Gaussian Sensor Data Buffering for Low-Memory Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing sensor data points require significant memory and computing resources, particularly in environments with large volumes of raw data, limiting the scalability and efficiency of data processing hardware.
Innovation Solution
A computer-implemented method using Gaussian distributions to process sensor data points by initializing a buffer with initial distributions, updating or merging them based on distance thresholds, creating a predetermined data structure that preserves data distribution, thereby reducing memory and computing requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of raw sensor data points are processed and stored to preserve data distribution, then measurement precision is improved, but memory usage and computing resources increase significantly
Solution Approach 1:
The patent transforms the representation parameters of sensor data from individual raw points to Gaussian distribution parameters (mean, covariance, amplitude). This parameter transformation allows preserving the essential distribution characteristics of sensor data while dramatically reducing the quantity of data that needs to be stored and processed, directly resolving the contradiction between measurement precision and memory usage
Solution Approach 2:
Instead of storing and processing actual raw sensor data points, the patent creates a simplified copy in the form of Gaussian distribution models that replicate the essential statistical properties of the original data. This copying approach maintains the distribution information needed for accurate processing while using minimal memory resources
2Measurement precision
If all raw sensor data points are processed in each cycle to maintain data accuracy, then measurement precision is improved, but computing power requirements increase
Solution Approach 1:
The patent changes the computational parameters from processing individual data points to processing Gaussian distribution parameters. The updating rule that computes new Gaussian parameters from previous parameters and new sensor points requires significantly fewer computational operations than processing all raw points, thus reducing computing power requirements while maintaining processing accuracy
Solution Approach 2:
The patent segments the large set of raw sensor data points into a small number of Gaussian distribution components. Each component represents a cluster of similar data points, allowing the system to process only the essential representative parameters rather than every individual point, thereby reducing computational load while preserving accuracy
3Measurement precision
If dense raw data points from static objects are processed to create environment maps, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent transforms the data representation from dense raw points requiring complex hardware processing to compact Gaussian parameters that can be handled by simpler systems. The updated Gaussian distributions efficiently represent environment features while reducing the computational complexity of the hardware architecture needed for environment mapping
Data Source
AI summary
This disclosure relates to a computer-implemented method for processing sensor data points by means of a data processing hardware where a distributions buffer for storing an initial plurality of Gaussian distributions is initialized, each Gaussian distribution of the initial plurality of Gaussian distributions including an initial plurality of sensor data points received sequentially from at least one sensor and having an associated predetermined distribution distance threshold, and where the distributions buffer is sequentially updated for a number n of new sensor data points based on a distribution distance condition, generating an updated plurality of Gaussian distributions including either a new Gaussian distribution, or an updated single Gaussian distribution or an updated merged Gaussian distribution.