Radar Data Peak Compression for Limited-Link Object Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radar devices generate a large amount of data, particularly in automotive applications, which can exceed the capacity of data links between the device and processing circuits, making it necessary to reduce the data while maintaining the ability to determine object characteristics.
Innovation Solution
A method for generating a compact representation of radar data by determining data peaks in the multi-dimensional representation and compressing radar data samples within a limited neighborhood around these peaks, including requantizing phase components and representing amplitude components with a reduced number of bits, allowing for efficient data reduction without significant loss of information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If joint processing of multiple radar data samples is performed to determine object characteristics, then measurement precision is improved, but device complexity increases due to significant computational load
Solution Approach 1:
The patent segments the multi-dimensional radar data into multiple one-dimensional data series by identifying and separating data peaks corresponding to different reflecting objects. Each data peak and its associated samples form an independent series that can be processed separately, transforming a complex multi-dimensional joint processing problem into multiple simpler one-dimensional processing tasks.
Solution Approach 2:
The patent extracts only the relevant portions of radar data by identifying data peaks and selecting samples within limited neighborhoods around these peaks. This extraction approach removes unnecessary data points that do not contribute to object characteristics determination, reducing the computational load while maintaining measurement precision.
2Measurement precision
If complete multi-dimensional radar data is transmitted to external processing circuits, then measurement precision is improved, but loss of substance increases due to exceeding data link capacity
Solution Approach 1:
The patent extracts and transmits only the essential data elements needed for accurate object characteristics determination. By identifying data peaks and selecting samples within limited neighborhoods around these peaks, the system transmits a compressed representation that contains the critical information while minimizing transmission volume to fit within data link capacity constraints.
Solution Approach 2:
The patent applies different processing and transmission quality levels to different portions of the radar data. Data samples within limited neighborhoods around identified peaks are preserved with high fidelity and transmitted, while other data samples are discarded. This local quality approach ensures that the most important data (around peaks) maintains measurement precision while reducing overall data transmission volume.
3Loss of substance
If data compression is applied to reduce transmission volume, then loss of substance decreases, but measurement precision may deteriorate due to information loss
Solution Approach 1:
The patent strategically extracts and preserves only the data samples that contain essential information for object characteristics determination. By focusing on data peaks and their limited neighborhoods, the compression process retains the most informative data while discarding redundant information, achieving both reduced transmission volume and maintained measurement precision.
Solution Approach 2:
The patent applies high-quality preservation selectively to data samples within limited neighborhoods around identified peaks, while other data samples are compressed or discarded. This local quality approach ensures that the compressed representation maintains sufficient information accuracy for determining object characteristics, as the preserved local regions contain the critical signal information.
Data Source
AI summary
A method for generating a compact representation of radar data, includes determining at least one data peak within a multi-dimensional representation of radar data; and compressing radar data samples of the multi-dimensional representation within a limited neighborhood around the at least one data peak to generate the compact representation.