Content uploaded by Othman Zahid
Author content
All content in this area was uploaded by Othman Zahid on Jan 18, 2022
Content may be subject to copyright.
URSI GASS 2020, Rome, Italy, 29 August - 5 September 2020
Long Term Rain Attenuation Measurements at Millimeter Wave Bands
for Direct and Side Short-Range Fixed Links
O. Zahid, J. Huang, and S. Salous*
Department of Engineering, Durham University, Durham, DH1 3LE U.K., sana.salous@durham.ac.uk
Abstract
Millimeter wave (mmWave) radio links are largely
affected by precipitation. In this paper, we use a custom-
designed continuous wave (CW) channel sounder to
record channel data at K band (25.84 GHz) and E band
(77.52 GHz) for direct line of sight link and a side non
line of sight link with dual polarizations. A high-
performance PWS100 disdrometer is utilized to collect
weather data, including rain rate and rain drop size
distribution (DSD) for rain attenuation study. The rain
attenuation for both links are compared. The side link
exhibits a slightly higher attenuation than the direct link.
The ITU-R P.838-3 model and DSD model are applied to
model the rain attenuation. The results will be useful for
the design of fixed links for fifth generation (5G)
mmWave communication systems.
1 Introduction
MmWave communication is a key technology for 5G and
short-range fixed links can be used for fronthaul,
backhaul, and building to building transmission to provide
high data transmission rate. A signal with frequency
above 10 GHz can be affected by rain, which causes
performance reduction and even outage of the
communication system, especially at mmWave bands.
Since rain is not uniform in space and time, long-term
measurements over several years and various locations are
needed. Its effects on mmWave propagation include
absorption and scattering. Rain attenuation depends on the
rain fall rate, rain drop size distribution, DSD, and
complex water refractive index, which is related to the
frequency band.
Several studies of rain attenuation at mmWave bands
have been reported. For terrestrial links, most of the
measurements are dedicated for long-range direct links
where the transmitter (Tx) and receiver (Rx) antennas are
in line-of-sight (LOS). In [1], a link was set up at 71-76
GHz over 1 km. The rain data were recorded by a
visibility and precipitation optical sensor. The measured
rain attenuation was shown to agree with the ITU-R
P.838-3 model. In [2], a link was set up at 93 GHz with a
link distance of 850 m. In [3], a link was set up at 38 GHz
with over 1.85 km. The wet antenna attenuation was
found to have a maximum value of 2.3 dB. In [4] and [5],
a microwave link was set up at 72 GHz and 84 GHz with
a link distance of 23.6 km. The rain data were recorded by
an optical disdrometer. The results indicated that the ITU-
R P.838 model may overestimate the rain attenuation. In
[6], four microwave links were set up at 14.5 GHz with a
distance range of 12.8-43 km and a new distribution
model was proposed for the fade-slope statistics. In [7], a
microwave link was set up at 26 GHz with a link distance
of 1.3 km. The worst month statistic obtained from the
real measurements was lower than what was predicted by
the ITU-R P.581-2 model. In [8], a link was set up at 73,
83, 148, and 156 GHz with a link distance of 325 m. In
[9], eight links were set up over a frequency range of
37.3-39.2 GHz with link distances of 48-497 m. The
results showed different wet antenna attenuations for
different rain rates, which indicated that the constant
value of wet antenna attenuation used in some studies
may overestimate the retrieved peak rain rates. In [10],
five microwave links were set up at 32 GHz with link
distances of 186-1810 m. The wet antenna attenuation
was about 3 dB per antenna for a rain rate of 100 mm/h.
In [11], a link was set up at 23, 25, 28, and 38 GHz with a
link distance of 700 m. The results showed that the wet
antenna effect, change of humidity level, and equipment
stability would also impact the rain attenuation.
Compared to the above studies, our research concentrates
on short-range building to building transmission
scenarios. Both the LOS direct link and non-line-of-sight
(NLOS) side links are set up at K band (25.84 GHz) and E
band (77.52 GHz) with dual polarizations. The distance
between the Tx and Rx is about 35 m for the direct link.
The rain attenuation study for both links is then related to
the ITU-R P.838-3 model using the rain fall rate and the
DSD model with Mie scattering.
The remainder of this paper is organized as follows.
Section 2 describes the measurement system setup. In
Section 3, the rain attenuation models are introduced. The
results and analysis are presented in Section 4. Finally,
conclusions are drawn in Section 5.
2 Measurement System Setup
2.1 Fixed link experimental setup
The experimental setup reported in [12, 13] was updated
as shown in Fig. 1 and in Fig. 2. Compared to the
previous system, the intermediate frequency (IF) unit is
replaced by a phase-locked loop (PLL) and integrated
with the radio frequency (RF) heads into a single box to
Authorized licensed use limited to: University of Durham. Downloaded on October 25,2020 at 12:58:29 UTC from IEEE Xplore. Restrictions apply.
reduce the complexity of the system and enhance its
stability.
In addition, the antennas which already had radomes are
further covered to reduce the wet antenna effect. For the
direct link, two antennas with vertical and horizontal
polarizations are used for the E band, while a dual
polarized antenna is used for the K band. For the side link
receiver, single vertical polarized antennas are used for
both bands. A 50 Hz trigger and 40 MHz clock are
connected to the two data acquisition cards where the 50
Hz signal is also connected to the Tx to switch between
vertical and horizontal polarizations and the 40 MHz
provides the sampling clock. The IF signal to be digitized
is set at 4 MHz and 12 MHz for the K band and E band,
respectively.
(a) Tx
(b) Rx for direct link
(c) Rx for side link
(d) Weather station
Figure 1. Tx and Rx boxes.
Figure 2. Block diagram of the channel sounder set up.
2.2 Weather station
The PWS100 weather station shown in Fig. 1(d) is
installed at Durham University, UK within 200 m distance
to the link. It is a laser-based sensor which can identify
precipitation type and measure the rain rate and DSD. The
recorded data include visibility, precipitation intensity,
precipitation type, temperature, relative humidity, DSD,
etc. The DSD is recorded with 300 values of the number
of drops corresponding to the diameter from 0.1 mm to 30
mm with 0.1 mm resolution.
The DSD in unit of (݉ିଷ݉݉ିଵ) can be calculated as
ሺ୧ሻൌ൫୧ǡ୨൯
୨
ଷ
୨ୀଵ ͳ
ο୧
(1)
where S = 40 cm2 is the measurement surface of the laser
beam of the PWS100 disdrometer, t = 60 s is the
integration time, n(Di,vj) is the number of particles
registered within the classes with mean diameter Di (mm)
and mean speed vj (m/s), dDi (mm) is the class width
associated with the diameter Di.
3 Rain Attenuation Models
Two models are widely used for rain attenuation. These
are the ITU-R P.838-3 model with distance factor and the
DSD model with Mie scattering. The ITU model relies on
the rain fall rate, while the DSD model with Mie
scattering relies on the DSD.
3.1 ITU-R P.838-3 model
The ITU model is given as
ߛൌ݇
(2)
where R is the rain fall rate (mm/h), k and Dare model
parameters dependent on the frequency f, and is the
specific attenuation in dB/km. The ITU model provides a
look-up table for the parameters for different frequencies
and polarizations.
The total attenuation for a specific distance depends on
the effective path length deff, between the Tx and Rx
antennas is expressed as
ܣൌߛ݀
(3)
where the effective path length, deff, of the link is obtained
by multiplying the actual path length d by a distance
factor r, given in the ITU-R P.530-17 model as
ݎൌ ͳ
ͲǤͶ݀ǤଷଷܴǤଵ
Ǥଷఈ݂Ǥଵଶଷ െͳͲǤͷͻሺͳ െሺെͲǤͲʹͶ݀ሻሻ
(4)
where R0.01 is the rain rate exceeded for 0.01% of the time
with an integration time of 1 minute.
Authorized licensed use limited to: University of Durham. Downloaded on October 25,2020 at 12:58:29 UTC from IEEE Xplore. Restrictions apply.
3.2 DSD model and Mie scattering
The DSD model is given as
ߛ ൌ ͶǤ͵Ͷ͵ ൈ ͳͲଷනߜ
௫௧
ஶ
ሺܦሻܰሺܦሻ݀ܦ
(5)
where γ is the specific attenuation in dB/km, ߜ௫௧ ൌ
ߨሺ
ଶሻଶܳ௫௧is the extinction cross section (ଶ) for water
drops of diameter D (mm), and N(D) is the drop size
distribution value ( ݉ିଷ݉݉ିଵ ) at diameter D. The
extinction efficiency Qext can be calculated from Mie
scattering or Rayleigh scattering theory depending on the
size parameter x = ߨ D/λ, where λ is the wavelength. The
complex water refractive indices, at an average
temperature 5oC, at 25.84 GHz and 77.52 GHz are
5.1253+2.7265i and 3.5330+1.6724i, respectively.
4 Results and Analysis
Four rain events and rain attenuation measurements were
recorded during 5-18 December 2019. The measurement
results are shown in Fig. 3 for the direct and side links.
The relative signal levels are compared to sunny days.
(a) Direct link
(b) Side link
(c) Rain rate
Figure 3. Direct and side link rain attenuations at K and E
band.
For the direct link, the attenuation follows the trend of
rain for both bands since there is a strong LOS
component. For the side link, it relies on NLOS reflected
paths to receive the signal and the attenuation can be
impacted by scattering from the rain. The mapped
received signal in dBm against the rain rate in mm/h
shows a correlation between rain rate and rain attenuation
for both links with the side link exhibiting higher
attenuation than the direct link.
The cross polarization shows higher attenuation for both
bands and both links. An accurate modeling of the side
link rain attenuation is possible by finding the propagation
paths through methods such as wideband channel
measurement and ray tracing simulation.
The rain attenuation calculated from the ITU-R P.838-3
model and the DSD model for both bands for the direct
link vertical polarizations are shown in Fig. 4. As can be
seen, both models are in good agreement.
(a) Modeled rain attenuation at K band
(b) Modeled rain attenuation at E band
Figure 4. Modeled rain attenuation at K band and E band.
When converting the calculated rain attenuations to the
specific path length, the distance factor should be
considered, as rain is not uniformly distributed in space.
Since the empirical equation for the distance factor is
mainly derived from long-range measurement datasets,
the recommended maximum value of 2.5 is not
appropriate for short-range links [13]. The calculated
distance factor without the maximum value constraint is
applied for the short-range link. Fig. 5 shows the
comparison of measured and modeled rain attenuations at
K band for the direct link. The measured attenuation is
larger than the ITU model and DSD model due to the wet
antenna effect (WAE), especially for a short-range link.
The model used to estimate the WAE is given as
Authorized licensed use limited to: University of Durham. Downloaded on October 25,2020 at 12:58:29 UTC from IEEE Xplore. Restrictions apply.
ܣൌܥሺͳെ݁ିௗ
ሻ
(6)
where ܣ is the measured attenuation, ܣ is the
attenuation from WAE, C, d, and n are the fitting
parameters [14].
For the direct link in the K band, the fitting parameters are
ܥൌͳͲ, ݀ൌെͲǤͲ, and ݊ൌͳǤʹ. The estimated
attenuation from the WAE is shown in Fig. 5(a). After
removing the WAE from the measurements, the measured
attenuation matches well with the predicted attenuation by
the ITU model, as shown in Fig. 5(b).
(a) Estimation of WAE
(b) Comparison of measured and estimated rain
attenuations
Figure 5. Measured and modeled rain attenuation for K
band direct link.
5 Conclusions
In this paper, we have conducted rain attenuation
measurements at two mmWave bands for both direct and
side links. The PWS100 disdrometer and custom-designed
channel sounder have been used to collect weather data
and channel data, respectively. The ITU-R P.838-3 model
and DSD model were used to model the rain attenuation.
The rain attenuation for the direct and side links have
been investigated. The side link is shown to have larger
attenuation, thus is more challenging for fixed link rain
attenuation modeling.
6 Acknowledgements
The authors would like to acknowledge the support of
WaveComBE project, under Horizon 2020 research and
innovation program with grant agreement No. 766231, the
support from Ofcom under grant No. 1362, and the
electronics and mechanical workshop technicians at
Durham University for the setup of the link.
7 References
1. J. Hansryd, Y. Li, J. Chen, and P. Ligander, “Long term
path attenuation measurement of the 71–76 GHz band in a
70/80 GHz microwave link,” in Proc. EuCAP’10, Barcelona,
Spain, Apr. 2010, pp. 1–4.
2 V. Kvicera, M. Grabner, and O. Fiser, “4-year hydrometeor
attenuation statistics obtained at 93 GHz on an 850 m
terrestrial path,” in Proc. EuCAP’11, Rome, Italy, Apr. 2011,
pp. 2501–2503.
3. M. Schleiss, J. Rieckermann, and A. Berne,
“Quantification and modeling of wet-antenna attenuation for
commercial microwave links,” IEEE Geosci. Remote Sens.
Lett., 10, 5, pp. 1195–1199, Sept. 2013.
4. E. S. Hong, S. Lane, D. Murrell, N. Tarasenko, and C.
Christodoulou, “Terrestrial link rain attenuation
measurements at 84 GHz,” in Proc. USNC-URSI NRSM’17,
Boulder, CO, 2017, pp. 1–2.
5. E. S. Hong, S. Lane, D. Murrell, N. Tarasenko, C.
Christodoulou, and J. Keeley, “Estimating rain attenuation at
72 and 84 GHz from raindrop size distribution measurements
in Albuquerque, NM, USA,” IEEE Geosci. Remote Sens.
Lett., 16, 8, pp. 1175–1179, Aug. 2019.
6. F. J. A. Andrade, A. A. M. de Medeiros, and L. A. R. da
Silva Mello, “Short-term rain attenuation predictor for
terrestrial links in tropical area,” IEEE Antennas Wireless
Propag. Lett., 16, pp. 1325–1328, May 2017.
7. I. Shayea, T. Abd. Rahman, M. Hadri Azmi, and M. R.
Islam, “Real measurement study for rain rate and rain
attenuation conducted over 26 GHz microwave 5G link
system in Malaysia,” IEEE Access, 6, pp. 19044–19064,
Mar. 2018.
8. L. Luini, G. Roveda, M. Zaffaroni, M. Costa, and C. Riva,
“EM wave propagation experiment at E band and D band for
5G wireless systems: Preliminary results,” in Proc.
EuCAP’18, London, UK, Apr. 2018, pp. 1–5.
9. M. Fencl, P. Valtr, M. Kvicera, and V. Bares,
“Quantifying wet antenna attenuation in 38-GHz commercial
microwave links of cellular backhaul,” IEEE Geosci. Remote
Sens. Lett., 16, 4, pp. 514–518, Apr. 2019.
10. P. Valtr, M. Fencl, and V. Bares, “Excess attenuation
caused by antenna wetting of terrestrial microwave links at
32 GHz,” IEEE Antennas Wireless Propag. Lett., 18, 8, pp.
1636–1640, Aug. 2019.
11. C. Han and S. Duan, “Impact of atmospheric parameters
on the propa-gated signal power of millimeter-wave bands
based on real measurement data,” IEEE Access, 7, pp.
113626–113641, Aug. 2019.
12. S. Salous, Y. Cao, and X. Raimundo, “Impact of
precipitation on millimetre wave fixed links,” in Proc.
EuCAP’19, Krakow, Poland, Apr. 2019, pp. 1–4.
13. J. Huang, Y. Cao, X. Raimundo, A. Cheema, and S.
Salous, “Rain statistics investigation and rain attenuation
modeling for millimeter wave short-range fixed links,” IEEE
Access, 7, pp. 156110–156120, Oct. 2019.
14. M. M. Z. Kharadly and R. Ross, ‘‘Effect of wet antenna
attenuation on propagation data statistics,’’ IEEE Trans.
Antennas Propag., vol. 49, no. 8, pp. 1183–1191, Aug. 2001.
Authorized licensed use limited to: University of Durham. Downloaded on October 25,2020 at 12:58:29 UTC from IEEE Xplore. Restrictions apply.
