Route Estimation via Mobility Model and Network Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for geolocating mobile devices, such as GPS and network overlay systems, face challenges in accuracy due to battery degradation, intrusiveness, and noise corruption, especially in non-stationary environments like road networks, where Kalman filtering may yield inconsistent results and fail to account for terrain and road networks.
Innovation Solution
A method that determines a mobility model with transition probabilities between road segments and processes time series network data to estimate the actual route traveled by a mobile device, using a directed graph representation of the road network to correlate communication events with road network data, thereby improving location history determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS is used to determine location data, then location accuracy is improved, but battery life is degraded
Solution Approach 1:
The patent combines multiple location determination methods (network overlay triangulation and GPS) into a hybrid system that leverages the strengths of each approach while mitigating their individual weaknesses, allowing accurate location tracking without continuous GPS usage
2Use of energy by moving object
If network overlay location systems are used to estimate location via triangulation, then battery consumption is reduced, but location accuracy is degraded in non-stationary environments
Solution Approach 1:
The system dynamically adapts the location determination strategy based on mobility detection. When stationary, it uses simple triangulation; when mobility is detected, it transitions to GPS-assisted methods, optimizing both energy consumption and accuracy for each operational state
3Object-affected harmful factors
If Kalman filtering is used to smooth noisy location estimates, then measurement noise is reduced, but consistency is degraded when error terms have non-Gaussian distributions
Solution Approach 1:
The patent transforms the location estimation problem by changing the parameter distribution assumption from Gaussian to non-Gaussian (e.g., using Student's t-distribution or other heavy-tailed distributions) that better models the actual error characteristics of multi-modal location data, thereby improving filter consistency
4Device complexity
If traditional filtering methods are used, then computational simplicity is maintained, but adaptability to terrain and road networks is lost
Solution Approach 1:
The patent introduces an intermediary component (mobility model or map-matching module) that acts as a bridge between the filter and the road network data, allowing the filter to incorporate terrain and road constraints without fundamentally changing the filtering algorithm itself
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A method for determining location history of a mobile device is provided. A plurality of states corresponding to a plurality of reachable road segments associated with a road network is determined based on one or more initial estimated locations of the mobile device. Mobility model defining transition probabilities between the determined plurality of states is generated. Time series network data associated with the mobile device is received from a network data source. The time series network data is associated with a given period of time. The time series network data is processed using the generated mobility model to estimate an actual route most likely travelled by the mobile device during the given time period.