation and triangulation) with a deep learning-based signal processing system. The positioning model processes reference signals through neural network layers to extract positioning information, substituting the mechanical/geometric calculation approach with an intelligent signal analysis approach that achieves higher accuracy while reducing power consumption.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If triangulation positioning based on geometric properties is used, then positioning can be obtained, but performance deteriorates significantly in NLOS scenarios

Engineering Contradiction:

Improvepositioning accuracyVSAvoidscenario adaptability

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

The patent changes the fundamental parameters used for positioning from geometric properties (angles, distances) to signal characteristics (channel state information, signal features). By processing reference signals through deep learning models that analyze signal parameters rather than relying on geometric calculations, the system adapts to different scenarios including NLOS conditions where geometric methods fail.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent substitutes the geometric calculation mechanism with a neural network-based signal processing mechanism. The positioning model processes reference signals through multiple neural network layers to extract positioning information, replacing the mechanical triangulation approach with an intelligent system that can handle complex propagation environments including NLOS scenarios.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If large fingerprint datasets are used for positioning, then positioning accuracy can be improved, but storage overhead increases significantly

Engineering Contradiction:

Improvepositioning accuracyVSAvoidstorage overhead

Core Design Contradiction:

Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent extracts positioning information directly from reference signals processed through the positioning model, eliminating the need for large fingerprint datasets. Instead of storing extensive environmental fingerprint data, the system extracts and processes only the necessary signal features through neural network layers to obtain positioning results.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent replaces the data-storage-based positioning approach with a signal-processing-based approach. Rather than relying on pre-stored fingerprint datasets, the system uses a positioning model that processes reference signals in real-time through neural network layers, substituting storage-intensive methods with computation-efficient signal analysis.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250193830A1Method for wireless communication, and electronic device and computer-readable storage medium

Publication Date: 2025.06.12 SONY GROUP CORP

AI summary

Provided are a method for wireless communication, and an electronic device and a computer-readable storage medium. The electronic device may comprise a processing circuit which is configured to: obtain a plurality of pieces of positioning information of a user equipment, which are respectively related to each base station in a plurality of base stations, wherein the positioning information, which is related to each base station, of the user equipment is acquired on the basis of a reception signal of a reference signal which is transmitted between the base station and the user equipment, and is acquired by using a positioning model related to the base station, and the plurality of pieces of positioning information of the user equipment have the same form as each other; and on the basis of the plurality of pieces of positioning information of the user equipment, obtain a positioning result of the user equipment.