Wireless Localization Using Signal Strength Change Patterns
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless localization techniques face challenges in achieving high accuracy indoors due to environmental changes, signal interference, and limitations in using short-range and wide-area wireless communication signals, particularly for vehicle navigation and autonomous driving, where GNSS is ineffective.
Innovation Solution
A method and apparatus that estimate a moving node's position by measuring signal strength from fixed nodes, generating a change pattern of signal strength over time, and using this pattern to estimate both relative and absolute positions, reducing the impact of environmental changes and enabling accurate localization with signals like LTE that have minimal strength variation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the triangulation technique is used to estimate position by measuring RSS from three or more access points, then the localization can be performed in indoor space, but the converted distance value includes large error due to attenuation, reflection, and diffraction of radio signals by walls and obstacles
Solution Approach 1:
The patent converts the harmful effects of signal attenuation, reflection, and diffraction into beneficial features by using them to create unique fingerprint patterns for different locations. Instead of trying to eliminate these effects, the system captures the actual signal characteristics (including reflections and attenuations) as they occur in the environment and uses them as identifiers for specific positions, thereby transforming measurement errors into useful localization data
Solution Approach 2:
The patent replaces the triangulation method (which relies on geometric calculations and distance conversions) with a fingerprint-based pattern matching approach. Instead of converting signal strength to distance and then to position through mathematical models, the system directly compares measured signal patterns with pre-stored fingerprint patterns to determine location, bypassing the error-prone distance conversion step entirely
2Measurement precision
If the fingerprint technique divides the indoor space into a grid structure and collects signal strength values in each unit area, then the localization accuracy can be increased to 2 to 3 meters, but the technique cannot adapt to changes in the wireless environment such as signal interference, expansion of access points, or occurrence of failures
Solution Approach 1:
The patent implements a dynamic fingerprint matching system that can adapt to environmental changes. Instead of using a static grid-based approach with fixed thresholds, the system dynamically adjusts matching criteria based on the number and strength of matching fingerprints. The patent introduces a flexible matching mechanism that can handle variations in signal environments by adjusting the matching threshold and considering multiple fingerprint matches, allowing the system to maintain accuracy even when access points are added, removed, or experience signal interference
Solution Approach 2:
The patent changes the matching parameters dynamically based on environmental conditions. The system adjusts the matching threshold and the number of required matching fingerprints based on the current wireless environment. When environmental changes are detected (such as new access points or signal interference), the system adapts its matching criteria to maintain localization accuracy, transforming the static fingerprint matching process into a dynamic one that responds to environmental variations
3Adaptability or versatility
If short-range wireless communication signals such as Bluetooth and Zigbee are used for localization, then the localization can be performed in indoor space, but the signals temporarily occur according to user needs and disappear, making them unsuitable for reliable localization
Solution Approach 1:
The patent creates a universal localization system that can work with multiple types of wireless signals (Wi-Fi, Bluetooth, Zigbee, cellular) simultaneously. The fingerprint-based approach is signal-agnostic, meaning it can capture and process fingerprints from any wireless signal type that is available in the environment. This multi-functionality allows the system to leverage whatever signals are present, making it adaptable to different indoor environments while maintaining reliability through the robustness of the pattern matching approach
4Adaptability or versatility
If wide-area wireless communication signals such as LTE are used for localization, then the signals are uniformly distributed in indoor and outdoor spaces, but the area where signal strength change is not large is wide, limiting localization accuracy
Solution Approach 1:
The patent moves from two-dimensional signal strength comparison to three-dimensional fingerprint pattern matching. Instead of relying solely on signal strength values, the system captures multi-dimensional characteristics including signal strength, signal phase, timing information, and spatial relationships between multiple access points. This additional dimensional information provides more discrimination power even when signal strength changes are minimal, enabling accurate localization with wide-area signals like LTE
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach significantly improves localization accuracy by reducing errors from environmental changes and allows for both indoor and outdoor localization, enabling reliable vehicle navigation and autonomous driving systems.
Implementation Method 1
measuring strength of a signal received from a fixed node
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to wireless localization method and apparatus of high accuracy, and measures strength of at least one signal that is transmitted from at least one fixed node, estimates a relative position of a moving node, generates a change pattern of at least one signal strength according to relative changes in positions of the moving node over a plurality of time points from at least one signal strength and the relative position of the moving node, and estimates an absolute position of the moving node, based on a comparison between the change pattern of the at least one signal strength and a map of a distribution pattern shape of signal strength in a region where the moving node is located. Accordingly, it is possible to accurately estimate a position of a moving node using a radio signal which not only accurately estimates the position of the moving node even in a change of wireless environment but also has almost no change in signal strength over a wide region.