Compressive Millimeter-Wave Radar Imaging With Sparse MIMO Arrays
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Millimeter wave radars suffer from low angular resolution due to small physical apertures and conventional signal processing techniques, limiting their effectiveness in high-resolution imaging, especially in visually degraded environments.
Innovation Solution
A sparse MIMO array design combined with an implicit neural network architecture, utilizing a compressive large aperture radar imaging system that reduces the number of antennas and leverages implicit neural networks for high-accuracy radar imaging, overcoming the limitations of conventional methods by being data set agnostic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional signal processing techniques are used with small physical apertures, then the radar system is simple and low-cost, but the angular resolution is low
Solution Approach 1:
The patent segments the continuous aperture into discrete sparse antenna elements arranged in a MIMO configuration. By dividing the aperture into strategically positioned individual antennas rather than using a continuous dense array, the system achieves equivalent resolution with fewer elements, resolving the contradiction between measurement precision and device complexity
Solution Approach 2:
The patent transitions from a traditional 2D dense antenna array to a sparse MIMO configuration that utilizes both spatial positioning and temporal coding dimensions. By adding the time dimension through coded excitation and processing, the system achieves high angular resolution with reduced spatial complexity, effectively moving the problem into another dimension
2Device complexity
If the number of antennas is reduced to lower cost and power consumption, then the device complexity decreases, but the read-out bandwidth requirement increases
Solution Approach 1:
The patent applies preliminary action by performing compressive sampling and signal processing on the received radar echoes before full digitization and storage. By preprocessing the analog signals through compressive receivers and applying sparse signal reconstruction algorithms, the system reduces the bandwidth requirements while maintaining imaging quality with fewer antennas
Solution Approach 2:
The patent substitutes the traditional mechanical/electrical approach of increasing antenna count with a computational approach using compressive sensing algorithms and machine learning. This replacement of physical expansion (more antennas) with computational processing (compressive reconstruction) reduces hardware complexity while managing information loss through intelligent signal processing
3Ease of manufacture
If a sparse MIMO array is used to reduce antenna elements, then the manufacturing cost decreases, but the image reconstruction accuracy may be compromised
Solution Approach 1:
The patent introduces an intermediary computational layer between the sparse physical measurements and the final image reconstruction. By using compressive sensing algorithms and trained neural networks as intermediaries to process and reconstruct the sparse MIMO data, the system recovers high-quality images despite the reduced antenna count, maintaining accuracy while reducing manufacturing cost
Solution Approach 2:
The patent changes the processing parameters by transitioning from conventional Fourier-based reconstruction to iterative compressive sensing algorithms and deep learning models. These parameter changes in the reconstruction process enable accurate image formation from sparse measurements, resolving the contradiction between ease of manufacture and measurement precision
Data Source
AI summary
Methods and systems are disclosed. The method includes obtaining one or more time-domain beat signals. The method further includes generating an under-sampled measured radar data cube based on the one or more time-domain beat signals, and determining, using a computer processor and a machine learning model, an initial scene reflectivity distribution image based on the under-sampled measured radar data cube. The method further includes synthesizing, using the computer processor, the initial scene reflectivity distribution image to obtain a synthetized full radar data cube, and processing, using the computer processor, the synthetized full radar data cube to obtain an under-sampled radar data cube. The method further includes determining, using the computer processor and the machine learning model, an enhanced scene reflectivity distribution image based on a loss function measuring a mismatch of the under-sampled radar data cube and the under-sampled measured radar data cube.