Radar Sensor Chip Preprocessing for Large Antenna Arrays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional radar systems face challenges in processing large-scale antenna array data due to limitations in computing power, cache capability, and data throughput, particularly when using a central processing unit, which restricts adaptability and compatibility with larger-scale radar sensors.
Innovation Solution
A radar system with microprocessors integrated into each radar monolithic chip to preprocess echo data, combined by a data combination module, and then post-processed by a central processing unit, sharing computing power and reducing bottlenecks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the antenna array scale of the radar sensor is increased to improve measurement capability, then the capability of measuring environment and distinguishing target angles is enhanced, but the computing power requirement and data throughput requirement increase significantly
Solution Approach 1:
The patent divides the radar system into multiple independent radar monolithic chips, each capable of autonomous preprocessing. This segmentation distributes the computing load across multiple chips, allowing each chip to handle a portion of the data processing independently before results are combined, thereby reducing the computing power burden on any single processor
Solution Approach 2:
The patent implements preliminary action by performing echo data preprocessing on each radar monolithic chip before data combination. The preprocessing operations (such as FFT, range cell migration correction) are executed in advance at the chip level, reducing the complexity and computing power required for subsequent central processing unit operations
2Adaptability or versatility
If data from multiple radar sensors is processed uniformly by a central processing unit, then data convergence is achieved, but the computing power of the central processing unit is overwhelmed
Solution Approach 1:
The patent segments the data processing function between radar monolithic chips and the central processing unit. Each chip performs autonomous preprocessing and outputs intermediate results, which are then combined and subjected to further processing by the central unit. This segmentation prevents the central processing unit from being overwhelmed by raw data from multiple sensors
Solution Approach 2:
The patent introduces an intermediary data combination module that collects preprocessed data from multiple radar monolithic chips and performs data combination before passing results to the central processing unit. This intermediary layer manages data flow and reduces the processing burden on the central unit
3Productivity
If data is processed by a processor at one end of the radar sensor before being sent to the central processing unit, then data throughput is improved, but the processor quickly enters a bottleneck in processing capability, cache capability, and data throughput capability
Solution Approach 1:
The patent embeds processing functionality directly within each radar monolithic chip, creating multiple distributed processing units rather than relying on a single end processor. This segmentation distributes the processing load and avoids the bottleneck that would occur in a centralized end processor
Solution Approach 2:
Each radar monolithic chip performs autonomous preprocessing of its own echo data without requiring external processing assistance. The chips are self-sufficient in performing critical processing operations, reducing the burden on external processors and avoiding bottleneck conditions
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
This application discloses a radar system, which may be applied to an in-vehicle radar in automated driving, such as an in-vehicle millimeter-wave radar. The radar system includes at least one radar sensor, a first radar sensor in the at least one radar sensor includes a data combination module (402) and a plurality of radar monolithic chips (401), and each radar monolithic chip (401) includes a first radio frequency front end (4011) and a first microprocessor (4012). The first microprocessor (4012) is configured to preprocess echo data obtained by the first radio frequency front end (4011), and after the data combination module (402) combines and transmits the preprocessed echo data, a next-level processor performs post-processing on the preprocessed echo data to generate point cloud data of the radar system. The first microprocessor (4012) preprocesses the echo data, so that computing power burden of a subsequent processor is reduced, a cache capability and a data throughput capability of the microprocessor are improved, and resource allocation is more proper, thereby improving adaptability of the radar system to a radar sensor with a large-scale antenna array.