Radar Direction of Arrival Estimation via AR Extrapolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radar systems face challenges in improving angular resolution without increasing computational complexity or system size, and existing methods have limitations such as limited maximum unambiguous detection ranges and requirements for prior target knowledge, which are not suitable for automated driving assistance systems.
Innovation Solution
The radar system employs a frequency-modulated continuous wave (FMCW) radar system that extrapolates input radar signals using auto-regressive models and the Burg method to determine the angle of arrival, balancing improved angular resolution with reduced computational complexity, without increasing the system's size on an integrated circuit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the size of the array and number of receiver channels are increased to avoid grating lobes, then angular resolution is improved, but the size of the radar system on an integrated circuit increases
Solution Approach 1:
The patent changes the parameter of array configuration from a full dense array to a sparse array with specific geometric patterns. This allows maintaining angular resolution through mathematical processing while keeping the physical array size reduced, thus avoiding grating lobes without increasing integrated circuit size.
Solution Approach 2:
The patent introduces a signal processing intermediary (sparse array processing algorithm) that mediates between the reduced physical array size and the required angular resolution. This intermediary process enables accurate direction of arrival estimation without requiring a large physical array.
2Area of stationary object
If a sparse array is used to maintain radar system size, then integrated circuit size is maintained, but computationally expensive sparse processing algorithms are required
Solution Approach 1:
The patent applies preliminary action by pre-designing optimized sparse array geometries and pre-computing processing parameters. This preparation work is done offline, allowing the actual radar processing to use simplified algorithms that are less computationally expensive while still achieving accurate results.
3Measurement precision
If super-resolution algorithms are used to improve angular resolution, then angular resolution is improved, but computational complexity increases and maximum unambiguous detection range is limited
Solution Approach 1:
The patent uses a copying approach by creating a virtual extended array through signal processing that replicates the information content of a larger physical array. This virtual copying enables super-resolution capabilities without the computational burden of actual super-resolution algorithms, and without limiting the maximum unambiguous detection range.
4Measurement precision
If the number of receiver channels is increased to improve angular resolution, then angular resolution is improved, but the radar system size on an integrated circuit increases
Solution Approach 1:
The patent merges the functions of multiple receiver channels by using a sparse array configuration where fewer physical channels are strategically positioned. Through signal processing, the system combines the information from these fewer channels to achieve the angular resolution that would otherwise require more channels, thus reducing the number of receiver channels needed.
Data Source
AI summary
A system includes a processor and a non-transitory computer-readable medium storing machine instructions executable by the processor. The processor obtains an input radar signal, determines coefficients for an auto-regressive model based on the input signal and an order size, and extrapolates the input signal based on the model and the determined coefficients to obtain an extrapolated signal. In some implementations, the input signal comprises a number N of samples, and the order size is approximately half of N. The coefficients are determined based on the N samples, and the processor extrapolates a right-side signal based on the model, the determined coefficients, and a second half of the N samples; extrapolates a left-side signal based on the model, complex conjugates of the determined coefficients, and a first half of the N samples; and generates the extrapolated signal based on the left-side signal, the input signal, and the right-side signal.