Radar Data Memory Layout for Faster Range and Velocity Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Prior art radar systems face limitations in processing large volumes of data efficiently due to insufficient high-speed local memory, leading to slow data transfer between local and remote memory, which hampers the overall speed of generating range data.
Innovation Solution
A method and apparatus for FMCW radar that transfers radar data across local and remote memory in a repeating pattern of single data lines for each chirp, storing data in blocks to facilitate fast transfer back to local memory, using a grid pattern in remote burst access memory to maximize efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large volumes of radar data are processed using conventional memory transfer methods, then measurement precision is improved, but processing speed deteriorates due to slow data transfer between local and remote memory
Solution Approach 1:
The patent segments the large volume of radar data into individual data lines corresponding to each chirp. Each data line is processed and transferred separately rather than moving entire data blocks, which reduces the total transfer time while maintaining the precision required for accurate range measurement.
Solution Approach 2:
The patent performs preliminary processing of radar data in the local memory to identify and prepare only the necessary data lines for transfer to remote memory. This preliminary action ensures that high-precision measurements are maintained while minimizing the amount of data that needs to be transferred, thus improving processing speed.
2Productivity
If more local storage memory is provided to avoid data transfer, then processing speed is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts only the essential data lines from local memory that are needed for subsequent processing, rather than keeping all data in local memory. This extraction approach maintains processing speed by minimizing transfers while avoiding the need for expanded local memory capacity or more complex memory architectures.
Solution Approach 2:
The patent implements periodic transfer of data lines from local to remote memory in a repeating pattern, where data is transferred in manageable intervals rather than all at once or never. This periodic action maintains processing throughput while using a simple, fixed memory architecture without requiring additional memory capacity.
3Loss of time
If data is transferred in large blocks between memory systems, then transfer efficiency is improved, but the need for local storage memory increases
Solution Approach 1:
The patent segments the data transfer into individual data line units, each corresponding to a single chirp. This segmentation allows for efficient periodic transfer of small data units rather than requiring large block transfers, thus reducing the local storage memory capacity needed while maintaining acceptable transfer efficiency.
Solution Approach 2:
The patent transfers only the partial amount of data that is immediately needed for processing (one data line per chirp cycle) rather than transferring excessive amounts of data in large blocks. This partial action reduces memory requirements while keeping transfer time acceptable for the application.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach reduces the need for local storage memory while enabling faster data processing by allowing efficient data transfer and subsequent Fourier Transforms, thereby enhancing the speed and efficiency of radar data generation.
Implementation Method 1
This can be produced relatively simply using voltage controlled oscillator (VCO)
Implementation Method 2
These data samples may then be processed using a discrete Fourier Transform to determine values for a set of range bins for the chirp
Implementation Method 3
Distance can be measured from the phase difference that is seen in the returned signal when the emitted signal is modulated by a small step change in frequency
Implementation Method 4
Similarly, the relative velocity can be measured by looking at the change in phase shift between closely spaced chirps
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of operating a radar apparatus is disclosed. The method comprises: (a) capturing a first chirp of data in the cycle which generates D data samples from each of N antennas and storing the data in a local storage memory; (b) processing the captured data using a Fourier Transform to generate a set of range bin data, one for each antenna, where each set comprises R range bins; (c) transferring in a sequence of bursts the range bin data for the chirp from the local storage memory to a remote burst access memory where the data is arranged in a grid pattern, the grid pattern comprising data filled rows containing two or more unique range bins of data for the chirp held in a continuous strip of the remote burst access memory, the data filled rows being separated vertically from adjacent data filled rows by multiple rows which do not contain any bursts of data associated with the chirp, (d) repeating the steps (a) and (b) for each subsequent chirp of data in the cycle, (e) repeating step (c) after the range bin data for each subsequent chirp has been stored in the local storage memory so as to transfer in a sequence of bursts the range bin data for each subsequent chirp of data from the local storage memory to the remote burst access memory where the data is arranged in the same grid pattern used for step (c) but offset vertically by one or more rows to fit within the rows in the column that have not been written with data for any previously processed chirp in the cycle, (f) transferring from the remote burst access memory into the local storage memory, in one or more bursts, at least one continuous vertical block of data from the remote block that has a length equal to the number of different range bins, (g) process the data transferred into the local storage memory using a Fourier Transform to generate a set of velocity data comprising a set of velocity bins where each velocity-bin is generated from the values of the corresponding range bin for every antenna of the array, and repeating steps (f) and (g) until all of the data in the remote burst access memory block of data has been processed by the second FFT.