ELKOMIKA: Jurnal Teknik Energi Elektrik, Teknik Telekomunikasi, & Teknik Elektronika

ISSN(p): 2338-8323 | ISSN(e): 2459-9638 | Vol. 12 | No. 4 | Halaman 968 - 982

DOI : http://dx.doi.org/10.26760/elkomika.v12i4.968 Oktober 2024

Polarimetric Calibration for Ground-based

Synthetic Aperture Radar based on Point Target
Approach
NURUL CHASANAH1, DADANG GUNAWAN1, FAROHAJI KURNIAWAN2, AGUS
WIYONO2, JEFRI ABNER2, BAMBANG SETIADI3
1Department of Electrical Engineering, Universitas Indonesia, Depok, Indonesia
2Research Center for Aeronautics Technology, BRIN, Bogor, Indonesia
3Research Center for Telecommunication, BRIN, Bogor, Indonesia

Email: nurul.chasanah31@ui.ac.id
*Corresponding author: guna@eng.ui.ac.id

Received 1 September 2024 | Revised 30 September 2024 | Accepted 30 Oktober 2024

ABSTRAK

Kalibrasi polarimetri (PolCal) dari citra radar aperture sintetis berbasis darat (GB
SAR) sangat penting dalam aplikasi analisis spasial empiris. Studi ini menyajikan
prosedur pra-pemrosesan dan metode kalibrasi untuk memperoleh citra kompleks
polarisasi tunggal dengan kualitas tinggi yang terkalibrasi. Pemrosesan data GB
SAR menjadi citra kompleks tunggal diterapkan dalam riset ini. Penelitian ini
bertujuan untuk menerapkan kalibrasi polarimetri eksternal pada radar gelombang
kontinu frekuensi bertahap (SFCW) di pita frekuensi C yang dihasilkan oleh
perangkat VNA menggunakan satu trihedral dan dua dihedral CR dengan sudut
kemiringan yang berbeda. Setelah pra-pemrosesan, teknik kalibrasi yang
diterapkan menunjukkan ketidakseimbangan saluran antara saluran polarisasi
yang berbeda (co-pol and cross-pol) dapat diatasi dengan baik.

Kata kunci: Kalibrasi Polarimetri, Radar Apertur Sintetis Berbasis Darat (GB SAR),
Ketidakseimbangan Saluran

ABSTRACT

Polarimetric calibration (PolCal) of ground-based synthetic aperture radar (GB SAR)

images is critical in empirical spatial analysis applications. This study provides the
preprocessing procedures and calibration method required for obtaining high-
quality calibrated polarimetric single-look complex imagery. The technique of
converting GB SAR data into single-look complex images is examined. This work
aimed to perform an external polarimetric calibration of the C-band stepped-
frequency continuous wave (SFCW) radar produced by VNA employing single
trihedral and two dihedral CRs with varying tilt angles. After preprocessing, the
calibration technique profoundly enhances channel imbalance between different
polarization channels (co-pol and cross-pol).

Keywords: Polarimetric Calibration, Ground-based synthetic aperture radar (GB

SAR), Channel Imbalance

ELKOMIKA – 968
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

1. INTRODUCTION

Recently, the development of SAR technology has evolved immensely, leading to its
widespread application in different areas such as terrain mapping, astronomy, and military
surveillance (Feng et al., 2022) (Peters et al., 2021) (Song et al., 2022). Due to its
principle of using microwave sensors, SAR offers day and night, all-weather conditions, and
high-resolution images (Richards et al., 2013). In general, synthetic aperture radar (SAR)
is an imaging radar that detects wideband echoes at multiple locations while the platform
moves relative to the target. After an accumulating period, SAR produces a two-dimensional
(2D) image of the target by processing received echoes coherently, enabling high-resolution
scrutiny of the actual image (Lu, 2019).

Polarization, the key characteristic of electromagnetic waves, may emit various scattering
properties when interacting with ground targets. By analyzing phase and amplitude information
from four-channel data using polarimetric decomposition and classification techniques, SAR
data can reveal scattering mechanisms on the ground surface caused by distinct objects
(Kirkpatrick et al., 2003) (Zhang et al., 2024). However, polarimetric distortions in SAR
data, known as distortion matrices (DMs), may contribute to incorrect inputs and misperception
of scattering and ground targets owing to non-ideal system polarization quality and
propagation variables (Jung & Park, 2018). The major polarimetric distortions relate to
channel imbalance, crosstalk, and Faraday rotation errors (Babu et al., 2022). The
misalignment in the transmit-receive modules (TRMs) and other antenna elements contributes
to the vertical component being more vulnerable to horizontal polarization (and vice versa),
leading to crosstalk faults. These mistakes contribute to the rising magnitude of cross-pol
images, resulting in distortion of volume scattering parameters (Chang et al., 2018). Second,
the disparity in gain of power amplifiers (PA) between horizontal and vertical polarizations of
the same and different TRMs, the distinction in sidelobe lowering capacity of Low Noise
Amplifiers (LNA), and the variations in attenuator ways between horizontal and vertical
polarizations can lead to channel imbalance and phase bias. Also, a Faraday rotation mistake
emerges when a linearly polarized electromagnetic wave propagates through the Earth's
ionosphere, solely affecting spaceborne SAR systems (Kumar et al., 2022). Therefore, the
benefits of polarimetric properties can only be fully realized by rigorous polarimetric calibration,
a vital preprocessing procedure that yields exact polarimetric properties.

Mainly, polarimetric calibration schemes are typically classified into three different kinds: corner
reflector (CR) calibration (Freeman et al., 1990) (Liang, 2020), distributed targets (DT)
calibration (Chang et al., 2023) (Han et al., 2023) (Zhou et al., 2022), and hybrid
calibration employing both CR and DT (Hou et al., 2022) (Quegan, 1994) (Tan & Hong,
2016). The methods that use distributed targets usually require certain assumptions and
accurate measurements. Thus, these assumptions will result in a lower calibration accuracy.
In the first method, polarimetric distortions are evaluated using artificial calibrators, which
might be a polarimetric active radar calibrator (PARC), which is commonly utilized in
spaceborne and airborne SAR systems (Shi et al., 2020) or passive corner reflectors used as
point targets. Besides, the point target method is precise, objective, and has the advantages
of lower cost and simple maintenance (Yu et al., 2022).

Utilizing the SAR idea of sensor antenna motion, Ground-based synthetic aperture radar (GB
SAR) is a type of terrestrial remote sensing that practically expands the sensor antenna
aperture by collecting radar signals from various points along its trajectory (Chasanah et al.,
2023). Contrary to satellite or aerial SAR systems, it is predominantly used for operations
requiring precise and steady observations, like surface deformation, monitoring of dams and
landslides, and stability evaluation (Qi et al., 2024). Moreover, Ground-based SAR (GB SAR)
sensors are appropriate for replacing orbital-based solutions since this technique is suitable for
ELKOMIKA – 969
 Chasanah et al.

monitoring smaller areas, such as city buildings or mountain hillsides. The sensor platform has
advantages such as robustness, swift reassessing time, ideal spatial resolution, and flexibility
in adjusting the illumination angle to the peculiar site geometry (Iglesias et al., 2014).
Provided that the Research Centre for Aeronautics, Indonesia, is developing a GB SAR system
to acquire data using a full polarimetric strip map. Therefore, polarimetric calibration needs to
be applied to get minimum distortion images.

Table 1 summarizes research on polarimetric calibration for GB SAR. The majority of earlier
research used frequency-modulated continuous wave (FMCW) radar. Polarimetric calibration
for GB SAR has been developed using a hybrid corner reflector using three types of calibration
targets, i.e., trihedral, dihedral, and 22.5°-rotated dihedral corner reflectors without an
assumption in FMCW signal that performs at Ku-band (Wang et al., 2021). In (Izumi et
al., 2017), the evaluation of the two calibration procedures by Wiesbeck et al. (Wiesbeck &
Riegger, 1991) , and Gau et al. (Gau & Burnside, 1995) was determined to be
implemented in circularly polarized GB SAR at C-band using stepped-frequency continuous
wave (SFCW) signal. In brief, a study calibration for linearly-polarized (LP) GB SAR in SFCW
mode is necessary for improved use in GB SAR. The contribution of this article is that the
distortion matrices are calculated concerning the calibration set in the experiment. The
calibrator set consists of three calibration targets, i.e., trihedral, dihedral, and 45° rotated
dihedral corner reflector in the C band using SFCW signal mode. Then, the polarimetric
calibration method (Wang et al., 2021) is adapted, making calibrated polarimetric GB SAR
in SFCW mode that is simple and quick yet requires no assumptions.

Table 1. Polarimetric Calibration Technique for GB SAR that are described in literature

Ref Year Radar Type Pol Method

(Krasnov et al., 2023) 2023 FMCW Linear Rotating dihedral
(Wang et al., 2021) 2021 FMCW Linear GCT
(Baffelli et al., 2018) 2018 FMCW Linear STCT
(Sun et al., 2017) 2017 FMCW N/A GCT
(Izumi et al., 2017) 2017 SFCW Circular GCT
Proposed Research SFCW Linear GCT

The subsequent contents are organized as follows. Section 2 provides a concise overview of
the SFCW radar employed in this investigation, along with the calibration technique. Section 3
presents experimental studies conducted to validate the calibration technique. Section 4
proceeds with the analysis and discourse of the gathered data. In conclusion, Section 5
presents the findings.

2. METHOD

2.1 SFCW Radar Signal

A fundamental necessity to produce calibrated polarimetric data is the availability of correctly
processed Single Look Complex (SLC) images for all elements of the polarimetric scattering
matrix. Therefore, it is vital to understand how data is received from the GB SAR system. The
signal in the present research is generated using a vector network analyzer (VNA) as a
measurement device. It operates in the frequency domain and employs the stepped-frequency
continuous-wave (SFCW) radar principle to obtain the scattering parameter (S21).

ELKOMIKA – 970
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

Figure 1. Transmitted signals of SFCW radar

VNA operating in SFCW is depicted in Figure 1. Its sweep time, Tdwell, can be defined as the
span within the transmission frequency. For (N) frequency points, the frequency rises in f
increments. The observation time (To), or coherent processing time, is the entire duration of
all sequential sweeps, whereas the pulse repetition interval Tp is the period of every single
step. Sweeping period (Tdwell), bandwidth (B), target range (R), and wave speed (c) define the
correct sampling frequency. The sampling frequency has to meet or exceed the maximum
frequency of baseband signals (Acar et al., 2021). The maximum frequency of return signals
(Fmax), is specified as

𝐹 = (1)

where the maximum range (Rmax) should be equal to or less than the unambiguous range,
which is

𝑅 ≤𝑅 = (2)
∆

Another crucial aspect of an SFCW signal is range resolution. The range resolution is referred
to as the smallest discernible difference between the radial distances among several targets.
The range resolution (R) of an SFCW radar can be measured by

∆𝑅 = (3)

The transmitting signal can be described as

𝑆 (𝑡) = 𝐴 𝑐𝑜𝑠 [2𝜋(𝑓 + 𝑖∆𝑓)𝑡 + 𝜃 ] (4)

𝑖 = 0, 1, … , 𝑁 − 1

In Eq.4, 𝐴 represents the amplitude, and i represents the phase at step i. Then, receiving
signals are delayed types of transmitting signals. The time delay (t) is defined as

( )
𝜏(𝑡) = (5)
/

Where (c) is the wave velocity, (R) is the range, and x(t) is the object displacement. The
objects' range and displacement are retrieved from baseband return signals of subsequent
sweeps. Afterward, the back-projection algorithm (BPA) will readily apply to the received raw
data SAR for all polarizations. The procedure consists of two fundamental stages: the inverse
FFT is fitted to every row of the complex data matrix to create the synthetic range matrix.
ELKOMIKA – 971
 Chasanah et al.

Next, interpolation is implemented to optimize object detection and display the SAR image
(Chasanah et al., 2023).

2.2 Polarimetric GB SAR Distortion Model

This research deploys two LP horn antennas in a quasi-monostatic set-up, collecting a single
polarization each time utilizing transmitter and receiving antenna sets. Hence, crosstalk
calibration is unnecessary because the radar has desirable polarization isolation. The only
potential contributor to crosstalk is when the main lobe and cross-polarized antenna sidelobes
are pointed in the same direction.

The backscattering characteristics of a ground scatterer can be properly represented by the

scattering matrix constructed in the horizontal (H) and vertical (V) polarization patterns as

𝑆 𝑆
[𝑆] = (6)
𝑆 𝑆

While first letter of the subscript denotes the scattered or received polarization; the second
letter denotes the incident or transmit polarization.

Polarimetric erroneous may affect the scattering coefficients observed by the polarimetric
synthetic aperture radar (PolSAR) system. These faults are divided into three distinct kinds:
crosstalk, channel imbalance in cross and co-polarization, and Faraday rotation. The Faraday
rotation error can be omitted for C-band Pol-SAR (Wright et al., 2003), whereas crosstalk
is excluded since we simply get a single polarization at each measurement. The correlation of
the measured scattering matrix [Sm] and the right target scattering matrix [Sc] is indicated as

[𝑆 ] = [𝐼] + [𝑅][𝑆 ][𝑇] (7)

Equation (7) can be expanded as

𝑆 𝑆 𝐼 𝐼 𝑅 𝑅 𝑆 𝑆 𝑇 𝑇
= + (8)
𝑆 𝑆 𝐼 𝐼 𝑅 𝑅 𝑆 𝑆 𝑇 𝑇

Where [R] is the received errors. [I] is the additive matrix, and [T] is the transmit errors.
During polarization calibration, the effects of additive white noise, [I], are usually ignored.
Therefore, the equation can be simplified while matrix [R] and [T] can be multiplied into
vector form, as follows:

𝑆 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑆
⎡ ⎤ ⎡ ⎤
𝑆 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑆
⎢ ⎥= ⎢ ⎥ (9)
⎢𝑆 ⎥ 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 ⎢𝑆 ⎥
⎣𝑆 ⎦ 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 𝑅 𝑇 ⎣𝑆 ⎦

By combining each of the prospective elements of Rij and Tij into one distortion matrix (Cij)
we can generate the calculation stated as

[𝑆 ] = [𝐶] [𝑆 ]
𝑆 𝐶 𝐶 𝐶 𝐶 𝑆
⎡ ⎤ ⎡ ⎤
𝑆 𝐶 𝐶 𝐶 𝐶 𝑆
⎢ ⎥= ⎢ ⎥ (10)
⎢𝑆 ⎥ 𝐶 𝐶 𝐶 𝐶 ⎢𝑆 ⎥
⎣𝑆 ⎦ 𝐶 𝐶 𝐶 𝐶 ⎣𝑆 ⎦

ELKOMIKA – 972
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

Thus, by sequential substitution of a homogenous system, the sixteen error coefficients of [C]
can be minimized into eight as follows:

𝑆 ⎡ 𝑐 ⎤ 𝑆
⎡ ⎤ ⎢ ⎥ ⎡ ⎤
𝑆 𝑐 𝑆
⎢ ⎥=⎢ ⎥ ⎢ ⎥ (11)
⎢𝑆 ⎥ ⎢ 𝑐 𝑐 𝑐 ⎥ ⎢𝑆 ⎥
⎣𝑆 ⎦ ⎢ 𝑐 𝑐 ⎥ ⎣𝑆 ⎦
⎣ 𝑐 ⎦

In principle, those eight unknowns can be ascertained by measuring two or more calibration
targets that yield four independently variable backscattering coefficients. The unknown values
c11, c22, c31, c32, c41, and c42 correspond to the co-pol signals, whereas c33 and c44 correspond
to the cross-pol signals. For the calibration process, dihedral and triangular trihedral corner
reflectors are widely used as calibration point targets due to their known theoretical scattering
matrices. Table 2 shows specific information on these calibration corner reflectors.

Table 2. Calibration Target Scattering Matrices Characteristics

Corner Theoretical Scattering
Reflectors Matrices

Vertical 1 0
Dihedral 0 −1

0 1
Dihedral 45
1 0

Triangular 1 0
Trihedral 0 1

The present study utilized three calibration targets with known scattering matrices [Sci] where
i = 1, 2, and 3 indicate a vertical dihedral, dihedral 45, and triangular trihedral. As a result,
the error coefficients are calculated by solving the equations as follows:

𝑐 = (12) 𝑐 = (15)

𝑐 = (13) 𝑐 = (16)

𝑐 = (14) 𝑐 = (17)

ELKOMIKA – 973
 Chasanah et al.

𝑐 = ± −𝑐 𝑐 (18)

𝑐 = ± −𝑐 𝑐 (19)

Finally, the calibrated scattering matrix can be readily acquired after finding the [C] matrix as
follows

[𝑆 ] = [𝐶] [𝑆 ] (20)

A flowchart (Fig 2) is presented to make the methods used in this study clear. Since there is
no assumption in the calculation, this method can accurately solve imbalance calibration.

Measured GB SAR Data

[𝑆 ] Trihedral [𝑆 ] Vertical [𝑆 ] 45

CR Dihedral CR Dihedral CR

Theoretical Reference target

Calculate distortion matrices
[𝑆 ]
Trihedral CR

Calibrate Channel Imbalance (CI)

[𝑆 ] Vertical
Dihedral CR
No
CI <0.5 dB? [𝑆 ] 45
Dihedral CR
Yes

Finish Calibration

Figure 2. Flowchart of polarimetric calibration procedure from GB SAR data

The technique for calibrating the images from SAR is laid out below.
Step 1: Measure all three reference targets.
Step 2: Compute distortion matrices by adding the theoretical values of the reference targets
in (11).
Step 3: Calculate the channel imbalance based on distortion matrices
Step 4: Verify that the polarimetric calibration is applied accurately if the CI is less than 0.5
dB; otherwise, return to Step 2.

2.3 Field Experiment

This section delineates the polarimetric radar measurement configuration employing the quasi-
monostatic methodology. A VNA-developed SFCW GB SAR system comprises 1601 sample
points and functions within the C-band frequency range of 5-7 GHz. The antenna configuration
included a pair of LP horn antennas with a synthetic aperture length of 9 meters and a 10 cm
ELKOMIKA – 974
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

aperture step in the cross-range. A longer rail scanning distance may lead to better azimuth
resolution. However, a lower stepping step is required to avoid multiples caused by azimuth
ambiguity. The visual design of the experiment is illustrated in Fig. 3, where one trihedral and
two dihedral CRs oriented at 0° and 45° alternatingly were deployed in the open area.
Absorbing substances were installed to minimize unwanted echoes from the ground and target
position. The target's side dimensions were 30 cm, and the radius between the target and its
midsection synthetic aperture antenna was set to 10 meters, allowing for far-field testing. A
Laptop can set all the settings, including frequency, output power, and frequency sampling
points. Besides, it can control the data acquisition process during the measurements.

Figure 3. Schematic representation of the GB SAR experiment

3. RESULTS AND DISCUSSION

3.1. SAR Images Processing Results

The SAR images were generated using the back-projection algorithm (BPA). This approach
required filtering the range-compressed data around the reflector and converting it back to the
time domain using an inverse Fourier transform. Then, the back-projected images were
validated by interpolation of the complex data. BPA was preferred for it is the most reliable
technique (predicting the matched filter ideal solution) and enables random observation areas.
Hence, the retrieved SAR images of the targets for every polarization are shown in Fig. 4-6.
The three calibration targets can then be identified in the co-polarimetric images. Nonetheless,
all polarimetric signatures demonstrate the 45°-oriented dihedral CR, indicating that the
calibration targets were appropriately positioned. The target location can be confirmed from
these images as about 10 meters in range direction. Fig. 4 presents the reconstructed images
of the trihedral corner reflector in the amplitude domain. Notably, the artificial target
constructed of the trihedral is apparent in each case in co-pol signatures (|𝑆VV| and (|SHH|),
with the amplitudes of (|𝑆VV|) being slightly higher than those of (|SHH|). Meanwhile, the
trihedral target is barely visible in cross-pol signatures (|𝑆VH| and (|SHV|), and a minor amount
of object noise is visible close to the trihedral target.

ELKOMIKA – 975
 Chasanah et al.

(a) VV (b) HH (c) VH (d) HV

Figure 4. Triangular Trihedral SAR Images
Figure 5 conveys the SAR images in four data channels for vertical dihedral targets. As
demonstrated in Fig.5 (c) and (d), the illumination amplitude is roughly -15 dB for cross-pol
signatures, while a co-pol signature of about 0 dB demonstrates noteworthy scattering. In
principle, assuming a dihedral structure is placed orthogonal to radar illumination, we get
notable |SHH| and |𝑆VV| components. Note that the dihedral has an intense co-pol signature
and behaves as a double-bounce scatterer when the angle tilts vertically relative to the radar
illumination. Meanwhile, the cross-polarized component (|𝑆VH| and (|SHV|) should not deliver
any backscattering signal due to a dihedral placed at 0° only reflecting signal in the identical
direction. However, Fig. 5 (c) and (d) illustrate a backscattering signal left in cross-pol images.
Therefore, to correct these images, a distortion matrix should be constructed and utilized to
calibrate polarimetric GB SAR.

(a) VV (b) HH (c) VH (d) HV

Figure 5. Vertical Dihedral 0 SAR Images
Forests, trees, vegetation, and sloped/oriented surfaces are the primary winding sources of
cross-pol SAR images. On top of the 45°-oriented dihedral, there is no straightforward physical
representation. Indeed, the dihedral CR is an ideal target for cross-polarization calibration as,
while orientated at 45°, the polarization direction shifts from horizontal to vertical direction
and vice versa. As a result of polarization alterations, there is no echo of 45°-oriented dihedrals
in the HH and VV images. Hence, the cross-polarized (|𝑆VH| and (|SHV|) images highlight the
45°-oriented dihedral. Meanwhile, as shown in Fig. 6 (a) and (b), some backscattering signals
in co-pol signatures remain noticeable, with (|𝑆VV|) amplitudes being slightly higher than those
of (|𝑆HH|). The measured |𝑆VH| and (|SHV| images manifest in the first place in magnitude that
show 0 dB for its highest value, with (|𝑆VH|) amplitudes being slightly greater than those of
(|𝑆HV|).

ELKOMIKA – 976
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

(a) VV (b) HH (c) VH (d) HV

Figure 6. 45 Oriented Dihedral SAR Images

3.2. Polarimetric Calibration

The final polarimetric calibration was carried out using the approach described in Section 2.2,
which involved three reflectors in the scene utilized for examining calibration performance.
Firstly, the scattering matrices of the target site (the pixel with the highest value) are recovered
and normalized to the highest polarization channel value. Table 3 shows the acquired data,
reflecting the uncalibrated and calibrated scattering matrices of the three CRs. As calibration
targets, these three CRs include a trihedral, a dihedral, and a 45°-oriented dihedral.

Table 3. Scattering Matrix of Three targets before and after calibration

CR Uncalibrated Theoretical Value Calibrated Amplitude error (dB)

Tri-SVV 1102.25 10 0.9345.454 0.069
Tri-SHH 0.235110.06 10 1.04248.182 0.042
Tri-SVH 0.007945.92 00 0.051-77.92 0.052
Tri-SHV 0.008179.01 00 0.052 102.01 0.052

Di0-SVV 1-7.878 1180 1.269171.819 0.269

Di0-SHH 0.487749.181 10 1.1828 37.905 0.183
Di0-SVH 0.067127.22 00 0.0308243.0391 0.031
Di0-SHV 0.0703145.906 00 0.0294 132.289 0.029

Di45-SVV 0.259115.05 00 0.227496.19 0.227

Di45-SHH 0.168154.02 00 0.0509 91.187 0.051
Di45-SVH 1113.24 10 0.98630.48 0.0137
Di45-SHV 0.42690.373 10 1.043 1.65 0.043

Table 3 reveals that the amplitude error before polarimetric calibration in |SHH| is 0.765 dB for
trihedral and 0.512 dB for vertical dihedral, whereas the amplitude error in |SHV| for 45°-
oriented dihedral CR is 0.574 dB. To obtain high-quality polarimetric SAR images, the amplitude
error should be less than 0.5 dB. Meanwhile, except for these three polarimetric modes, the
amplitude error exceeds the requirement. After polarimetric calibration, the amplitude error
shows good performance for all these three conditions that |SHH| is reduced to 0.042 dB for
trihedral and 0.183 dB for vertical dihedral, while the amplitude error in |SHV| for 45°-oriented
dihedral CR is 0.043 dB.

After polarimetric calibration, the channel imbalance value improved to 0.08 dB for the vertical
dihedral target, 0.176 dB for the 45°-oriented dihedral target, and 0.11 dB for the trihedral
target. The experiment demonstrates that the technique performs satisfactorily regarding
channel imbalance accuracy. Nonetheless, the residual distortion of imbalances persists under
ELKOMIKA – 977
 Chasanah et al.

certain circumstances. As considering channel balancing, we aim to preserve the orthogonality

of the signals in the copol and cross-pol channels, meaning that the amplitude, phase, and
latency differences must be regulated to acquire accurate measurements of the target
characteristics. This systemic issue needs to be addressed at the hardware level. However, if
the impacts of residual channel imbalance are alleviated afterward, we must address it
mathematically by utilizing data-driven approaches. In contrast, the measured phases before
and after calibration are near the predicted ones for a 45°-oriented dihedral in cross-pol (0.48°
phase error in VH and 1.65° phase error in HV). This result creates a channel phase imbalance
of 1.17°. Additional phase values for different targets and polarizations may not be the same
theoretically. The occurrence was mostly caused by the impact of the tilt angle GB SAR system,
which varied between experiments. It should be noted that the signal phase refers to the
number of oscillation cycles of a waveform that go between the radar and surface before
returning. As a result, interaction with the ground surface may impact data collection.

Finally, the above analysis shows that polarimetric calibration can be carried out successfully
through quasi-monostatic SFCW waveform mode in C band signal, resulting in <0.2 dB channel
imbalance in amplitude and meeting calibration requirements. Meanwhile, the residual phase
channel imbalance persists and will require phase correction in future research. Furthermore,
this study can be enhanced by using polarimetric decomposition to examine the effect of
calibration on the obtained images.

4. CONCLUSION

The primary aim of this research is to calibrate polarimetric GB SAR in SFCW mode, which is
efficient and straightforward yet eliminates any assumptions. The study focuses on an
algorithm incorporating a single trihedral and two dihedral CRs at distinct rotation angles.
Calibration accuracy has been shown by examining the scattering matrices of the selected CRs
target before and after calibration, revealing significant conformity between calibrated and
theoretical polarimetric responses is readily achievable by assuming zero crosstalk. Regarding
channel imbalance, the findings suggest that it can be effectively removed with great precision
by utilizing the proposed method. The amplitude error tends to be minimized to 0.2 dB.
However, the phase error needs to be resolved in future research.

ACKNOWLEDGEMENTS

The authors would like to thank all Synthetic Aperture Radar Research Group members for
their help with data preparation and measurement. Also, this research was funded by the
Rumah Program Research Centre for Aeronautics Technology, BRIN and the Lembaga
Pengelola Dana Pendidikan (LPDP) as part of the Riset dan Inovasi untuk Indonesia Maju
(RIIM) program.

REFERENCES

Acar, Y. E., Saritas, I., & Yaldiz, E. (2021). An S-band zero-IF SFCW through-the-wall radar for
range, respiration rate, and DOA estimation. Measurement, 186, 110221.
https://doi.org/10.1016/j.measurement.2021.110221

ELKOMIKA – 978
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

Babu, A., Kumar, S., & Agrawal, S. (2022). Polarimetric Calibration and Spatio-Temporal
Polarimetric Distortion Analysis of UAVSAR PolSAR Data. Earth and Space Science, 9(4),
e2020EA001629.
Baffelli, S., Frey, O., Werner, C., & Hajnsek, I. (2018). Polarimetric Calibration of the Ku-Band
Advanced Polarimetric Radar Interferometer. IEEE Transactions on Geoscience and
Remote Sensing, 56(4), 2295–2311. https://doi.org/10.1109/TGRS.2017.2778049
Chang, Y., Li, P., Yang, J., Zhao, J., Zhao, L., & Shi, L. (2018). Polarimetric Calibration and
Quality Assessment of the GF-3 Satellite Images. Sensors, 18(2), 403.
https://doi.org/10.3390/s18020403
Chang, Y., Zhao, L., Liu, B., He, H., Fan, J., & Li, P. (2023). Polarimetric Calibration of Airborne
SAR Images Using Low Helix Scattering Distributed Targets. IGARSS 2023 - 2023 IEEE
International Geoscience and Remote Sensing Symposium, 4529–4531.
https://doi.org/10.1109/IGARSS52108.2023.10282499
Chasanah, N., Abner, J., Wiyono, A., Rahayu, N., Kurniawan, F., Widada, W., Setiadi, B., &
Gunawan, D. (2023). Experimental Evaluation of Polarimetric Ground Based Synthetic
Aperture Radar Imaging using Portable VNA. 2023 International Conference on Radar,
Antenna, Microwave, Electronics, and Telecommunications (ICRAMET) , 6–11.
https://doi.org/10.1109/ICRAMET60171.2023.10366606
Feng, H., Zhang, L., Dong, J., Li, S., Zhao, Q., Luo, J., & Liao, M. (2022). Mapping the 2021
October Flood Event in the Subsiding Taiyuan Basin by Multitemporal SAR Data. IEEE
Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 15,
7515–7524. https://doi.org/10.1109/JSTARS.2022.3204277
Freeman, A., Shen, Y., & Werner, C. L. (1990). Polarimetric SAR calibration experiment using
active radar calibrators. IEEE Transactions on Geoscience and Remote Sensing, 28(2),
224–240. https://doi.org/10.1109/36.46702
Gau, J.-R. J., & Burnside, W. D. (1995). New polarimetric calibration technique using a single
calibration dihedral. IEE Proceedings - Microwaves, Antennas and Propagation, 142(1),
19-25(6).
Han, Y., Lu, P., Liu, X., Hou, W., Gao, Y., Yu, W., & Wang, R. (2023). On the Method of Circular
Polarimetric SAR Calibration Using Distributed Targets. IEEE Transactions on
Geoscience and Remote Sensing, 61, 1–16.
Hou, W., Zhao, F., Liu, X., Liu, D., Han, Y., Gao, Y., & Wang, R. (2022). Hybrid Compact
Polarimetric SAR Calibration Considering the Amplitude and Phase Coefficients
Inconsistency. Remote Sensing, 14(2).

ELKOMIKA – 979
 Chasanah et al.

Iglesias, R., Fabregas, X., Aguasca, A., Mallorqui, J. J., Lopez-Martinez, C., Gili, J. A., &
Corominas, J. (2014). Atmospheric Phase Screen Compensation in Ground-Based SAR
With a Multiple-Regression Model Over Mountainous Regions. IEEE Transactions on
Geoscience and Remote Sensing, 52(5), 2436–2449.
Izumi, Y., Demirci, S., Baharuddin, M. Z., Waqar, M. M., & Sri Sumantyo, J. T. (2017). The
Development And Comparison Of Two Polarimetric Calibration Techniques For Ground-
Based Circularly Polarized Radar System. Progress In Electromagnetics Research B, 73,
79–93. https://doi.org/10.2528/PIERB17010604
Jung, Y. T., & Park, S.-E. (2018). Comparative Analysis of Polarimetric SAR Calibration
Methods. Remote Sensing, 10(12). https://www.mdpi.com/2072-4292/10/12/2060
Kirkpatrick, C. J., Bittinger, F., Nozadze, K., & Wessler, I. (2003). Expression and function of
the non-neuronal cholinergic system in endothelial cells. Life Sciences, 72(18), 2111–
2116.
Krasnov, O. A., Zhang, Q., & Yarovoy, A. (2023). Polarimetric Calibration of an FMCW Doppler
Radar with Dual-Orthogonal Signals. 2023 IEEE Conference on Antenna Measurements
and Applications (CAMA), 312–317.
https://doi.org/10.1109/CAMA57522.2023.10352897
Kumar, S., Babu, A., Agrawal, S., Asopa, U., Shukla, S., & Maiti, A. (2022). Polarimetric
calibration of spaceborne and airborne multifrequency SAR data for scattering-based
characterization of manmade and natural features. Advances in Space Research, 69(4),
1684–1714. https://doi.org/10.1016/j.asr.2021.02.023
Liang, W. (2020). A Set of Point-Targets-Based Polarimetric Calibration Methods Based on
General Polarimetric System Model. IEEE Journal of Selected Topics in Applied Earth
Observations and Remote Sensing, 13, 2502–2519.
Lu, J. (2019). Design Technology of Synthetic Aperture Radar. Wiley.
https://books.google.co.id/books?id=6tOdDwAAQBAJ
Peters, S. T., Schroeder, D. M., Haynes, M. S., Castelletti, D., & Romero-Wolf, A. (2021).
Passive Synthetic Aperture Radar Imaging Using Radio-Astronomical Sources. IEEE
Transactions on Geoscience and Remote Sensing, 59(11), 9144–9159.
https://doi.org/10.1109/TGRS.2021.3050429
Qi, Y., Hui, J., Hou, T., Huang, P., Tan, W., & Xu, W. (2024). A Clustering Approach for
Atmospheric Phase Error Correction in Ground-Based SAR Using Spatial
Autocorrelation. Sensors, 24(13). https://doi.org/10.3390/s24134240

ELKOMIKA – 980
 Polarimetric Calibration for Ground-based Synthetic Aperture Radar based on Point Target Approach

Quegan, S. (1994). A unified algorithm for phase and cross-talk calibration of polarimetric data-
theory and observations. IEEE Transactions on Geoscience and Remote Sensing, 32(1),
89–99. https://doi.org/10.1109/36.285192
Richards, M. A., Scheer, J. A., & Holm, W. A. (2013). Principles of Modern Radar: Basic
Principles. https://api.semanticscholar.org/CorpusID:114114032
Shi, L., Li, P., Yang, J., Sun, H., Zhao, L., & Zhang, L. (2020). Polarimetric calibration for the
distributed Gaofen-3 product by an improved unitary zero helix framework. ISPRS
Journal of Photogrammetry and Remote Sensing, 160, 229–243.
https://doi.org/10.1016/j.isprsjprs.2019.12.003
Song, Y., Li, J., Gao, P., Li, L., Tian, T., & Tian, J. (2022). Two-Stage Cross-Modality Transfer
Learning Method for Military-Civilian SAR Ship Recognition. IEEE Geoscience and
Remote Sensing Letters, 19, 3162707. https://doi.org/10.1109/LGRS.2022.3162707
Sun, G., Huang, L., Chen, K., & Han, C. (2017). An Efficient Polarimetric SAR Calibration
Algorithm Using Corner Reflectors. Canadian Journal of Remote Sensing, 43(3), 286–
296. https://doi.org/10.1080/07038992.2017.1330142
Tan, H., & Hong, J. (2016). Calibration of Compact Polarimetric SAR Images Using Distributed
Targets and One Corner Reflector. IEEE Transactions on Geoscience and Remote
Sensing, 54(8), 4433–4444.
Wang, S., Chen, K.-S., & Sato, M. (2021). Performance of SAR Polarimetric Calibration Using
Hybrid Corner Reflectors: Numerical Simulations and Experimental Measurements. IEEE
Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 14, 440–
451.
Wiesbeck, W., & Riegger, S. (1991). A complete error model for free space polarimetric
measurements. IEEE Transactions on Antennas and Propagation, 39(8), 1105–1111.
https://doi.org/10.1109/8.97343
Wright, P. A., Quegan, S., Wheadon, N. S., & Hall, C. D. (2003). Faraday rotation effects on L-
band spaceborne SAR data. IEEE Transactions on Geoscience and Remote Sensing,
41(12), 2735–2744. https://doi.org/10.1109/TGRS.2003.815399
Yu, T., Li, M., Li, W., Mao, H., Wang, R., Hu, C., & Long, T. (2022). Polarimetric Calibration
Technique for a Fully Polarimetric Entomological Radar Based on Antenna Rotation.
Remote Sensing, 14(7), 1551. https://doi.org/10.3390/rs14071551
Zhang, S., Cui, L., Zhang, Y., Xia, T., Dong, Z., & An, W. (2024). Research on Input Schemes
for Polarimetric SAR Classification Using Deep Learning. Remote Sensing, 16(11).
https://doi.org/10.3390/rs16111826

ELKOMIKA – 981
 Chasanah et al.

Zhou, Y., Zhuang, L., Duan, J., Zhang, F., & Hong, W. (2022). Synthetic Aperture Radar
Radiometric Cross Calibration Based on Distributed Targets. IEEE Journal of Selected
Topics in Applied Earth Observations and Remote Sensing, 15, 9599–9612.

ELKOMIKA – 982