[Translation] Texas Instruments Millimeter Wave Radar SAR Imaging Tutorial

This translation is based on the original document released on August 13, 2018, by the University of Texas at Dallas, titled “SAR IMAGING TUTORIAL” (Synthetic Aperture Radar Imaging Tutorial), authored by Muhammet Emin Yanik and Professor Murat Torlak. Thanks to those who laid the groundwork, the author only translates for learning purposes.

Table of Contents

1.Introduction

2.System Configuration and Recorded Data Format

2.1.System Configuration

2.2.Recorded Data Format

3.Simplified 2-D Imaging Algorithm

4.Recorded Data Scenarios

4.1.Planar 2-D Targets

4.2.Two Planar 2-D Targets Concealed in a Box

4.3.Cascaded Concealed Targets

5.Processing Software

6. 2-D Imaging Results

1.Introduction

Within the scope of this project, a two-dimensional (2-D) near-field imaging system was designed, combining synthetic aperture radar (SAR) processing techniques with low-cost millimeter-wave frequency-modulated continuous wave (FMCW) radar. To create a synthetic aperture over the target scene, a dual-axis automated track system was constructed and integrated with the Texas Instruments (TI) IWR1443 77 GHz millimeter-wave FMCW radar sensor.

This tutorial introduces simplified signal processing techniques for near-field 2-D image formation and details the specifications of recorded SAR data samples.

2.System Configuration and Recorded Data Format

2.1System Configuration

To reconstruct the 2-D image of the scene, data collection was performed by moving the radar along a trajectory in the x−y plane, creating a rectangular grid of measurement points, as shown in Figure 1. Assuming the transmitting and receiving antennas are very close to each other, they are represented by their midpoint. In the established (x, y, z) Cartesian coordinate system, the x-axis, y-axis, and z-axis represent the horizontal, vertical, and range directions, respectively.

System Configuration

As shown in the measurement configuration, the transceiver is located at position (x′, y′, 0) at a specific measurement time, and a general point on the target is located at position (x, y, z0) at a distance z0 from the imaging system. The detailed scanning aperture configuration is shown in Figure 2. The total scanning size is represented by Dx and Dy on the x-axis and y-axis, respectively, with corresponding sampling distances of dx and dy.

Aperture Configuration

The Chirp parameters used in the experiment are shown in Figure 3.

Chirp Parameters

2.2. Recorded Data Format

The 3-D recorded data cube is shown in Figure 4, and its properties are detailed below.

Data Recording

rawData3D is an nBeam × nVertical × nHorizontal 3-D data matrix of nSample.

Details of the parameters are as follows. Dimensions

nSample: Number of samples in the time domain

nVertical: Number of sampling points in the vertical (y) axis

nHorizontal: Number of sampling points in the horizontal (x) axis

Dx: Aperture size in the horizontal (x) axis

Dy: Aperture size in the vertical (y) axis

dx: Sampling distance in the horizontal (x) axis

dy: Sampling distance in the vertical (y) axis

z0: Distance to the target

3.Simplified 2-D Imaging Algorithm

After the target and aperture coordinates coincide, the 2-D reflectivity image can be represented as −1[FT2D[s(x, y)]FT2D[h(x, y)]], where FT2D and FT2D f(x, y) = FT2D −1 represent the two-dimensional Fourier and inverse Fourier transform operations on the −y plane, f(x, y) is the two-dimensional target reflectivity function, s(x, y) is the measured radar signal, and h(x, y) is the impulse response or point spread function of the imaging system, calculated for each (x, y) measurement point as h(x, y) = e−j2k√x2+y2+z02.

The reconstruction algorithm summarized in Figure 5 does not consider any visibility conditions. Therefore, the built-in MATLAB 2-D FFT function can be used directly.

4.Recorded Data Scenarios

4.1.Planar 2-D Targets

The scenario is shown in Figure 6, and its properties are detailed below.

Data Name rawData3D_simple2D

Parameters

nSample: 512

nVertical: 100

nHorizontal:407

Dx:200 mm (dx:200/406 mm)

Dy:198 mm (dy:2 mm)

z0:280 mm

4.2. Two Planar 2-D Targets Concealed in a Box

The scenario is shown in Figure 7, and its properties are detailed below.

Data NamerawData3D_twoConcealed2D

Parameters

nSample:512

nVertical:101

nHorizontal:407

dx:200 mm (dx:200/406 mm)

dy:200 mm (dy:2 mm)

z0:260 mm (front of the box), 320 mm (back of the box)

4.3. Cascaded Concealed Targets

The scenario is shown in Figure 8, and its properties are detailed below.

Data Name rawData3D_cascadedConcealed

Parameters

nSample:512

nVertical:101

nHorizontal:407

Dx:200 mm (dx:200/406 mm)

Dy:200 mm (dy:2 mm)

z0:250 mm (first target), 340 mm (second target)

5.Processing Software

The main script of the processing software is shown in Table 1.

Main Code:

This code block preprocesses rawData3D for 2-D imaging and calls other SAR imaging functions.

The following parameters should be updated based on the scenario.

%% Load rawData3D

dataName = ‘rawData3D_simple2D’; % Change only this line

rawData = load(dataName);

rawData = rawData.(dataName);

%% Define parameters, update based on the scenario

nFFTtime = 1024; % Number of FFT points for Range-FFT

z0 = 280e-3; % Range of target (range of corresponding image slice)

dx = 200/406; % Sampling distance at x (horizontal) axis in mm

dy = 2; % Sampling distance at y (vertical) axis in mm

nFFTspace = 1024; % Number of FFT points for Spatial-FFT

The following lines call the matched filter creation function.

%% Create Matched Filter

matchedFilter =

createMatchedFilterSimplified(nFFTspace,dx,nFFTspace,dy,z0*1e3);

The following lines call the image reconstruction function.imSize parameter.

%% Create SAR Image

imSize = 200; % Size of image area in mm

sarImage =

reconstructSARimageMatchedFilterSimplified(sarData,matchedFilter,dx,dy,imSize

);

Matched filter function

matchedFilter = createMatchedFilterSimplified(xPointM,xStepM,yPointM,yStepM,zTarget) This function creates a 2-D matched filter. Input parameters and outputs are detailed below:

SAR Imaging Function

sarImage = reconstructSARimageMatchedFilterSimplified(sarData,matchedFilter,xStepM,yStep M,xySizeT)

This function creates a 2-D SAR image. Input parameters and outputs are detailed below:

6.2Imaging Results

Figure 9 shows the 2-D imaging results of the planar 2-D target scene.

Figure 10 shows the 2-D imaging results of the two planar 2-D target scenes concealed in a box.

Two different slices were obtained using two different range boxes.

Imaging results of the two planar 2-D target scenes concealed in a box

(Note: These are preliminary results and will be adjusted after 3-D imaging)

Figure 11 shows the 2-D imaging results of the cascaded concealed target scene.

Two different slices were obtained using two different range boxes.

Imaging results of the cascaded concealed target scene (Note: These are initial results and will be adjusted after 3D imaging)