Radar Imaging With Clock Ambiguities Using Convex Sparse Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Radar imaging systems with distributed antennas face challenges due to clock ambiguities, leading to ill-posed problems with a vast number of unknowns, making it difficult to recover accurate radar images from measurements affected by asynchronous clocks.
Innovation Solution
The approach transforms the radar image recovery problem into a convex sparse recovery problem by representing the time shift of radar measurements as a convolution with a one-sparse shift kernel, allowing for the use of sparse reconstruction techniques to reduce the solution space and solve for the radar image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If distributed antennas are deployed to achieve large physical aperture for high resolution imaging, then cross-range resolution is improved, but clock synchronization accuracy deteriorates due to geographical distribution
Solution Approach 1:
Instead of attempting to synchronize clocks across distributed antennas (the conventional approach), the patent inverts the problem by allowing clock ambiguities to remain and formulating the imaging problem in terms of these ambiguities. The radar image is reconstructed by solving an optimization problem that accounts for the unknown time shifts, effectively turning the synchronization challenge into a solvable inverse problem.
Solution Approach 2:
The patent changes the fundamental parameters of the imaging formulation by introducing time shift parameters to account for clock ambiguities. Rather than assuming synchronized clocks, the model explicitly includes time shift parameters for each antenna, transforming the problem into a parameter estimation problem that can be solved through optimization while maintaining imaging resolution.
2Measurement precision
If phase error estimation methods are applied to correct clock drift, then phase accuracy is improved, but computational complexity increases due to non-linearity and phase wrapping
Solution Approach 1:
The patent inverts the conventional phase error correction approach by not attempting to estimate and correct phase errors directly. Instead, it formulates the problem as estimating time shift parameters that, when applied, would correct the phase errors. This inversion transforms the non-linear phase error estimation into a more tractable time shift parameter estimation problem.
Solution Approach 2:
The patent introduces time shift parameters as intermediary variables that mediate between the observed phase errors and the desired image reconstruction. These time shift parameters serve as a bridge, allowing the system to account for clock ambiguities without directly solving the complex phase error estimation problem, thereby reducing computational complexity.
3Manufacturing precision
If conventional phase error correction methods are used, then imaging resolution is improved, but reliability decreases due to failure to capture true error nature
Solution Approach 1:
The patent changes the fundamental parameters of the error modeling by introducing time shift parameters that can capture the true nature of clock ambiguities. These parameters are more reliable than phase error models because they directly represent the physical time displacement caused by asynchronous clocks, leading to more accurate error correction and higher reliability in the imaging process.
Data Source
AI summary
A radar system for generating a radar image of a scene includes an input interface to accept radar measurements of a scene collected from a set of antennas with clock ambiguities, wherein the radar measurements are measurements of reflections of a radar pulse transmitted to the scene, a hardware processor configured to solve a convex sparse recovery problem to produce a radar image of the scene, wherein the convex sparse recovery problem matches a time shift of the radar measurements with a signal generated by propagation of the radar pulse through a radar propagation function of the scene, wherein the time shift of the radar measurements is represented as a convolution of the radar measurements with a shift kernel that is one-sparse in time, and an output interface configured to render the radar image.