University of Birmingham
Phenomenology of Doppler forward scatter radar
for surface targets observation
Gashinova, Marina; Daniel, Liam; Hoare, Edward; Sizov, Vladimir; Cherniakov, Mikhail
DOI:
10.1049/iet-rsn.2012.0233
Document Version
Publisher's PDF, also known as Version of record
Citation for published version (Harvard):
Gashinova, M, Daniel, L, Hoare, E, Sizov, V & Cherniakov, M 2013, 'Phenomenology of Doppler forward scatter
radar for surface targets observation', IET Radar, Sonar and Navigation, vol. 7, no. 4, pp. 422.
https://doi.org/10.1049/iet-rsn.2012.0233
Link to publication on Research at Birmingham portal
General rights
Unless a licence is specified above, all rights (including copyright and moral rights) in this document are retained by the authors and/or the
copyright holders. The express permission of the copyright holder must be obtained for any use of this material other than for purposes
permitted by law.
•Users may freely distribute the URL that is used to identify this publication.
•Users may download and/or print one copy of the publication from the University of Birmingham research portal for the purpose of private
study or non-commercial research.
•User may use extracts from the document in line with the concept of ‘fair dealing’ under the Copyright, Designs and Patents Act 1988 (?)
•Users may not further distribute the material nor use it for the purposes of commercial gain.
Where a licence is displayed above, please note the terms and conditions of the licence govern your use of this document.
When citing, please reference the published version.
Take down policy
While the University of Birmingham exercises care and attention in making items available there are rare occasions when an item has been
uploaded in error or has been deemed to be commercially or otherwise sensitive.
If you believe that this is the case for this document, please contact UBIRA@lists.bham.ac.uk providing details and we will remove access to
the work immediately and investigate.
Download date: 10. Aug. 2022
Published in IET Radar, Sonar and Navigation
Received on 16th August 2012
Revised on 26th October 2012
Accepted on 12th November 2012
doi: 10.1049/iet-rsn.2012.0233
ISSN 1751-8784
Phenomenology of Doppler forward scatter radar for
surface targets observation
Marina Gashinova
1
, Liam Daniel
1
, Vladimir Sizov
2
, Edward Hoare
1
, Mikhail Cherniakov
1
1
Department of Electronic, Electrical and Computer Engineering, University of Birmingham, Birmingham B15 2TT, UK
2
Department of Radioelectronics, National Research University of Electronic Technology, Zelenograd, Moscow 124498,
Russia
E-mail: m.s.gashinova@bham.ac.uk
Abstract: In this study, the forward scatter Doppler phase signature formation is analysed to show the rationale for the forward
scatter radar in the true sense of the meaning, where a target actually crosses the baseline; so the advantage of the main shadow
lobe is taken and, therefore a forward scatter effect occurs to enhance signal to clutter ratio. The modelling approach suggested is
based on the consideration of the Doppler phase signature as a result of superposition of the direct path signal and the shadow
radiation signal. It is shown that the target signature may be represented as a Doppler signature of a point-like target specified by
its trajectory and speed, which is modulated according to forward scatter cross-section of an actual extended target specified by its
silhouette at each moment of motion. The proposed model may be recommended to provide matched filtering in coherent
processing. Finally, the approach is verified experimentally using calibrated targets with conductive and absorbing coating in
the controlled environment and maritime targets in the real sea conditions.
1 Introduction
Forward scatter radar (FSR) is historically thought of as the
first type of bistatic radar, which has been reported in [1].
FSR could be viewed as subclass of bistatic radar (BR)
where the bistatic angle is close to 180° [2–4], the ph ysical
operational principle of FSR is however essentially
different from that of BR, which is inherently a backscatter
radar. One of the major differences is that the target
signature is formed as a result of interference between the
direct path (or leakage) sign al and the scattered signal from
the target, rather than correlation of the received waveform
and locally generated hetero dyne reference in backscatter
radars. This dif ference affects the met hod of opti mal signal
processing [5–7]. Fundamentally the Doppler signature of
the target depends on the target electrical dimensions, its
shape , trajectory parameters and speed. Practically all these
parameters are unknown a priory and, therefore in order to
achieve optimal signal pro cessing on the reception side,
one should know the m eans of generating the waveform,
which replicates the real signal. To predict a signature the
correct model should be available based on co mprehensive
analysis of the forward scattering (FS) phenomena – this
constitutes the main subject o f the pape r. The propos ed
modelling procedure is based on consi deration of the two
main mechanisms contribu ting to Doppler signature
formation: first is the chi rp-like Do ppler waveform
formation because of frequency variation cau sed by
the target motion and the second is the envelope pattern
related to the target’s electrical dimensions and shape.
Optimal signal processing in this case provides not only
maximisation of the sign al to nois e rati o for better target
detection, but also allows an estimation of the target’s
trajectory and speed.
Recently a wave of interest has emerged in FSR ; firstly
this is a consequence of the introduction of ‘stealth’
targets. These targets have a significantly reduced radar
cross-section (RCS) because of their specific shapes and/or
coatings which m ay greatly suppress backscattering, yet
their s hadows will still render them perfectly ‘visible’ to
FSR. Secondly, interest in FSR has appeared because of
the establishment of passive coherent location concepts
[8–9] where illuminators of opportunity are used to form a
bistatic radar network.
This paper is concerned with the analysis of a class of FSR
where the target actually crosses the baseline (true FSR),
introducing a perturbation of the direct path signal and
producing a Doppler phase signature with a relatively long
observation time. Strictly speaking the radar under analysis
is a special-purpose radar intended to detect low profile,
low-speed targets, which represent a class of ‘difficult
surface targets’ where traditional monostatic radar has
limitations.
FSR exploits the so-called forward scattering (FS) effect
for enhanced target detection [4, 10, 11], which occurs if
the target’s electrical size produces scattering in the Mie
and optical regions. The FS effect is the strong increase of
RCS in the forward direction caused by the co-phase
interference of the waves arising from the shadow contour
of the scatterer. This results in a field focusing on the line
perpendicular to the object aperture in the shadow area.
With an increase in frequency the main shadow lobe
www.ietdl.org
422 IET Radar Sonar Navig., 2013, Vol. 7, Iss. 4, pp. 422–432
&
The Institution of Engineering and Technology 2013 doi: 10.1049/iet-rsn.2012.0233
narrows and its peak intensity becomes signi ficantly larger
than that of the backscattering lobe, with a maximum along
the axis of the main shadow lobe, that is, when the
transmitter, receiver and target form ∼180° bistatic angle
[2, 12].
Initially the FS phenomenon was thoroughly studied in
optics, where it was predicted by Mie [13], reported to be
observed for the Mie scattering region [14] and
quantitatively evaluated by the optical theorem for particles
[15]. Subsequently, the effect was investigated in studies
dedicated to the estimation of bistatic RCS of objects [11]
for microwaves and Ufimtsev [16] developed the physical
theory of diffraction.
Although the target is in motion, the scattering mechanism
undergoes a fundamental change: bistatic scattering (mainly
of the reflective nature) when the receiver is outside the
target main shadow lobe transforms into purely FS when
the target crosses the baseline and the receiver is in the
shadow. Signal analysis with a view to find the transition
between the two scattering mechanisms will, firstly, provide
a valuable insight into responses from so called ‘stealth
targets’ and, secondly, define the margins of applicability
for the suggested model where shadow radiation is the
prevailing diffraction mechanism. Results of measurements
in the controlled environment of an anechoic chamber will
be shown for metallic and absorbing targets of the same
geometrical cross-section.
The remainder of the paper is structured as follows:
initially, the concept of FSR is described. Next analytical
formulae for the Doppler frequency shift and target Doppler
phase signature along with a simplified forward scatter
cross-section model are presented. RCSs estimated
according to the presented approach are compared with
those simulated in three-dimensional (3D) full-wave
software. Then the validity of the suggested models in
optical and sub-optical scattering regions is discussed and
comparison with experimental results in both the anechoic
chamber and the real environment is given. The paper
concludes with a summary of the study’s research
contributions.
2 FSR phenomenology
In a bistatic radar configuration sp atia lly separated Tx and
Rx antennas are pointed to the area where the target of
interest app ears and it is assumed t hat the baseline
distance is comparable with the distances from Tx/Rx to
the target [17]. The signal at the input of the receive
anten na represents mainly bistatic reflections of the
transmitted signal from the interrogated target and only
this reradiated signal is required to extract information
about the target if the transmit and receive signals are
synchronised.
In contrast, in FSR the Rx and Tx antennas are facing each
other and there are two signals, which play equ ally
important roles in forming the sensed interference or ‘beat’
[4] chirp-like signal: the first is the strong direct path
signal, or ‘leakage’, which may be used only f or detection,
and the second is the much weaker forward scatter signal
from the mov ing target ‘modulating’ the leakage, which in
fact carries information on target trajectory, speed and
even size and profile.
Thus, the operational principle of FSR is based on
availability of both signals at the input of the receive antenna.
To provide the presence of both leakage and scattered
signal the target-radar topology should satisfy far-field
conditions for the target and the receive antenna: the size of
the target must be significantly smaller than distances to
both Tx and Rx and to the radius of the first Fresnel zone.
The target represents a source of secondary radiation with
respect to primary radiation from Tx according to
Huygens–Fresnel principle.
Using Fresnel parameter S = D
2
/(4λ), where D is a largest
effective size of target and λ is a wavelength we will
consider the scattering mechanism from a target as a
Fraunhofer diffraction (or far-field) at distances larger than
S. However when considering the time-varying Doppler
signature of a moving target, we should not confuse
Fraunhofer diffraction on the individual target with the
Fresnel-like diffraction on the effective inverse aperture
defined by the whole path of the moving target which is
‘seen’ by the radar.
Measured signatures of targets with sizes defining different
diffraction mechanisms are shown in Fig. 1 for 7.5 GHz
carrier and 300 m baseline: (a) Fraunhofer (far-field)
diffraction from a small inflatable boat of size 2.9 m × 1 m
(length and height above the surface), S = 60 m; (b)
boundary Fresnel to Fraunhofer, medium size sailing yacht
(5 m × 3 m), S = 160 m; and (c) Fresnel diffraction from
large motor boat (15 m × 4 m), S = 630 m.
The measured signals shown are the received signal
strength indicator (RSSI) signals, which contain the
oscillating Doppler signature on top of a DC level
indicating the strength of the direct path signal. In all
signatures, the typical Fresnel diffraction behaviour
(positive and negative contribution of phases of interfered
signals) is visible at least at the edges of the target signal
for cases (b) and (c). Obviously all three signals are liable
to detection and, moreover, (b) and (c) are not difficult
targets because their scattered signals are comparable with
the leakage signal. However, only the first signal is suitable
for the extraction of target motion parameters. Indeed, its
waveform is fully defined by diffraction reflecting specific
positions and speed of the target passing through the
constructive (in phase) and destructive (out of phase) zones
over the path. It should be stressed that in the middle of the
two-sided chirp-like signal the signal intensity is the same
as the intensity of the incident (direct path) signal, although
intuitively there should be a global minimum because of
shadowing. This phenomenon is similar to the Poisson
phenomenon (Arago spot) known in optics for Fresnel
diffraction.
The Arago spot is quite challenging to observe in optics
where the very small wavelength imposes the following
conditions: (i) target to be small, perfectly symmetrical
and having ideal edges, (ii) distances to the source and
the illuminated screen to be in the Fresnel zone and (iii)
source of light to be point-like. However scaling up the
wavelength, we can expect that a patter n similar to
Fresnel rings will appear if the target/wavelength ratio
and the total range of target movement satisfy the
conditions of Fresnel diffraction. Moreover, the much
larger scale and the use of the Doppler signature instead
of operational carrier frequency signals weakens the
strict conditions on symmetry and smooth edges of the
target silhouette on the line of sig ht, so that the Arago–
Poisson spot may appear as soon as the target is on the
line of sight. Mathematically it is expressed by the
presence of the non-zero imaginary part of the FS
amplitude f (θ = 0), which relates to the total scattering
www.ietdl.org
IET Radar Sonar Navig., 2013, Vol. 7, Iss. 4, pp. 422–432 423
doi: 10.1049/iet-rsn.2012.0233
&
The Institution of Engineering and Technology 2013
cross-section σ
tot
as
Im f (
u
= 0) =
k
s
tot
4
p
(1)
and explained by optical theorem [18, 19].
It should be stressed that the symmetry of the target
silhouette to the incident wave is still required for the
quasi-optical region; however, it is less strict when we
move down in frequency. We can suppose that if the
wavelength is nearly comparable to the effective ta rget
dimensions its asymmetry will be not resolved by the
incident and diffracted waves.
When the target moves, sequentially passing constructive
and destructive zones, the time domain waveform of the
Doppler signature develops in time in the same manner as
the Fresnel rings progress in space. The larger electrically
the target the less observable is the Arago spot and the
larger the intensity of shadow radiation.
3 Target signature in FSR
3.1 Doppler signature
In this paper, we concentrate on the true FS Doppler signature
when the target shadows the receiver. Following Ufimtsev
[10] the physical idea of the shadowed direct path signal
can be understood by considering the field at the receiver as
the result of interference of the incident electromagnetic
field and the shadow radiation from the scattering body.
Hence, the physical optics (PO) approximation will be the
right amount of theory to gain insight into the phenomenon.
The question of more accurate electromagnetic diffraction is
out of the scope of this paper and may be found in [16].
We assume that target has a uniform linear trajectory. This
is nearly always true for surface targets: they have a relatively
narrow FS CS pattern, consequently visibility time, or
signature length, is the order of seconds and it is not likely
that they will make a significant manoeuvre or change of
speed over this short time.
Shadow radiation is cast upon the receiver while the target
moves in the vicinity of the baseline according to the width of
the shadow lobe. The maximum of the shadow radiation
corresponds to the case when the target is on the baseline.
Being ‘shadow’ the forward scatter signal is π/2 phase
shifted (imaginary along the FS axis) relative to the direct
path signal [10, 19].
At this stage we omit any amplitude modulation of the
signal caused by propagation loss and by the FS CS
pattern. Only the phase signature of the point-like target
(yet casting shadow on the receiver) will be initially derived.
Later the total target signature will be presented as the
result of superposition of the point target phase signature
and complex envelope defined by FS CS of the extended
target.
At the receiver input a composition of the direct path signal
and delayed scattered signal from the moving target is
S
RI
(t) = S
DP
+ S
TG
= A
DP
cos(
v
0
t) + A
Tg
sin
v
0
(t + t
sh
)
(2)
where ω
0
is the carrier, t
sh
is the delay time of the signal from
moving target, S
DP
and S
TG
are direct path signal and
scattered target signal, respectively. The initial phase of
coherently acquired signals can be omitted without loss of
generality.
Owing to the fact that the Doppler signature in FSR
fundamentally occupies a very low-frequency band, in the
order of few Hz, the only way to detect this signal within
the background of the transmitter phase noise is to use self
mixing heterodyne, that is, to segregate this signature by
mean of non-linear transformation of the input signal. We
Fig. 1 Recorded Doppler signature of the targets crossing the centre of 300 m baseline
a Small inflatable boat
b Medium size yacht
c Large motor boat
www.ietdl.org
424 IET Radar Sonar Navig., 2013, Vol. 7, Iss. 4, pp. 422–432
&
The Institution of Engineering and Technology 2013 doi: 10.1049/iet-rsn.2012.0233
will consider as an example an envelope detector with
quadrature characteristic.
After passing the square law detector and low-pass filter in
the receiver the input signal transforms into
S
RO
(t) ≃
SLD
A
DP
cos (
v
0
t) + A
Tg
sin
v
0
(t + t
sh
)
2
≃
LPF
DC − A
Sc
sin
v
0
t
sh
t
sh
= R(t)+R(t)−L/c
()
= DC − A
Sc
sin
2
p
R
T
(t) + R
R
(t) − L
l
(3)
where DC =
A
2
DP
+ A
2
Tg
/(2)
≃
A
Tg
≪A
DP
A
2
DP
/2
is the power
of the leakage signal, A
Sc =
A
DP
A
Tg
characterises phase
signature amplitude, L is a baseline distance, R
T
(t) and
R
R
(t) are time-dependant ranges Tx-to target and
target-to-Rx accordingly.
In terms of Doppler phase shift
S
RO
(t) ≃
SLD
A
DP
cos(
v
0
t) + A
Sc
sin
v
0
+
v
d
t
2
≃
LPF
DC + A
Sc
sin(
v
d
t) (4)
where ω
d
is the Doppler frequency shift of the moving target.
Thus
v
d
t ; −
2
p
l
R
T
(t) + R
R
(t) − L
+ 2
p
n, n [ Z (5)
The last term can be omitted without loss of generality.
Traditionally [3, 20] bistatic Doppler frequency shift is
presented as a derivative of the variable phase shift
v
d
=−k
d R
T
(t) + R
R
(t)
dt
(6)
which to a first approximation may be solved considering ω
d
as a constant in form
v
d
t + C =−kR
R
(t) + R
T
(t)
(7)
Setting the origin of the time coordinate system as zero at the
moment of crossing the baseline we can find that the
integration constant C = − kL using the initial condition
R
T
(t)+R
R
(t) |
t =0
= L.
Finally
v
d
=
−(2
p
/
l
) R
R
(t) + R
T
(t) − L
t
which coincides with the expression obtained (5) for the
phase shift with removable sin gularity ω
d
=0at t =0
v
d
=−
2
p
l
R
T
(t) + R
R
(t) − L
t
(8)
Actual Doppler target signature extracted at the receiver has
an envelope A(t)defined by both path propagation loss and
FS CS
S
r
t()=At()sin
c
t()
[
]
(9)
where ψ(t)=2π/λ(T
T
(t)+R
T
(t) − BL) is the phase of the
moving point-like target.
3.2 Signature envelope
Analytical solutions for the FS CS are only available for the
few convex shapes using physical theory of diffraction
(PTD) approaches for optical and sub-optical scattering
regions [21–23]. For the Rayleigh region the diffraction
mechanism is m ore sophisticated and correct analytical
solutions are only available for the sphere and infinitely
long cylinder. Thus either approximated models, such as
suggested by PO or PTD [4, 22, 23] or full-wave EM
simulation methods [12, 24] must be used for target FS CS
estimation. In general without a prior knowledge of the
target shape it is only the phase target signature which can
be used for information extraction on target motion. Yet its
envelope defined by FS CS as function of time while the
target is moving in a specific bistatic configuration indicates
the size and shape of the target and enables rough
classification based on comparison with the database of
known targets [25].
In optical approximation for estimation of the target RCS at
FS direction the Babinet principle is used. A flat absorbing
screen of finite dimensions may be replaced by a
complementary infinite plane screen with an aperture
shaped exactly like the original screen. The incident field
diffracted at the aperture gives rise to the field coinciding
with the shadow field of the original absorbing screen
(except for the phase). If the incident wave is a plane wave,
as assumed by the target being in the Fraunhofer zone, then
the shadow field of a target at a distant receiving point
tends to be the radiation field of a flat aperture placed
perpendicular to the incident wave propagation direction
and determined by the target shadow silhouette. Thus the
target could be considered as an antenna of the silhouette
aperture with a negative gain which reduces the field
intensity at the reception side.
Summarising this, the shadow contour theorem [16]
declares that the shadow radiation in the optical case is
completely determined by the size and geometry of the
shadow contour.
As was stressed in [26] the RCS of a complex shape target
can be analysed by decomposing it into a number of basic
shapes which, when put together represent a replica of the
actual target. This is undoubtedly an approximation as the
finer-level interactions between parts will not be taken into
account, this is however acceptable for our purposes. For
the shadow radiation analysis we can represent the complex
shape of a target as a composition of the elementary shapes
with rectangular cross-section. Thus we will consider
scattering on the target with rectangular cross-section,
which is equivalent to the radiation by a rectangular
aperture antenna with the phase shifted by π/2 [22] and
fundamentally does not depend on the incident wave. Then
we can use the same methods of power budget and signal
analysis as for conventional radars, including the concept of
target RCS σ
tg
as a measure of the power, re-radiated by
the target in the direction of the observation point at
distance R:
s
tg
= lim
R1
4
p
R
2
E
r
2
/ E
i
2
, where E
i
and
E
r
are incident and tar get re-radiated electric field intensity,
respectively. In far field approximation σ
tg
in the direction
of Rx will be defined by the attitude of the aperture at
every moment of motion to the incident transmitted field
(Fig. 2) and viewing angles θ, φ from the receiver/
observation point (Fig. 2a).
www.ietdl.org
IET Radar Sonar Navig., 2013, Vol. 7, Iss. 4, pp. 422–432 425
doi: 10.1049/iet-rsn.2012.0233
&
The Institution of Engineering and Technology 2013
