Multi-Modal Kalman Filter for Mobile Device Location Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing location determination methods in mobile devices often yield inconsistent and erratic location estimates due to the use of single Kalman filters on multi-modal location data from different sources, such as GPS and wireless access points, leading to inaccurate positioning.
Innovation Solution
Implementing a multi-modal Kalman filter that maintains multiple location approximations, each determined by filtering a subset of location estimates using a respective Kalman filter, and designates one as active based on consistency with recent data, allowing for smoother and more consistent location estimation by switching between approximations as data quality changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single Kalman filter is used to process location estimates from multiple sources, then the device complexity is reduced, but the location estimation becomes inconsistent and erratic
Solution Approach 1:
The patent divides the location estimation problem into multiple independent Kalman filters, each handling a specific subset of location data sources (e.g., GPS-only, Wi-Fi-only, cellular-only). This segmentation allows each filter to process homogeneous data with consistent characteristics, avoiding the inconsistency that arises when a single filter processes heterogeneous multi-modal data together.
2Reliability
If multiple independent Kalman filters are used to process different subsets of location data, then the location estimation consistency is improved, but the device complexity increases
Solution Approach 1:
The patent implements a dynamic selection mechanism that chooses which Kalman filter approximation to use based on current data quality and availability. The system dynamically adjusts the active filter based on real-time conditions such as GPS signal strength, Wi-Fi availability, and cellular network status, allowing the system to maintain consistency while adapting to changing environments without requiring all filters to operate simultaneously at full complexity.
3Ease of manufacture
If location estimates from different sources are averaged, then the computation is simple, but the location accuracy deteriorates when estimates differ significantly
Solution Approach 1:
Instead of averaging all location estimates together, the patent segments the estimates by source type and processes them through separate Kalman filters. This allows the system to maintain computational simplicity within each homogeneous group while avoiding the accuracy degradation that occurs when significantly different estimates from heterogeneous sources are averaged together.
Data Source
AI summary
Methods and systems for determining a location of a mobile device using a multi-modal Kalman filter are described. According to an example method, a mobile device may maintain multiple approximations of a location of a mobile device. Each approximation includes an estimated geographic location of the mobile device that is determined by filtering a respective subset of location estimates received by the mobile device using a respective Kalman filter, and one of the multiple approximations is designated as an active approximation. The method also involves receiving data indicating an estimate of a geographic location of the mobile device and, based on a distance between the estimate of the geographic location and a given approximation of the multiple approximations, updating the given approximation using the estimate of the geographic location. Additionally, the method involves providing for display a visual indication of an estimated geographic location associated with the active approximation.