Neural Network RSS Fingerprint Deployment for Indoor Navigation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Indoor navigation systems face challenges in maintaining accurate and up-to-date fingerprint maps due to the dynamic nature of indoor infrastructure, requiring time-consuming and error-prone manual calibration efforts.

Innovation Solution

A neural network training system that integrates a postulated mathematical model of wireless received signal strength (RSS) features with machine learning methods, allowing for crowd-sourced training and deployment of an RSS fingerprint dataset, reducing the need for manual supervision and improving accuracy in indoor positioning.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual calibration efforts are used to generate and maintain the fingerprint map, then the positioning system can be calibrated, but the process is time-consuming and error-prone

Engineering Contradiction:

Improvepositioning accuracyVSAvoidcalibration time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system enables self-service by allowing mobile devices to automatically contribute RSS measurements to the fingerprint database without manual intervention. The neural network automatically processes these measurements and updates the fingerprint map, eliminating the need for manual calibration efforts while maintaining positioning accuracy.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical calibration process with an automated neural network system. The neural network automatically processes RSS measurements from multiple mobile devices, generates fingerprint profiles, and updates the database, substituting human-operated mechanical calibration with an automated computational system.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If manual calibration is used to maintain the fingerprint map, then positioning data can be updated, but the process is error-prone

Engineering Contradiction:

Improvefingerprint map accuracyVSAvoidcalibration complexity

Core Design Contradiction:

ReliabilityVSEase of operation

Solution Approach 1:

The system performs self-updates by automatically processing RSS measurements from mobile devices through the neural network. The fingerprint map maintains itself by continuously incorporating new measurements and updating existing profiles without human intervention, thereby improving reliability while simplifying operation.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces error-prone manual calibration operations with an automated neural network system that processes measurements, detects patterns, and updates the fingerprint database automatically. This substitution eliminates human errors while reducing operational complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Device complexity

If traditional RSS modeling is used, then the system can function with minimal data, but the model is an over-simplification of indoor space complexities

Engineering Contradiction:

Improvesystem simplicityVSAvoidRSS modeling accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent transforms the RSS modeling approach by changing the parameters from simple distance-based calculations to complex neural network parameters that learn from actual measurements. The neural network adjusts weights and biases based on training data, enabling accurate modeling of indoor space complexities while maintaining system functionality.

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent creates a composite modeling approach by combining traditional RSS measurement data with neural network learning capabilities. The system integrates multiple data sources and processing methods to create a more accurate and comprehensive model that captures the complexities of indoor environments.

Inventive Principle:

Principle #40Composite materials

4Speed

If a neural network trained model is deployed, then convergence speed improves, but data and processor resources are required

Engineering Contradiction:

Improvemodel convergence speedVSAvoiddata resources

Core Design Contradiction:

SpeedVSQuantity of substance

Solution Approach 1:

The patent applies partial action by deploying the neural network model incrementally. The system processes and stores RSS measurements from mobile devices, training the model progressively rather than requiring all data upfront. This approach enables faster convergence while managing data resource requirements through staged implementation.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS10716089B1Deployment of trained neural network based RSS fingerprint dataset

Publication Date: 2020.07.14 MAPSTED CORP

AI summary

A method and system of deploying a trained neural network-based RSS fingerprint dataset for mobile device indoor navigation and positioning. The method comprises: based on RSS parameters acquired from a plurality of mobile devices acquired at a set of positions within an indoor area, accumulating the RSS parameters as a trained neural network-based RSS fingerprint dataset in a fingerprint database of the indoor area; and when a density of points represented by the set of positions having accumulated RSS parameters exceeds a deployment threshold density, deploying the RSS fingerprint dataset within a fingerprint map for mobile device navigation of the indoor area, the fingerprint map encompassing the set of positions.