4200 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 65, NO.
8, AUGUST 2017
Dust Storm Attenuation Modeling Based
on Measurements in Sudan
Elfatih A. A. Elsheikh, Student Member, IEEE, Md. Rafiqul Islam, Member, IEEE,
Mohamed H. Habaebi, Senior Member, IEEE, Ahmad F. Ismail, Senior Member, IEEE,
and Alhareth Zyoud, Student Member, IEEE
Abstract— Microwave (MW) and millimeter-wave (mmW)
propagation are severely affected by dust storms and sand
storms in arid and semi-arid areas. Electromagnetic waves may
suffer from attenuation due to suspended particles during a dust
storm. This paper proposes an empirical model to predict the
attenuation due to dust storms based on a one-year measurement
of visibility, humidity and their effects on MW links in Sudan.
Signal strength variations on two operational MW links at 14 and
22 GHz as well as visibility were monitored simultaneously. The
model is developed empirically using measured attenuation and
measured storm characteristics (e.g., visibility, dielectric constant,
frequency, and moisture content). The predicted attenuation from
the proposed empirical model is compared with the attenuation
at frequencies ranging from 7.5 to 40 GHz measured at different
locations, and good agreement is found. Additionally, this method
is characterized by simplicity and capability to predict reliable
dust storm attenuation for a wide range of frequencies and
moisture levels.
Index Terms— Attenuation, dust storm, microwaves (MWs),
millimeter waves (mmWs).
I. I NTRODUCTION
D UST storms affect not only human and animal health
but also communication links. The problems inherent
in modeling dust storm effects on microwave (MW) and
millimeter-wave (mmW) links have attracted considerable
interest recently because of the increasing number of terrestrial
links, satellite links, and high altitude platforms that encounter
sand and dust storms [1]–[4]. Such storms can reduce visibility
to nearly zero, halt flying operations, and break communi-
cation links for the duration of the storms. Sand and dust
storms occur in many parts of the world, such as the Middle
East, African Sahara, parts of Asia, arid parts of Australia,
and dry states such as Texas and Arizona in the Southwest
USA [5]. When MWs pass through a medium containing
precipitation, such as rain, snow, or dust particles, the MW
signals become attenuated from absorption and scattering of
energy by these particles [6], [7]. All theories and techniques
for scattering computation are generally based on solving
Maxwell’s equation either analytically or numerically.
Manuscript received July 31, 2016; revised May 16, 2017; accepted
June 4, 2017. Date of publication June 14, 2017; date of current ver-
sion August 2, 2017. (Corresponding author: Elfatih A. A. Elsheikh;
Md. Rafiqul Islam.)
The authors are with the Department of Electrical and Computer Engi-
neering, Faculty of Engineering, International Islamic University Malaysia,
Malaysia (e-mail: elfatihaa2005@hotmail.com; rafiq@iium.edu.my).
Color versions of one or more of the figures in this paper are available
online at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TAP.2017.2715369
0018-926X © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Analytical models have been developed to interpolate atten-
uation induced by dust storms. These models have been based
either on Rayleigh approximation [8]–[11] or Mie scattering
theory [12], [13]. Rayleigh approximation is suitable when the
radius of the particle is small compared to the wavelength [14].
On the other hand, the Mie solution does not have the
limitations of Rayleigh approximation.
However, Rayleigh approximation and Mie scattering are
based on single particle scattering theory and do not include
the multiple scattering effect. Although these attenuation
models are based on theoretical assumptions, they depend
on certain empirical inputs, such as frequency, visibility,
and dust particle characteristics, including shape, dielectric
constant, and size. Investigations have shown that dust par-
ticles are random in shape and cannot be assumed to be
spheres, ellipsoids, or any other shapes [15]. Sand and dust
particle distributions reported in the literature range from
0.05 to 0.150 μm [16]–[18].
Xiao-Ying [19] developed a model by formulating the wave
propagation constant based on the equivalent complex permit-
tivity using the Maxwell-Garnett formula. However, Maxwell
recognized that his formula is valid only when the volume
fraction of inclusions is small [20], [21].
On the other hand, the finite-difference time-domain
(FDTD) method has been implemented with the turning bands
method to formulate the scattering, absorption, and extinction
cross section of dust particles based on multiple scattering
effects [22]. However, the FDTD method must perform volu-
metric integration over a volume larger than the particle, which
is computationally time consuming. Additionally, the FDTD
method is inappropriate for applications involving groups of
randomly oriented particles [23].
All available models were found to be unable to predict
measured attenuation accurately. They were limited to certain
frequencies and were complicated and time consuming.
In this paper, the first empirical dust storm attenuation
prediction model is proposed. The model is developed to over-
come the complex nature of the real dust storm phenomena,
which is difficult to describe in analytical or numerical models.
The proposed empirical model is based on data from long-
term monitoring of MW links and an automatic weather
station (AWS) in Khartoum, Sudan. The model predicts the
attenuation as a function of visibility, frequency, dielectric
constant and moisture content. The model is characterized by
its simplicity, flexibility, reliability, and efficiency.
ELSHEIKH et al.: DUST STORM ATTENUATION MODELING BASED ON MEASUREMENTS 4201

The rest of this paper is organized as follows. Section II TABLE I

shows the available dust storm attenuation prediction models. D UST D IELECTRIC C ONSTANT AND E QUIVALENT
PARTICLE R ADIUS IN M ETERS
Section III presents the measurement setup and data collection.
Section IV compares the predicted attenuation with the mea-
sured data. Section V presents the proposed empirical model.
The validation of the proposed model and discussion of the
results is presented in Section VI. The paper is concluded in
Section VII.

II. AVAILABLE P REDICTION M ODELS

follows:
The effect of dust storms on MW and mmW attenuation are
1
generally estimated by solving the forward scattering ampli-
tude function of a single particle. The solutions are determined
using the Rayleigh approximation, Mie equations or numerical
methods. For randomly distributed particles in air, the sin-
gle scattering approximation or multiple scattering theory is
applied, according to the concentration of the particles in
the air.
Based on Rayleigh approximation, Ahmed et al. [9] devel-
oped a dust-induced attenuation prediction model expressed as
  
1 re
α = 566.97 (G)[d B/km] (1)
V0 λ following equation:
k
where α is the attenuation, λ is the wavelength in meters, and
ε’ and ε” are the real and imaginary parts of the dielectric
constant of the dust particles. G is a constant that depends

on the particle dielectric constant (ε = ε + j ε" ).V0 is the
optical visibility in kilometers. re is the equivalent particle
radius in meters, equal to 48.3 μm. Goldhirsh [10] derived an
attenuation prediction model for dust storms expressed by
2.317 · 10−3 · ε" 1 tion models discussed earlier.
A= · [d B/km]
[(ε + 2)2 + ε"2 ] · λ V γ
where V is the visibility in kilometers, and γ is a constant
value equal to 1.07. λ is the wavelength in meters.
Based on Mie theory, Elabdin et al. [12] derived another
attenuation prediction model as follows:
ae f   Maygooma–Koukou operating at 14.4 and 21.3 GHz with
A= x + yae2 f 2 + zae3 f 3 [d B/km]
V
where f is the frequency in GHz. ae is the equivalent particle
radius in meters, equal to 30 μm. x, y, z are constants that

depend on the particle dielectric constant (ε = ε + j ε").
Sharif [13] developed a model based on Mie scattering. Sharif
considered the variation of the dust particles dimensions in his
model as follows:
1 
AP = c1" f + c2" f 3 + c3" f 4 [d B/km]
V 1.07
where A P is the attenuation considering particle dimensions.
c1" , c2" , and c3" are constants that depend on the particle dielec-

tric constant (ε = ε + j ε" ).
Xiao-Ying et al. [19] developed a model for attenuation
prediction using the effective material property technique and
the general formulation of the complex propagation factor as
εeq   2

2π
α = 8.686 · · 1+tan2 δ−1 [d B/km] (5)
λ 2
εeq
"
tan δ =  (6)
εeq
where ε and ε" are the real and imaginary parts of εeq
∗ , and ε ∗
eq
is the complex relative permittivity of sand and dust.
Chen and Ku [22] used the FDTD method implemented
with the turning band method to predict attenuation due to
dust storms based on multiple scattering effects given in the
A = 8.686 × 103 σext (k r)N(k r ) r [d B/km] (7)
k=1
where K is an integer number of rmax / r, and rmax and
r are the maximum particle radius in the storm and the
incremental radius, respectively. σext is the extinction cross
section.
Table I gives the values of the important parameters required
to predict the attenuation using dust storm attenuation predic-
(2)
III. E XPERIMENTAL S ETUP AND DATA C OLLECTION
The experimental system consists of radio links, metrolog-
ical sensors, and data acquisition and processing hardware.
The MW links under monitoring were Shakeer–Magharba and
(3) 2.6 and 2.8 km, respectively.
These links were located in northeast Khartoum, approx-
imately 5 km from the Khartoum airport. Meteorological
data were collected from an AWS, operating in Khartoum
International Airport. Visibility was measured using Vaisala
Transmissometer model LT31 [24]; an optical range from
10 to 10 000 m with 1 min integration time and accuracy
of ±3% were obtained. Rain fall rate is measured with 1 min
integration time and accuracy of ±1%.
(4) Additionally, the AWS is equipped with sensors for tem-
perature (T ), relative humidity (RH), wind speed and wind
direction [24]. Tables II and III show detailed specifications
of the MW links; Fig. 1 shows a map of the MW links and
the automatic weather station.
From careful monitoring of the MW links, signal level
variations of 0.7 and 2.5 dB have been recorded due to diurnal
and seasonal temperature variation, respectively.
4202 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 65, NO. 8, AUGUST 2017

TABLE II
D ETAILED D ESCRIPTION OF S HAKEER –M AGHARBA

TABLE III
D ETAILED D ESCRIPTION OF M AYGOOMA –K OUKOU

Fig. 2. Cumulative distribution of measured visibility from June 1, 2014 to

May 31, 2015 in Khartoum.

Fig. 1. Locations of MW links and the AWS in Khartoum.

A. Data Collection
During the experimental period from June 1, 2014 to
May 31, 2015, more than 22 dust storms were experienced
in the city of Khartoum. Metrological parameters as well as
transmitted and received signal levels were recorded during
the storm events.
1) Visibility: Optical visibility has typically been used
to measure the severity of dust storms. Consequently,
Fig. 2 shows the cumulative distribution function of the optical
visibility based on the collected data using an LT31 sensor.
Fig. 2 indicates that 0.01% of the time during the year
visibility was ≤83 m, whereas 20% of the time during the
year visibility was ≤9 km; beyond the point where visibility
equals 9 km an exponential improvement in the visibility has Fig. 3. (a) Received signal level and visibility on June 6, 2014. (b) Received
been experienced. signal level and RH on June 6, 2014.
Moreover, a thorough investigation of the effects of an
actual dust storm on the MW signals provides a better under-
standing of the complex phenomena. Fig. 3(a) and (b) shows dropped from 3000 to 500 m and returned to 6000 m at
an analysis of dust storm events on June 6, 2014. 18:14. The second attack started at 18:15 when the visibility
Fig. 3(a) shows that Khartoum has been hit by a dust reduced from 4800 m to a very low value during the storm.
storm with several peaks. In the first attack at 16:15, visibility Visibility then increased to 1000 m after half an hour and
ELSHEIKH et al.: DUST STORM ATTENUATION MODELING BASED ON MEASUREMENTS 4203

Fig. 4. Cumulative distribution of measured attenuation at 14.4 and 21.3 GHz Fig. 5. Measured and predicted attenuation at 14 GHz.
links from June 1, 2014 to May 31, 2015 in Khartoum.

continued to return to its prestorm value for another half an

hour. However, at the first attack, the visibility was reduced to
500 m and a small change was recorded in the received signal
level. It has been observed that some turbulent events moved
slowly in arbitrary directions and did not occur concurrently
at weather stations and along the link; the first attack might
have been a similar event. On the other hand, simultaneous
degradation in the signal level was observed at the second
attack when the visibility nearly approached zero. At a high
visibility of V > 500 m, limited changes in the received
signal level are experienced, whereas in a severe dust storm
with V< 500 m, a significant degradation in the received signal
level is observed.
2) Relative Humidity: Fig. 3(b) shows that the RH increased
drastically from approximately 20% to 70% during the dust
storm measurement. This rapid increase in the RH directly Fig. 6. Measured and predicted attenuation at 22 GHz.
affects the moisture content of the dielectric constant of the
dust and consequently degrades the signal significantly due to
changes in the dielectric characteristics of the dust particles. were collected in Khartoum, Sudan from June 1, 2014 to
3) Attenuation Due to Dust Storm: Fig. 4 shows the cumu- May 31, 2015 with 92% availability. Visibility data were
lative distribution functions of the measured attenuation due collected for 21 dust storm events; only one event was missed
to the dust storm for two MW links with 2.6- and 2.8-km during the monitoring period.
path lengths operating at 14.4 and 21.3 GHz, respectively.
Measured attenuation has been converted to dB/km, assuming IV. P REDICTION M ODELS V ERSUS M EASURED DATA
uniform intensity of dust storm over the link.
However, the intensity of sand/dust storms is not uniform; Measured attenuation is compared at the same con-
a study in Riyadh, Saudi Arabia concluded that the dust storm dition to the predicted attenuation obtained by Gold-
cell size was generally in excess to 20 km [25]. Consequently, hirsh [10], Ahmed [9], Zain [12], Xiao-Ying [19], Chen [22],
in Sudan, attenuation has been measured based on short links and Sharif [13]; the measured and predicted attenuation
with path lengths of 2.6 and 2.8 km as shown in Fig. 4. at 14 and 22 GHz are plotted in Figs. 5 and 6, respectively.
Hence, the assumption of uniform dust storm intensity can be Equations (1)–(5) and (7) are used to predict attenuation
considered valid over these short path lengths. Additionally, at 14 and 22 GHz.
if the intensity of the dust storm is not uniform over the entire Fig. 5 shows that the values of the measured attenuation are
link, a slight overestimation or a slight underestimation may 6.3, 1.95, and 0.71 dB/km at visibilities of 0.084, 0.279, and
occur during analysis of each event; the misestimations tend 0.936 km, respectively. All six models highly underestimated
to cancel each other out after adding several events throughout the measured attenuation. Ahmed’s model predicts the high-
a year. est attenuation: 0.58, 0.175, and 0.052 dB/km at visibilities
of 0.084, 0.279, and 0.936 km, respectively. Fig. 6 shows
B. Data Availability that the values of the measured attenuation were 11.42,
The receive signal levels for the MW links with 2.71, and 0.925 dB/km at visibilities of 0.083, 0.279, and
2.6 and 2.8 km at 14.4 and 21.3 GHz, respectively, 0.936 km, respectively. Ahmed’s predictions were 0.92, 0.275,
4204 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 65, NO. 8, AUGUST 2017
and 0.082 dB/km at visibilities of 0.084, 0.279, and 0.936 km,
respectively.
Therefore, the available models are not capable of predicting
the attenuation accurately at both frequencies, especially at
low visibility with the high particle concentrations, as shown
in Figs. 5 and 6. Attenuation in dust storms occurs mainly
from two physical mechanisms, which are absorption and
scattering of the dust particles. However, at higher frequen-
cies, the scattering effect becomes more pronounced. The
computations of these effects require knowledge of electrical
properties of the scattering particles in addition to some other
mechanical properties, such as particle shape and size. For all
analytical and numerical models proposed earlier, attenuation
prediction was based on the assumption of the spherical shape
of dust particles, arbitrarily considered average dust particle
sizes, and very limited measurements on dust particle dielectric
constant [13].
From our recent measurements in Sudan, it has been deter-
mined that dust particles are random in shape and cannot
be assumed to be spheres, ellipsoids, or any other shape.
Additionally, dust particle samples collected in Khartoum had
a normal distribution of the dust particle diameters with means
of 47 and 66 μm. Moreover, the dielectric constant of sand and
dust varies according to frequency and moisture content [18].
From careful study, it can be observed that the dust particle
shape, size and variation of dielectric constants with moisture
contents are the main source of underestimation in attenuation
predictions.
V. P ROPOSED E MPIRICAL M ODEL
Even though many mathematical, physical, and numeri-
cal models have been developed recently, an experimental
approach is still needed. Such an approach can integrate all
complex effects during the dust storm, which may be very
difficult to incorporate in other approaches. Hence, empirical
models are considered more reliable for investigating the
effects of dust storms on the operation and design MW and
mmW links [26].
Generally, in cases when the investigator has little or no
control over events, the experimental approach is preferred,
especially when focusing on a contemporary phenomenon
within the context of a real environment. However, direct
observation is the only certain method for extrapolating the
magnitude and frequency of propagation impairments on
microwave links [27].
In developing the empirical model, the following observa-
tions and derivation steps have been followed.
1) The model is developed based on measured visi-
bility, RH, and attenuation at 22 GHz as shown
in Figs. 2, 3, and 4, respectively.
2) Fig. 3(b) shows that during the dust storm, there was
a certain amount of RH present in the sand and dust
particles.
3) If the moisture is present in the dust and sand particles,
electromagnetic characteristics of dust and sand particles
change significantly.
4) When moisture is present in the dust and sand particles,
the real and imaginary parts of the dielectric constant
TABLE IV
VALUES OF c, d , AND e TO E STIMATE ε" IN (16) AT
D IFFERENT M OISTURE C ONTENTS
will vary due to their dependence on frequency and
moisture [18]. Moreover, measurement has shown that
dust can absorb 5.1% of moisture by weight in air with
82% RH [28].
5) A study on the effect of the frequency and moisture
showed that real part ε has no significant relation with
the frequency at moisture content (m v < 20%). On the
other hand, it has been found that ε" has a significant
correlation with the frequency in the 1–20 GHz range,
as shown in Fig. 7 [29].
6) Using regression, it is apparent that attenuation
(i.e., the response variable) is affected by visibility
and the imaginary part of dielectric constant (i.e., the
predictors) as follows:
A ≈ F(V, ε") (8)
Equation (8) can be re-expressed to separate visibility
and the dielectric constant, thereby proposed as
A ≈ F(V )F( f, m) (9)
where V is visibility, and ε" is the imaginary part of
the dielectric constant. f is the frequency, and m is the
moisture content.
7) An equation that relates the imaginary part of the dielec-
tric constant ε" to the frequency at different moisture
contents can be extrapolated from Fig. 7 as follows:
ε" = c × ( f d ) + e (10)
where the values of c, d, and e depend on the volu-
metric moisture content present in the dust, as shown
in Table IV.
8) Using the measured attenuation at 22 GHz from
Fig. 4 and the known relation of the imaginary part of
the dielectric constant (10) at a given frequency and
moisture content, the contribution of visibility in the
attenuation is estimated as follows:
W = A/(ε" ) (11)
where W is the contribution of visibility in the attenua-
tion and A is the measured attenuation at 22 GHz.
ELSHEIKH et al.: DUST STORM ATTENUATION MODELING BASED ON MEASUREMENTS
Fig. 7. Frequency dependence of imaginary part of permittivity of
sand for pure water for several volumetric moisture content ranging
from 0.05–0.6 cm3/cm3.
TABLE V
VALUES OF a AND b IN (18) TO E STIMATE THE C ONTRIBUTION
OF V ISIBILITY IN THE M EASURED ATTENUATION AT
D IFFERENT M OISTURE C ONTENTS
Based on the attenuation measured at 22 GHz, the values
of W are calculated at different moisture levels.
9) Using the trust region curve fitting algorithm, the con-
tribution of visibility to the attenuation due to the dust
storm is extrapolated as follows:
W = a × Vb (12)
where V is the visibility in km, a and b are constants
representing the contribution of the visibility as shown
in Table V.
10) Using the values of a and b to predict the attenua-
tion at different moisture contents, 0.1 cm3 /cm3 vol-
umetric moisture content achieves good predictions at
14 and 22 GHz. Therefore, parameters a and b are fixed
at this 0.1 cm3 /cm3 value.
11) Finally, the proposed prediction model is as follows:
A = (a × V b ) × ((c × f d ) + e)d B/km (13)
n (y − yi )2
where V is the visibility in km, a= 0.6301, b = −1.123,
f is the frequency in GHz, mv is the volumetric moisture
content, c, d, and e are parameters representing the
imaginary part of the dielectric constant at different
frequencies and moisture per Table V.
The empirical model, which was developed based on long-
term data collection, is a function of visibility in terms of
kilometers, dielectric constant, frequency in terms of GHz, and
4205
Fig. 8. Measured and predicted attenuation at 14 and 22 GHz.
moisture content. As shown in (13), the model is very simple
and requires only the measured visibility and frequency data.
The analytical and numerical prediction models depend
on empirical or semi-empirical input parameters, such as
particle shape, particle size, dielectric constant, visibility, and
frequency.
VI. VALIDATION OF THE P ROPOSED M ODEL
Measurements were carried out in Khartoum on two
MW links. The links operate at 14 (Ku-band) and 22 GHz
(Ka-band) with path lengths of 2.6 and 2.8 km, respectively.
During the measurement period from June 1, 2014 to
May 31, 2015, more than 22 dust storms were recorded in
Khartoum. The measured attenuation at 14 and 22 GHz was
compared with those predicted by the proposed model in (13)
and presented in Fig. 8. Additionally, similar measurements
reported in Riyadh, Saudi Arabia at 40 GHz over a 14 km
MW link [25] were also compared with the proposed model
and presented in Fig. 9. Figs. 8 and 9 show that the attenuation
obtained by the proposed empirical model shows good agree-
ment with the measured attenuation. Generally, attenuation
due to sand and dust storms depends heavily on frequency,
visibility, and complex permittivity of the particle, as shown
in Figs. 8 and 9.
To validate the proposed empirical model, the root-mean-
square error (RMSE) values of attenuation in dB/km for
the proposed model and the available models at 14, 22,
and 40 GHz were calculated using (14) [30] and presented
in Table VI. The proposed model achieved the best RMSE
values of 0.35, 0.26, and 0.022 dB/km at frequencies of 14,
22, and 40 GHz, respectively. RMSE is calculated as follows:

RM S E = [d B/km] (14)
=1=1 n
where y is the predicted specific attenuation and yi is the
measured specific attenuation.
A few additional measurements were reported in Khartoum
on a 20-km link operated at 7.5 GHz with visibility less than
200 m and on a 15-km link operated at 13 GHz with visibility
less than 50 m [31] [32].
4206
C ALCULATED RMSE FOR THE P REDICTION M ODELS AT 14, 22, AND 40 GH Z
M EASURED AND P REDICTED ATTENUATION AT 7.5 AND 13 GH Z
Fig. 9. Measured and predicted attenuation at 40 GHz.
Table VII shows the measured and predicted attenuation at
7.5 and 13 GHz. The results, shown in Table VII, indicate
that the proposed model predictions are in good agreement
with measured attenuation at 7.5 and 13 GHz.
Thus, a simple and efficient empirical model to predict
attenuation due to dust storms has been proposed. The model
has been validated at frequencies ranging from 7.5 to 40 GHz.
Furthermore, the dust particles scattering effects and pre-
diction increase as frequency increases. Moreover, measured
and predicted attenuations increase as the visibility decreases
(e.g., particle concentration increases). Therefore, attenuation
is directly proportional to frequency and inversely proportional
to visibility.
VII. C ONCLUSION
This paper investigates the effects of propagation on MWs
and mmWs in a dust storm. Meteorological parameters, such
as visibility, humidity and their effects on 14 and 22 GHz MW
propagation, were measured concurrently for a one-year period
in Sudan. The first empirical model was developed based on
measured data and can be used to predict attenuation due to
dust storms. The proposed model is a function of visibility,
frequency, moisture content, and dielectric constant to predict
IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 65, NO. 8, AUGUST 2017
TABLE VI
TABLE VII
attenuation with high reliability. The proposed model reflects
all measurements more accurately than contemporary physical
and mathematical models.
The model is simple in its expression and is characterized
by flexibility in accommodating a wide range of frequencies,
as well as moisture levels varying from dry to wet dust.
The proposed model has been validated with measurements
at different locations at 7.5, 13, 14, 22, and 40 GHz and has
found very good agreement. However, to improve the accu-
racy of the proposed model, measurements of dust dielectric
constant behavior at higher frequency bands with different
moisture contents are required.
ACKNOWLEDGMENT
The authors would like to thank MTN Sudan for giving
permission to monitor their links and to the Sudan Mete-
orological Authority for providing metrological information
on a regular basis. The authors would also like to thank the
Research Management Center, International Islamic University
Malaysia, for support this research project.
R EFERENCES
[1] K. Harb, S. Abdillah, and S. Abdul-Jauwad, “Dust & sand (DUSA)
storms impact on LEO satellite microwave radio links,” in Proc.
ASMS/SPSC, 2014, pp. 442–447.
[2] E. A. A. Elsheikh, R. Islam, A. Zyoud, A. H. M. Z. Alam, and
Z. O. Elshaikh, “A proposed vertical path adjustment factor for dust
storm attenuation prediction,” in Proc. ICOM, 2011, pp. 1–3.
[3] E. A. A. Elsheikh, M. R. Islam, J. Chebil, S. Bashir, and
A. F. Ismail, “Preliminary analysis of dust effects on microwave prop-
agation measured in Sudan,” in Proc. ICOM, 2013, pp. 1–7.
[4] M. R. Islam, E. A. A. Elsheikh, A. H. M. Z. Alam, and O. O. Khalifa,
“Proposing a horizontal path adjustment factor for microwave link’s
attenuation prediction based on the analysis of dust storm behavior,”
in Proc. MICC, 2011, pp. 44–48.
[5] Z. E. O. Elshaikh, M. R. Islam, O. O. Khalifa, and H. E. Abd-El-Raouf,
“Mathematical model for the prediction of microwave signal
attenuation due to duststorm,” Prog. Electromagn. Res. M, vol. 6,
pp. 139–153, 2009. [Online]. Available: http://www.jpier.org/
PIERM/pier.php?paper=09021906
[6] R. K. Crane, “Propagation phenomena affecting satellite communication
systems operating in the centimeter and millimeter wavelength bands,”
Proc. IEEE, vol. 59, no. 2, pp. 173–188, Feb. 1971. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1450053
[7] Q. Lei, “Study of microwave attenuation from dust particle,” in
Advanced Research on Computer Education, Simulation and Modeling,
vol. 176, S. Lin and X. Huang, Eds. Berlin, Germany: Springer, 2011,
pp. 379–383.
ELSHEIKH et al.: DUST STORM ATTENUATION MODELING BASED ON MEASUREMENTS
[8] T. S. Chu, “Effects of sandstorms on microwave propagation,”
Bell Syst. Tech. J., vol. 58, no. 2, pp. 549–555, 1979. [Online].
Available: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=
6770986
[9] A. S. Ahmed, A. A. Ali, and M. A. Alhaider, “Airborne dust
size analysis for tropospheric propagation of millimetric waves into
dust storms,” IEEE Trans. Geosci. Remote Sens., vol. GE-25, no. 5,
pp. 593–599, Sep. 1987. [Online]. Available: http://ieeexplore.ieee.org/
stamp/stamp.jsp?tp=&arnumber=4072690
[10] J. Goldhirsh, “Attenuation and backscatter from a derived two-
dimensional duststorm model,” IEEE Trans. Antennas Propag.,
vol. 49, no. 12, pp. 1703–1711, Dec. 2001. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=982449
[11] A. Musa and S. O. Bashir, “Electromagnetic waves propagation in
dust storms at millimetre wave band,” J. Emerg. Trends Eng. Appl.
Sci., vol. 4, no. 2, pp. 162–167, 2013. [Online]. Available: http://jeteas.
scholarlinkresearch.com/articles/Electromagnetic%20Waves.pdf
[12] Z. Elabdin, M. R. Islam, O. O. Khalifa, H. E. A. Raouf, and
M. J. E. Salami, “Development of mathematical model for the prediction
of microwave signal attenuation due to duststorm,” in Proc. ICCCE,
2008, pp. 1156–1161.
[13] S. M. Sharif, “Attenuation properties of dusty media using Mie
scattering solution,” Prog. Electromagn. Res. M, vol. 43, pp. 9–18,
Jun. 2015. [Online]. Available: http://www.jpier.org/PIERM/pierm43/
02.15022403.pdf
[14] R. E. Collin, “Radio wave propagation,” in Antennas and Radio-Wave
Propagation. New York, NY, USA: McGraw-Hill, 1985, pp. 401–410.
[15] S. I. Ghobrial and S. Sharief, “Microwave attenuation and cross polar-
ization in dust storms,” IEEE Trans. Antennas Propag., vol. 35, no. 4,
pp. 418–425, Apr. 1987. [Online]. Available: http://ieeexplore.ieee.org/
stamp/stamp.jsp?tp=&arnumber=1144120
[16] A. S. Ahmed, “Role of particle-size distributions on millimetre-wave
propagation in sand/dust storms,” IEE Proc. H-Microw., Antennas
Propag., vol. 134, no. 1, pp. 55–59, Feb. 1987. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4642968
[17] E. A. A. Elsheikh, M. R. Islam, A. H. M. Z. Alam, A. F. Ismail,
K. Al-Khateeb, and Z. Elabdin, “The effect of particle size distributions
on dust storm attenuation prediction for microwave propagation,” in
Proc. ICCCE, 2010, pp. 1–5.
[18] A. J. Ansari and B. G. Evans, “Microwave propagation in sand
and dust storms,” IEE Proc. F-Commun., Radar Signal Process.,
vol. 129, no. 5, pp. 315–322, Oct. 1982. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4645433
[19] X. Y. Dong, H. Y. Chen, and D. H. Guo, “Microwave and millimeter-
wave attenuation in sand and dust storms,” IEEE Antennas Wire-
less Propag. Lett., vol. 10, pp. 469–471, 2011. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5770171
[20] A. H. Sihvola and J. A. Kong, “Effective permittivity of dielec-
tric mixtures,” IEEE Trans. Geosci. Remote Sens., vol. 26, no. 4,
pp. 420–429, Jul. 1988. [Online]. Available: http://ieeexplore.ieee.org/
stamp/stamp.jsp?tp=&arnumber=3045
[21] R. Landauer, “Electrical conductivity in inhomogeneous media,” in Proc.
AIP Conf., 1978, pp. 2–45.
[22] H.-Y. Chen and C.-C. Ku, “Calculation of wave attenuation in
sand and dust storms by the FDTD and turning bands methods
at 10–100 GHz,” IEEE Trans. Antennas Propag., vol. 60, no. 6,
pp. 2951–2960, Jun. 2012. [Online]. Available: http://ieeexplore.
ieee.org/stamp/stamp.jsp?tp=&arnumber=6183464
[23] F. M. Kahnert, “Numerical methods in electromagnetic scattering the-
ory,” J. Quant. Spectrosc. Radiat. Transf., vols. 79–80, pp. 775–824,
Jun./Sep. 2003.
[24] V. Oyj, “Vaisala transmissometer LT31 RVR with new eyes,” Vaisala,
Helsinki, FL, USA, Tech. Rep. B210416en-C, 2010.
[25] M. A. Alhaider and A. A. Ali, “Experimental studies on millimeterwave
and infrared propagation in arid land: The effect of sand storms,” in Proc.
ICAP, vol. 2. 1989, pp. 268–270.
[26] M. R. Islam, E. A. A. Elsheikh, A. F. Ismail, S. O. Bashir, and J. Chebil,
“Development of an empirical dust storm attenuation prediction model
for microwave links in arid area—A proposed framework,” in Proc.
ICCCE, 2014, pp. 224–227.
[27] E. A. A. Elsheikh, M. R. Islam, A. F. Ismail, M. H. Habaebi, and
J. Chebil, “Dust storms attenuation measurements at 14 GHz and 21 GHz
in Sudan,” in Proc. ICCNEEE, 2015, pp. 11–16.
[28] S. M. Sharief, “Backscatter properties of dusty medium at X band,” in
Proc. ICAP, vol. 2. 1989, pp. 424–427.
4207
[29] Y. Lasne et al., “Effect of salinity on the dielectric properties of
geological materials: Implication for soil moisture detection by means
of radar remote sensing,” IEEE Trans. Geosci. Remote Sens., vol. 46,
no. 6, pp. 1674–1688, Jun. 2008. [Online]. Available: http://ieeexplore.
ieee.org/stamp/stamp.jsp?tp=&arnumber=4524018
[30] J. F. Kenney, Mathematics of Statistics. New York, NY, USA: D. Van
Nostrand Company, 1962, p. 59.
[31] S. I. Ghobrial, M. A. Hemeidi, and M. E. Tawfig, “Observations
on 2 and 7.5 GHz microwave links during dust storms,” Electron.
Lett., vol. 23, no. 1, pp. 44–45, Jan. 1987. [Online]. Available:
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=4257240
[32] Z. Elabdin, M. R. Islam, O. O. Khalifa, and A. F. Ismail, “Duststorm
measurements for the prediction of attenuation on microwave signals in
Sudan,” in Proc. ICCCE, 2008, pp. 1181–1185.
Elfatih A. A. Elsheikh (S’17) received the B.Sc.
degree in electrical and computer engineering from
Omdurman Islamic University, Omdurman, Sudan,
in 2000, and the M.Sc. and Ph.D. degrees in elec-
trical engineering from International Islamic Uni-
versity Malaysia, Malaysia, in 2011 and 2017,
respectively.
His current research interests include wireless
channel modeling, radio link design, RF propagation
measurement, and modeling in arid area.
Md. Rafiqul Islam (M’02) received the B.Sc.
degree in electrical and electronic engineering from
the Bangladesh University of Engineering and Tech-
nology, Dhaka, Bangladesh, in 1987, and the M.Sc.
and Ph.D. degrees in electrical engineering from
the University of Technology Malaysia, Malaysia,
in 1996 and 2000, respectively.
He is currently a Professor with the Depart-
ment of Electrical and Computer Engineering, Fac-
ulty of Engineering, International Islamic University
Malaysia, Malaysia. He has supervised more than
50 Ph.D. and M.Sc. students. He has authored more than 200 research papers
in international journals and conferences. His current research interests include
wireless channel modeling, radio link design, RF propagation measurement
and modeling in tropical and desert, RF design, smart antennas and array
antennas design, free space optics propagation, and modeling.
Dr. Islam is a Life Fellow of Institute of Engineers Bangladesh.
Mohamed H. Habaebi (M’99–SM’16) received
the bachelor’s degree from the Civil Aviation and
Meteorology High Institute, Sabah, Libya, in 1991,
the M.Sc. degree in electrical engineering from Uni-
versiti Teknologi Malaysia, Malaysia, in 1994, and
the Ph.D. degree in computer and communication
system engineering from Universiti Putra Malaysia,
Malaysia, in 2001.
He is currently an Associate Professor and the Post
Graduate Academic Advisor with the Department of
Electrical and Computer Engineering, International
Islamic University Malaysia, Malaysia, where he heads the research works on
Internet of Things. He has supervised many Ph.D. and M.Sc. students. He has
authored more than 120 articles and papers and serves on editorial board
of many international journals. His current research interests include M2M
communication protocols, wireless sensor and actuator networks, cognitive
radio, small antenna system, radio propagation and wireless communications,
and network performance evaluation.
Dr. Habaebi is an active reviewer for many international journals.
4208 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 65, NO. 8, AUGUST 2017
Ahmad F. Ismail (M’06–SM’13) received the Elec-
trical Engineering (Cum Laude) degree from Gannon
University, Erie, PA, USA, the M.Sc. degree in
telecommunication and information systems from
the University of Essex, Colchester, U.K., and the
Ph.D. degree in electronic from the University of
Bath, Bath, U.K.
He is currently a Lecturer with the Depart-
ment of Electrical and Computer Engineering, Fac-
ulty of Engineering, International Islamic University
Malaysia, Malaysia. He is acknowledged by the
Board of Engineering Malaysia as a Professional Engineer with Practicing
Certificate. His current research interests include microwave and milimetre-
wave propagation studies, radio frequency, and wireless systems technology
enhancement.
Alhareth Zyoud (S’15) received the bachelor’s
degree in electrical engineering from Palestine poly-
technic University, Hebron, West Bank, in 2006,
and the master’s degree in communication engineer-
ing from International Islamic University Malaysia,
Malaysia, in 2011, where he is currently pursuing
the Ph.D. degree with the Department of Electrical
and Computer Engineering.
His current research interests include interference
cancellation, modeling, and propagation studies in
femtocells 4G and 5G networks.