Dynamic Clustering for Mobile Location Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for geolocating wireless mobile devices in challenging environments, such as urban canyons, face limitations due to variability in RF measurements, leading to inaccurate location determination, especially when using Network Measurement Reports (NMRs) that include power measurements.
Innovation Solution
The system employs dynamic clustering in observation space, grouping measurement vectors by similarity without constraining them to pre-defined regions or shapes in location space, allowing for flexible association of measurement clusters and improving location accuracy by re-clustering candidate NMRs with existing clusters based on common parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional geolocation methods using RF measurements are used, then location determination can be performed, but measurement variability in challenging environments leads to inaccurate location determination
Solution Approach 1:
The patent segments the measurement space by dividing it into multiple clusters based on similarity of RF measurement vectors. Instead of treating all measurements uniformly, the system partitions them into distinct groups that represent different environmental conditions or spatial regions, allowing more accurate location determination by selecting appropriate clusters for comparison.
Solution Approach 2:
The patent transforms the measurement data by changing its representation parameters. It converts raw RF measurements into clustered measurement vectors that capture essential patterns while reducing variability. This parameter transformation enables the system to maintain location accuracy despite environmental fluctuations by comparing patterns rather than raw values.
2Ease of operation
If pre-defined regions or shapes are used to constrain measurements, then location estimation can be simplified, but flexibility in adapting to dynamic environments is reduced
Solution Approach 1:
The patent implements dynamic clustering where measurement vectors are grouped into clusters based on their similarity without being constrained by fixed geographic boundaries. The clusters are formed dynamically according to the actual measurement patterns observed, allowing the system to adapt to changing environmental conditions while maintaining a structured approach to location estimation.
Solution Approach 2:
The patent introduces a new dimension of analysis by clustering measurements in a transformed space rather than directly in geographic space. By organizing measurements into clusters based on their vector similarity, the system creates an additional organizational layer that simplifies location estimation while preserving environmental adaptability.
3Measurement precision
If dynamic clustering in observation space is implemented, then location accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent performs clustering operations in advance during a calibration phase, organizing measurement vectors into clusters before actual location determination is needed. This preliminary organization stores the computational work done in creating the cluster structure, so that during operational use, the system only needs to compare measurements against the pre-formed clusters rather than performing complex clustering computations in real-time.
Solution Approach 2:
The patent creates representative copies of measurement patterns by forming clusters that capture the essential characteristics of groups of similar measurements. Instead of storing and processing every individual measurement vector, the system uses cluster representatives or centroids that copy the essential information, significantly reducing computational complexity while maintaining accuracy.
Data Source
AI summary
A system and method for determining the location of a mobile device in a communications network. A set of network measurement reports (NMR) may be provided for a region in the network. NMRs in the first set may be clustered as a function of observed measurement information in the NMRs. A candidate NMR for a mobile device may be received, and the candidate NMR may then be associated with a selected cluster of NMRs. The candidate NMR and a second set of NMRs may then be clustered, and a location of the mobile device determined as a function of the clustered candidate NMR and second set of NMRs, where the NMRs in the second set are associated with the selected cluster.