Dynamic Spatial Anonymization for Vehicle Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current spatial anonymization techniques in cloud environments for vehicle data lack flexibility, fail to adjust to varying traffic densities, result in inaccurate data granularity, neglect sparsely populated areas, and struggle with big data loads, leading to inefficient data processing and deletion of data in low-traffic regions.
Innovation Solution
A dynamic spatial anonymization method using geospatial indexing for adaptive grid sizes and two-level data aggregation, which includes iterative splitting of vehicle data into subsets with varying resolutions and combining data points to enhance privacy and reduce computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If static grid structures with predefined grid sizes are used for spatial anonymization, then the anonymization process is simple to implement, but the data granularity is insufficient and sparsely populated areas are neglected
Solution Approach 1:
The patent implements dynamic spatial anonymization by replacing static grid structures with dynamic grid structures that can adapt their size and configuration based on traffic density. The system dynamically adjusts grid parameters in real-time to match the underlying data distribution, allowing for arbitrary dynamic adjustments based on traffic density while maintaining implementation feasibility through automated detection and adjustment mechanisms.
Solution Approach 2:
The patent changes the parameters of the grid structure dynamically based on detected traffic density. When traffic density varies, the system modifies grid size, resolution, and configuration parameters to optimize anonymization quality. This allows the system to adapt to different spatial patterns and maintain appropriate data granularity across various regions without requiring manual reconfiguration.
2Ease of operation
If predefined grid sizes are used for spatial anonymization, then the system is easier to operate, but it fails to meet required data granularity levels and deletes data from sparsely populated areas
Solution Approach 1:
The patent implements feedback mechanisms that continuously monitor traffic density and data distribution patterns. Based on this feedback, the system automatically adjusts grid parameters to maintain optimal anonymization quality. The feedback loop ensures that the system responds to changing conditions and maintains reliable data quality without requiring manual intervention or complex operational procedures.
3Productivity
If current spatial anonymization techniques are applied to big data loads from large vehicle fleets, then the processing time is excessive, but reducing processing time may compromise anonymization quality
Solution Approach 1:
The patent segments the large dataset into smaller manageable portions based on spatial partitions and traffic density zones. By dividing the big data load into smaller chunks that can be processed in parallel, the system reduces processing time while maintaining anonymization quality. The segmentation allows for distributed processing across multiple computational resources, significantly improving productivity without sacrificing precision.
Data Source
Figure 1a~1c
Figure 2
Figure 3
AI summary
The invention relates to a method for dynamic spatial anonymization of vehicle data in a cloud environment, the method comprising: - collecting vehicle data (D(t)), - spatial partitioning the vehicle data (D(t)) into data subsets (DZ(t)) associated with different geographical areas (Z) of various sizes and comprising a maximal amount (N) of records within each data subset (DZ(t)), - spatial aggregation of the vehicle data (D(t)) within the data subsets (DZ(t)), providing two level aggregation: - aggregation of vehicle data (Di(t)), coming from a single vehicle (10), to a corresponding data point (Di_mean) and - aggregation data points (Di_mean), coming from a group (G) of vehicles (10) comprising a particular number (k) of vehicles (10), to a spatial aggregated data set (Dj_mean).