Automotive Radar Peak Detection Under SRAM Memory Limits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automotive radar systems face memory constraints due to the reliance on expensive SRAM memory, necessitating a reduction in memory footprint to reduce costs and form factor, particularly in corner radar systems where DRAM is not feasible.
Innovation Solution
Implement a progressive peak detection scheme that iteratively processes radar cube data in subsections, performing R-FFT, D-FFT, and peak detection on each subsection before discarding it, combining candidate peaks into a single dataset using a final pass-criteria to identify true peaks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radar cube data is stored completely in memory for processing, then peak detection accuracy is maintained, but memory requirements increase significantly
Solution Approach 1:
The radar cube data is divided into multiple smaller subsections that are processed iteratively. Each subsection is processed through R-FFT, D-FFT, and peak detection independently, then combined with previous results. This segmentation allows the system to maintain peak detection accuracy while reducing the memory footprint from storing complete radar cubes to only storing current subsections and candidate peak datasets.
2Quantity of substance
If iterative processing of radar cube subsections is implemented, then memory footprint is reduced, but processing time increases
Solution Approach 1:
Candidate peaks are identified and stored in a candidate peak dataset during the iterative processing of subsections. This preliminary identification allows the final peak detection to focus only on verifying and refining these candidate peaks rather than searching through the entire radar cube, thereby reducing the time required for final peak determination while maintaining reduced memory footprint.
3Productivity
If complete radar cubes are processed at once, then processing efficiency is maximized, but memory constraints are violated
Solution Approach 1:
The radar cube is segmented into manageable subsections that fit within the limited SRAM memory capacity. Each subsection is processed efficiently through the complete signal processing chain (R-FFT, D-FFT, peak detection), and results are accumulated. This segmentation enables processing efficiency to be maintained at the subsection level while respecting the memory constraints of the radar system.
Solution Approach 2:
After each subsection is processed and candidate peaks are identified, the processed subsection data is discarded from memory to free up space for the next subsection. Only the lightweight candidate peak dataset is retained. This approach allows the system to process data efficiently without being constrained by total memory capacity, as each subsection is handled and then removed from memory.
Data Source
AI summary
An automotive radar system includes at least one transmitter and at least one receiver and a processor configured to receive, from the at least one receiver, received radar signals, generate a first subsection of a radar cube using the received radar signals, detect a first set of candidate peaks in the first subsection of the radar cube, generate a second subsection of the radar cube using the received radar signals, detect a second set of candidate peaks in the second subsection of the radar cube, determine a set of locations in a candidate peak dataset, wherein each location in the set of locations is associated with candidate peaks in both the first set of candidate peaks and the second set of candidate peaks, and estimate a direction of arrival of an object using the candidate peaks associated with the set of locations.