Article citation information:
Ślesicki,
B., Ślesicka, A., Kawalec,
A. Improve the safety of air transport, especially in militarized
terrain, by use of side looking airborne radar and space time adaptive
processing. Scientific Journal of Silesian University of Technology. Series
Transport. 2024, 123,
335-346. ISSN: 0209-3324. DOI: https://doi.org/10.20858/sjsutst.2024.123.17.
Błażej ŚLESICKI[1], Anna ŚLESICKA[2], Adam KAWALEC[3]
IMPROVE THE SAFETY OF AIR TRANSPORT, ESPECIALLY IN MILITARIZED TERRAIN,
BY USE OF SIDE LOOKING AIRBORNE RADAR AND SPACE TIME ADAPTIVE PROCESSING
Summary. The paper
explores the potential to enhance aviation safety, particularly in militarized
regions, by outfitting aircraft with Side Looking Airborne Radar (SLAR) and employing space-time adaptive processing (STAP) algorithms. The research objective revolves around
implementing a model of side-looking airborne radar and the corresponding STAP algorithms. This technology enables the detection of
slow-moving targets amidst strong interference, encompassing both passive
(clutter) and active (jammer) elements. Slow-moving targets relative to the
aircraft's speed include tanks, combat vehicles, command vehicles, artillery,
and logistical assets of enemy forces. The theoretical framework of space-time
adaptive processing is presented, elucidating the sequential steps of the
classical Sample Matrix Inversion Space-Time Adaptive Processing (SMI STAP) algorithm. The paper
underscores the significance of characteristic parameters delineating a linear STAP processor. The proposed solution facilitates the
detection of enemy combat measures and enhances aviation safety. It outlines a
radar model installed beneath the aircraft's fuselage and elucidates algorithms
for space-time adaptive processing of radar signals. The simulations conducted
within the article were executed using the MATLAB
environment. The simulation results indeed suggest that the proposed solution
holds promise for deployment in equipping aircraft of one's own military and
those engaged in operations within conflict zones. This paper stands as one of
the few contributions in the literature addressing the augmentation of aircraft
safety through radar and space-time adaptive processing.
Keywords: radar,
safety, space-time adaptive processing, signal processing, airborne radar, air
transport
1.
INTRODUCTION
Today,
radars can be categorized based on their function, operational mode, range,
probing signal type, or signal processing methodology. Nevertheless, this
merely provides an overview of the wide range of radar systems customized for
diverse applications. Additionally, radars can be classified based on the
platform on which they are deployed.
Deploying
radar on an aircraft surface inevitably introduces several technical
challenges. Firstly, the high velocity of the aircraft leads to moving received
echoes, necessitating specialized algorithms for their mitigation. Secondly,
achieving effective target detection amidst significant clutter poses another
substantial challenge. Consequently, this study explores the implementation of
Side Looking Airborne Radar (SLAR) installed beneath
the aircraft fuselage, coupled with Space-Time Adaptive Processing (STAP) algorithms to counteract interfering signals.
STAP represents a contemporary signal
processing approach applied in radar systems. This technique is instrumental in
detecting ground-based moving targets via radar systems mounted on airborne
platforms. In the literature, radar interference is categorized into passive
interference (referred to as “clutter”) and active interference
(known as “jammer”). Passive interference encompasses echoes from
signals reflected off buildings, vegetation, and other terrain obstacles, while
active interference includes radio interference transmitters hindering the
operation of one's own radar systems [1].
Aircraft
equipped with active radar sensors and state-of-the-art signal processing
capabilities possess real-time control and surveillance capabilities over their
operational airspace. Enhanced operational awareness for pilots mitigates risks
and threats posed by enemy combat assets. Therefore, the proposed solution
holds potential for deployment in outfitting aircraft for military forces and
those engaged in operations within conflict zones.
2. A
LITERATURE REVIEW
The
history of the beginning of STAP technology dates
back to the early 1970s [2].
As a result of the work done by researchers, several monographs were published
summarizing the current state of knowledge on STAP
[3]. The publications of the aforementioned authors laid the foundation for
future STAP research [4]. Until the late 1990s, STAP algorithms were based
on statistical methods for estimating the disturbance covariance matrix. A very
popular method, which later often served as a reference for new methods, was
the Sample Matrix Inversion (SMI) method [3].
The
publication [5] marked a significant advancement in scientific inquiry
regarding the evolution of methods for estimating interference covariance
matrices. It introduced a novel form of non-statistical interference covariance
matrix estimation termed the Direct Data Domain method, commonly abbreviated as
D3, making its debut appearance in the literature.
Subsequently,
research on statistical methods for estimating clutter covariance matrices was
discontinued, which seems appropriate, given the drawbacks with which they were
characterized. The main ones include the need to access a huge amount of
training data contained in distance bins, which was not infrequently difficult
to fulfil. In the following years, efforts were made to introduce improvements
to D3 methods by eliminating the disadvantages that
occurred, which caused phenomena such as difficulties in detecting a target
against a background of inhomogeneous interference or confusing target
detection [6].
Another
approach to estimate the clutter covariance matrix was through the
Knowledge-Aided STAP (KA-STAP)
method [7]. The concept behind this technique was to derive the clutter
covariance matrix using information about the radar-scanned area or the present
targets [8]. It is important to highlight that during the early 21st century,
one of the primary research objectives of the US Defence
Advanced Research Projects Agency (DARPA) was the
development of the KA-STAP method [9]. Consequently,
it is noteworthy that, based on the existing literature, it can be inferred
that assessing the accuracy of prior knowledge about the scanned terrain
compared to the actual radar data received. This constituted the primary focus
of scientific research on KA-STAP algorithms.
In
recent years, a fresh avenue of research has emerged in the advancement of
non-statistical techniques for clutter covariance matrix estimation [10]. These
approaches rely on sparse recovery algorithms within STAP
[11]. Consequently, according to reports from the literature, there has been a
surge in research publications over the past few years focusing on the
refinement of methods for clutter covariance matrix estimation in STAP, particularly emphasizing the utilization of sparse
recovery algorithms [12].
Fig. 2. STAP signal
processing [3]
In order
to realize the essence of the issues discussed in the publication, it seems
very helpful to present a diagram of STAP processing,
on which the process of estimating the clutter covariance matrix can be clearly
located [3]. Hence, the above figure highlights the most important steps in the
processing of the radar signal in STAP technology.
3. RESEARCH
PROBLEM
The
problem of this research is the implementation of a model of side looking
airborne radar and the corresponding algorithms for space-time adaptive signal
processing.
Fig. 2. Geometry of radar [4]
The
paper adopts a model of the localization of the radar and the target according
to the diagram in Figure 2. The letter P was used to denote the target,
the distance of the radar to the target was denoted Rs,
the elevation angle between the radar and the target was denoted θ, and the
azimuth angle between the radar and the target was denoted φ. The flight
altitude of the aircraft is denoted H. The aircraft is moving in a
straight line at a constant speed Va. Located under the
fuselage, the radar has a uniform linear antenna (ULA) radiating the target P
at an angle α.
The radar emits electromagnetic waves at a wavelength λ, the radar
signal is emitted as a sequence of M coherent pulses with a pulse
repetition rate fr.
The pulses are emitted by N antennas spaced at a fixed distance d
from each other. Accordingly, the received echo forms a radar data cube. The
radar data cube consists of composite signal samples collected for M
pulses by N antenna array elements for distance intervals from 1 to K.
Every
STAP algorithm analyses the unprocessed data,
utilizing a particular snapshot of the radar cube at a predetermined distance
bin k [3]. Following this, assessments regarding the presence or absence
of the target within the specified distance bin are commonly conducted.
To accomplish this, a filter is devised, distinguished by its strong
amplification of the pertinent signal from the target, alongside substantial
suppression of all other signals (including interference from stationary
objects and jammers). STAP is engineered to eliminate
echoes originating from interference sources while preserving the signal
originating from the target of interest [1].
4. RESEARCH
METHOD
The
data contained for the k-th snapshot of the
raw data cube is a matrix [1]:
For
further processing, the matrix should be regrouped into a vector:
According
to the literature, the first step in STAP processing
is to determine the control vector
where
where Vr is the radial
velocity between the target and the radar. The temporal steering vector is
given as [1]:
where
Finally, the
steering vector takes the form [1]:
where
the symbol
Subsequently,
the following stage involves the calculation of the clutter covariance matrix.
To achieve this, the distance bin under examination, denoted as k, is partitioned into
of clutter. It is postulated that the clutter component within a specific
distance bin results from the amalgamation of signals emanating from each
clutter patch. Consequently, the covariance matrix encompassing noise, clutter,
and jammer for a particular snapshot of the radar data cube is delineated as
[1], respectively [13]:
where
By establishing both the steering vector and the
covariance matrix encompassing clutter, jammer, and noise, ascertain the weight
vector [14]:
where
As previously noted, the STAP
processor functions as a linear filter, with its primary objective being the
elimination of clutter, jammer, and noise to facilitate target detection. The
relationship governing these processes is detailed in [16]:
where
At this point it is worth quoting a very
important parameter describing the degree of clutter in the received echo
signal. This parameter is the clutter ridge slope β. In general,
the higher its value, the more difficult it is to detect the target. Its
relationship is given by the formula [17]:
This equation defines the position of the
clutter in the space-time plane. It is worth noting that the distance between
antennas d takes a constant value at the stage of radar antenna design.
Hence, when operating the radar at a specific pulse repetition frequency
Another crucial measure indicating the
clutter level in the received signal is the rank of the clutter covariance
matrix. The rank of a matrix signifies the highest count of linearly
independent vectors that form its rows (or columns). The rank of a matrix is
expounded upon by Brennan's equation in the referenced paper [3]. The authors
in question approximated the rank of the clutter covariance matrix as expressed
by [18]:
The parentheses ⌊ ⌋ denote
rounding to the nearest integer.
5. RESEARCH RESULTS
The
simulations conducted in this paper were executed using the MATLAB
platform. Initially, simulation analyses were undertaken to assess the scope of
application and practical viability of the proposed solution. In this context,
the impact of the platform's velocity on radar performance under fixed
parameters, as commonly referenced in literature [19], was investigated.
Furthermore, the investigations were enhanced by examining the eigendecomposition of the clutter covariance matrix across
various radar and platform configurations.
The
last section presents the results of a complex simulation of a heterogeneous
environment in which a target was randomly placed. The task of the radar and
the STAP processing used was to eliminate clutter and
correctly detect the target.
The
radar and platform models as specified in paragraphs 3 and 4 were adopted for
the simulation, with the parameters included in the corresponding tables.
Tab. 1
Parameters adopted to simulation.
|
Parameter |
Value |
|
Antennas |
10 |
|
Pulses |
10 |
|
Radar operating
frequency |
8 GHz |
|
Wavelength |
0.0375 m |
|
Distance between antennas |
0.01875 m |
|
Platform flight
altitude |
2000 m |
|
Clutter ridge β |
0.4…2.0 |
|
Pulse repetition
frequency |
12 000 Hz |
|
Clutter to noise
ratio |
30 dB |
|
Signal-to-noise
ratio |
10 dB |
5.1. Scope of applicability and operational use of the developed
solution
Figure
3 depicts a graph illustrating the clutter's location in the received signal
for various velocities of the platform's flight. As previously established, the
velocity of the platform's flight directly influences the β parameter.
The β
parameter values are presented within the range of 0.4 to 2.0. For the
parameters outlined in Table 1 and a β parameter value of 1, a velocity
of 90 m/s was determined, representing the optimal scenario in terms of
target detection precision. Figure 3 illustrates that with an increase in the β parameter,
clutter (represented by more and more lines) emerges across a wider range of
azimuth angles observed by the radar.
5.2.
Analysis of the eigendecomposition of the clutter
covariance matrix for individual radar and platform parameters
Additionally,
the computational complexity, which directly impacts the processing time and
target detection efficiency, was calculated for the parameters listed in Table
1. The figure below uses vertical lines to represent the rows of the clutter
covariance matrix determined directly from Brennan's equation for various
values of the parameter β.
For the given simulation data, the row of the clutter covariance matrix is
5.5. Correct target detection
To
verify the accuracy of the proposed STAP method using
the SMI algorithm, an additional simulation was
conducted. A radar system consisting of 10 antennas operating at 10 GHz was
assumed. The distance between the antennas was set to half the wavelength,
which in this case is 0.015 m. The radar was mounted on an airborne platform
moving along the axis of the antennas at a constant speed of 225 m/s at an
altitude of 2000 m. For the given speed and pulse repetition frequency, the
parameter β is
equal to 1. The radar-cross section (RCS) of the target was set to 1 m².
Fig. 3. Graphs
of the location of clutter in the received signal for
different values of the platform's flight velocity
Fig. 4. Plot
of the eigendecomposition of the clutter covariance
matrix against the value of
the parameter β
A
commonly utilized model of heterogeneous forest-covered terrain, known in the MATLAB environment as the gamma model, was employed as
clutter. Additionally, a jammer was positioned in the field. The parameters
used for the simulation are presented in the following tables.
Tab.
2
Parameters adopted to simulation
|
Parameter |
Value |
|
Velocity
of target [ x, y, z] |
[40 m/s, 40 m/s, 0 m/s,] |
|
Location of target |
x = 1500; y = 1500; z = 0; |
|
Power of jammer |
100 W |
|
Location
of jammer [x, y, z] |
x = 1100; y = 1200; z = 0; |
|
Noise |
Gauss noise |
As a result of
the simulation, a string of 10 pulses was transmitted through the antenna
array, with a pulse repetition frequency of 30 kHz and a pulse duration of 33
µs. Then the signal reflected from the target but also the clutter from
the ground surface, the interference signal from the transmitter and the noise
on the receiving side were received.
Figure 5a illustrates the values of signals received by the
radar's antenna array as a function of range, following the transmission of the
first pulse. At this point, due to the presence of significant clutter, the
radar is unable to accurately determine the target's position. It is evident
that the radar incorrectly suggests the target is located 1000 meters away.
Figure
5b depicts the values of signals received by the
radar's antenna array as a function of range, after the first pulse
transmission. This time, however, the raw data has undergone STAP processing using the proposed SMI-STAP
algorithm implemented in the MATLAB environment. It
is clear that the radar correctly identifies the target as being approximately
2200 meters away in a straight line.
Fig. 6. Space
time spectrum of interference before (a) and after (b) SMI-STAP
processing
6.
SUMMARY AND CONCLUSION
The
article explores the potential for improving the safety of air transport,
particularly in militarized areas, by outfitting aircraft with SLAR and the STAP algorithm. It
outlines the theoretical foundations of STAP
processing and describes the specific steps of the classical SMI STAP processing algorithm.
The theoretical analyses, calculations, and simulation results conducted during
the study lead to the following conclusions:
-
in STAP processing, several approaches to
estimating the clutter covariance matrix are suitable, with a shift away from
statistical methods. At present, the most advanced techniques for estimating
the clutter covariance matrix are non-statistical methods;
-
the influence of the flight
parameters of the flying platform on which the radar is mounted, such as speed
and the angle of deviation from the axis of the antennas through the parameter β, which
uniquely determines the degree of clutter and the frequency of its occurrence,
is emphasized;
-
the paper is one of the few items in the literature on enhancing aircraft
safety with radar and STAP processing. It represents
an original contribution to the development of knowledge and radar technology,
and at the same time, this item, through the publication of some of the results
in world journals, represents an important point on the map of the development
of the field of STAP;
-
the theoretical analyses presented in the article, the calculations performed,
and the simulation results obtained are of great practical importance. The use
of the STAP technique in radar systems contributes to
increasing their usability and reducing computational complexity requirements.
In
summary, this paper demonstrates that it is feasible to detect slow-moving
objects (such as tanks, artillery, and drones) and enhance pilots' situational
awareness during flight missions, especially in air transport operations.
References
1.
Klemm Richard. 1998. Space-time
Adaptive Processing: Principles and Applications. The Institution of Electrical Engineers. ISBN:
978-0-8529-6946-5.
2.
Reed Irving, John Mallett,
Lawrence Brennan. 1974. „Rapid convergence
rate in adaptive arrays”. IEEE Trans. Aerosp. Electron. Syst. 10(6): 853-863. DOI:
10.1109/TAES.1974.307893.
3.
Ward James. 1994. Space-Time Adaptive Processing
for Airborne Radar. Lincoln Laboratory
Technical Report 1015.
4.
Guerci Joseph. 2014. Space-Time
Adaptive Processing for Radar. Artech House. ISBN:
978-1-6080-7820-2.
5.
Sarkar Tapan, Srikanth Nagraja, Michael Wicks. 1998. „A deterministic
direct data domain approach to signal estimation utilizing non uniform and
uniform 2D arrays”. Dig. Sig. Proc. 8:
114-125.
6.
Carlo Jeffrey, Tapan Sarkar,
Michael Wicks. 2003. „A Least Squares
Multiple Constraint Direct Data Domain Approach for STAP”.
In: 2003 IEEE
Radar Conference: 431-438. 5-8 May 2003, Huntsville,
AL, United States.
7.
Melvin Wiliam, Gregory
Showman. 2006. „An approach to knowledge-aided
covariance estimation”. IEEE Transactions on Aerospace and Electronic Systems 42(3):
1021-1042. DOI: 10.1109/TAES.2006.248216.
8.
Zhu Xumin, Peter Stoica. 2011. „Knowledge-aided
space-time adaptive processing”. IEEE Trans. Aerosp. Electron. Syst. 47(2): 1325-1333.
DOI: 10.1109/TAES.2011.5751261.
9.
Peng Hao, Yuze Sun,
Yang Xiaopeng. 2019. „Robust
knowledge-aided sparse recovery STAP method for
non-homogeneity clutter suppression”. The
Journal of Engineering 20: 6373-6376. DOI: 10.1049/joe.2019.0273.
10.
Chen Jie, Xiaoming Huo. 2006. „Theoretical results
on sparse representations of multiple-measurement vectors”. IEEE Trans. on Signal
Processing 54(12): 4634-4643. DOI: 10.1109/TSP.2006.881263.
11.
Duan Keqing, Zetao Wang, Wenchong Xie. 2017. „Sparsity-based STAP algorithm with multiple measurement vectors via sparse
Bayesian learning strategy for airborne radar”. IET Signal Processing 11(5): 544-553. DOI: https://doi.org/10.1049/iet-spr.2016.0183.
12.
Zang Wei. 2019. „Reduced dimension STAP based on sparse recovery in heterogeneous clutter environments”.
IEEE Trans. on Aerospace
and Electronics Systems 56(1): 785-795. DOI:
10.1109/TAES.2019.2921141.
13.
Cristallini Diego, Wolfram Bürger. 2012. „A robust direct data
domain approach for STAP”. IEEE Trans. on Sig.
Proc. 60(3):1283-1294.
DOI: 10.1109/TSP.2011.2176335.
14.
Cristallini Diego. 2012. „Exploiting robust
direct data domain STAP for GMTI
in very high resolution SAR”. In: 2012
IEEE Radar Conference: 0348-0353. 7-11 May 2012,
Atlanta, GA, United States.
15.
Guo Yiduo,
Guisheng Liao, Weike Feng.
2017. „Sparse
representation-based algorithm for airborne radar in beam-space post-Doppler
reduced-dimension space-time adaptive processing”. IEEE Access 5:
5896-5903. DOI: 10.1109/ACCESS.2017.2689325.
16.
Li Ming, Guohao Sun, Zishu He. 2019. „Direct Data Domain STAP Based on Atomic Norm Minimization”. In: 2019 IEEE Radar
Conference: 1-6. 22-26 April 2019, Boston,
Massachusetts, United States.
17.
Ma Zeqiang, Yimin Liu, Huadong
Meng. 2013. „Jointly sparse
recovery of multiple snapshots in STAP”. In: 2013 IEEE Radar Conference.
1-4. 29 April - 3 May 2013, Ottawa, Ontario, Canada.
18.
Satyabrata Sen. 2015. „Low-rank matrix
decomposition and spatio-temporal sparse recovery for
STAP radar”. IEEE Journal of Selected Topics
in Signal Processing 9(8): 1510-1523. DOI:
10.1109/JSTSP.2015.2464187.
19.
Knee Peter. 2012. Sparse representations for Radar
with MATLAB. Examples. Morgan
& Claypool. ISBN: 978-1-6270-5034-0.
Received 05.12.2024; accepted in
revised form 10.03.2024
Scientific Journal of Silesian University of Technology. Series
Transport is licensed under a Creative Commons Attribution 4.0
International License
[1]
Faculty of Aviation Division, Polish Air Force University, Dywizjonu
303 no 35 Street, 08-520 Dęblin, Poland. Email: b.slesicki@law.mil.pl. ORCID:
https://orcid.org/ 0000-0002-0857-1081
[2]
Institute of Navigation, Polish Air Force University, Dywizjonu 303 no 35 Street, 08-520 Dęblin, Poland. Email: a.slesicka@law.mil.pl. ORCID: https://orcid.org/ 0000-0002-6313-030X
[3]
Faculty of Mechatronics, Armament and Aerospace, Military University of
Technology, gen. Sylwestra
Kaliskiego 2 Street, 00 -908 Warszawa, Poland. Email:
adam.kawalec@wat.edu.pl. ORCID:
https://orcid.org/0000-0003-3930-7504