PS.

04/08/2017

PUSAT PENGAJIAN SISWAZAH

CENTRE FOR GRADUATE STUDIES
UNIVERSITI TUN HUSSEIN ONN MALAYSIA

PROGRESS REPORT FOR RESEARCH STUDENT

SEMESTER 2 SESSION 2022/23

TITLE
Performance and Measurement Study for Rain Rate and Rain Attenuation for 6G in Tropical
Region (Malaysia)

BACKGROUND

PROBLEM STATEMENT
In this current world, technology and innovation are rapidly increasing while getting more
complicated and interconnected. Technology is applied in most tasks, including communication
systems, medical devices, financial transactions, and electrical systems, which help make lifestyle
easier and time-saving. The latest wireless technology that humans have reached is 5G technology.
Every decade, technology has revolved around a strong reason for the limitation. As an example,
when the first generation mobile network began, it was limited to slower data speeds. As a
breakthrough, the second generation was introduced to enhance the data speed by a few hundred
along with text messaging features. From here, the technology evolved to a satellite internet
connection and improved with a data speed of a few megabits which is called a third-generation
mobile network. Later in the 2000s, this wireless technology improved with fourth-generation
networks to improve the quality of streaming from third-generation mobile networks. Started with
wired channels for long-distance communications which stay within the boundary and limit the
 PS.04/08/2017

usage level, with continuous experiments and technology leading to wireless medium
communication all around the world. As can conclude the main reason triggering engineers,
scientists and inventors was the limitation, quality and problems faced during the usage of every
invention. Even though there were many changes, enhancements and improvements made from
the previous technologies which are from the first generation(1G), second generation(2G), third
generation(3G), and also fourth generation(4G), with this current 5G technology there are still
interferences and limitations occur. The fact is that the fifth generation has reached worldwide and
is currently in use at a few locations in the major pullback for this technology. It is only able to
benefit the cities and some of the remote areas still have not been able to receive coverage for
years. Furthermore, fifth-generation (5G) technology has a smaller broadcasting range compared
to fourth-generation (4G). Even larger structures and trees block the bandwidth of 5Gbps. To
overcome this issue, there is a solution to have additional towers for better coverage which is
expensive and more time-consuming as well. Besides that, the weather is one of the major reasons
to interfere with or slow down the wireless connection in the fifth generation. The rain which is in
a liquid form had a broad absorption spectrum in microwave wavelengths. When it is rainy weather,
between the transmitter and receiver, signals fall into these frequencies and some of the signals
will be absorbed. Additionally, the rain also scatters the signals if the frequency of the signal and
size of the liquid droplets gets along in the right way. To improve this connection with a higher
signal rate, researchers, engineers, scientists and inventors are working on sixth-generation
technology(6G), to get away from the limitations that occur from the fifth generation(5G). As the
sixth generation is still not in use in real life, characterizing the performance of the sixth generation
will be carried as an effect of rain attenuation. The performance of sixth-generation technology will
be analyzed and measured using MATLAB Simulink. Finally, a deep learning model will be developed
as a prediction for rain attenuation.

SCOPE

1. A set of rain rate data will be collected for the experiment.

2. Terahertz frequency will be used to represent the sixth generation(6G)
network.
3. Using MATLAB Simulink, will produce predicted results on rain attenuation
using rain rate data.
4. A deep learning model will be developed to predict rain attenuation.

OBJECTIVES
1. To characterize the performance of the sixth generation(6G) network as an effect of rain
attenuation.
2. To measure the sixth generation(6G) network performance using rain rate data in
MATLAB Simulink.
3. To develop a deep learning model for the prediction for the rain attenuation model.

LITERATURE REVIEW
 PS.04/08/2017

As the 6G technology uses higher frequencies, the effects of weather become more clear. Most
problems with transmission systems caused by weather happen at mm and THz frequencies.
Attenuation of a terahertz link with four frequency bands over a distance of one kilometre when
it's raining. Rain absorption is more affected by the rate and frequency of rain than by how the
raindrops are spread out. The effect of rain is clear, but the effects of how often it rains are not
well known. The current statistical method, which uses average attenuation, is less accurate,
especially for the high frequencies that will be used soon. It is also wrong because it doesn't take
into account how the rate of rain changes during a single rain event. So, rain absorption needs to
be looked at in places where it rains a lot in the tropics to see how it affects the performance of
6G links.

1. 6G Terahertz Wireless Communication Rain Attenuation Characterization

In the approaching decades, wireless communication will need tremendous bandwidth, data
throughput, and latency. Demand is rising and will soon be great. An ultra-high-speed connection
in the terahertz (THz) frequency spectrum (275 GHz to 3 THz) can satisfy demand. Sixth-generation
(6G) communication systems have high data rates and capacity needs that millimetre wave
technology cannot provide. Terahertz links are impacted by atmospheric conditions, yet highly
directed antennas reduce attenuation and open the door to future uses. This study showed
terahertz uses, losses, a brief comparison between mm-wave and terahertz, and rain attenuation.
This study outlines inclement weather applications, losses, and attenuation profiles to tackle this
problem and open the door to a new ultra-high-speed future.

i) THz Attenuation

Attenuation limits ultra-high-frequency spectrum use. Terahertz links must account for numerous
sorts of losses. In terahertz, correlate atmospheric dispersion, turbulence, and attenuation in three
dimensions to propagation effects in a communication medium. Power loss from atmospheric
attenuation reduces receiver signal amplitude. The atmosphere disperses frequency components
with different phase velocities, causing group delay. Atmospheric turbulence fluctuates the THz
signal. Amplitude and phase may vary. The losses caused three reasons. Atmospheric turbulence
is very unpredictable due to temperature, humidity, air velocity, and pressure. Near Earth, it
increases amplitude and phase deviations. This study will describe how rain attenuation depends
on raindrop size because raindrops spread the signal. Sleet and snow attenuate signals, although
their refractive indices and number density distributions differ. Cloud particle dispersion causes
cloud attenuation. Water droplets, ice crystals, or both make up a cloud, according to the WMO.
When a moving wave interacts with a cloud, electromagnetic wave absorption and dispersion cause
extinction and attenuation. Gas attenuation from oxygen, water vapour, and nitrogen dominate
typical climatic circumstances. The gas particle absorbs the signal. Each particle has its own
chemical structure and spectral absorption line. The extinction process affects scattering loss,
which is caused by meteorological conditions. Any item between RX and TX might block the direct
route. When line-of-sight communication is disrupted, multipath propagation causes propagation
loss.

ii) Rainfall Model

Rain attenuation shifts from millimetres to sub-millimetre wavelengths when raindrop size
approaches electromagnetic wave wavelength. Increased frequency attenuates the terahertz
connection. Particulate size and precipitation rate determine attenuation, which is 3 dB/km at 0.313
THz for 2 mm/h precipitation. Mie scattering shows rain attenuation. ITU-R and Mie scattering
 PS.04/08/2017

theory anticipate precipitation attenuation. If dispersion is smaller than absorption, use the
Rayleigh approximation.

iii) Performance Analysis

Attenuation depends on precipitation rate and rain particle quantity and dispersion. Attenuation
grows nonlinearly with precipitation rate and frequency (Figure 1). In Fig. 1, the RX-TX
communication connection is 100 meters. Four links with different lengths are assessed here. From
1 GHz to 1000 GHz, attenuation (dB) rises with frequency. Link distance lowers attenuation.

This study simulates circumstances and precipitation attenuation. Signal attenuation rises with
frequency and rain precipitation rate at a fixed connection. Simulations show that signal
attenuation increases with distance. When the transmission distance is maintained constant at 1
kilometer and signal attenuation is simulated for varying tilt degrees, higher frequencies have
higher attenuation. Finally, high-frequency links attenuate less and behave virtually consistently at
varying elevation angles. Future 6G wireless communication systems may use terahertz
communication. Terahertz band's exhilarating properties will dominate worldwide communication.
Many applications use this frequency spectrum. Casualties have hampered this band's
performances. This research models rain attenuation under different settings and explores this
projected frequency band's properties and losses. Due to its huge bandwidth, ultra-high data
throughput, low power consumption, and minimal end-to-end latency, terahertz band will be vital.
6G will boost wireless data rates to 1 Tbps per device. 6G's high data throughput requires the
terahertz spectrum. This study aids wet channel modelling. In snowfall, fog, and gaseous settings,
terahertz links will be studied.

2. Communication systems performance at mm and THz based on rain rate probability density
function model.

6G will revolutionise human-machine communication with greater frequencies. Weather, especially

rain, attenuates transmission signals at higher frequencies. Long-term averaging is used to
determine how precipitation affects the symbol mistake rate. Inaccurate techniques lead to
inefficient system design. Due to bandwidth and energy constraints, this requires a higher margin
of error. Recently developed a more accurate approach for estimating communication system
performance in rain using the probability density function of rain rate. Our research shows that the
symbol error rate is orders of magnitude higher than ITU standard methods for high rainfall rates
(more than 10 mm/h) and usual parameters. Our model also suggests detecting and monitoring
rain rate likelihood to give customers the required bit error rate. This will enable the creation of
more efficient, adaptable systems that function better under wet situations. The authors believe
this analysis is innovative. It may improve performance metrics for 6G communication and reduce
system design-related breakdowns. Theoretical impairment models were used to simulate and
assess fade mitigation techniques. Signal frequency and rain intensity increase signal attenuation.
Rainfall forecasts influence attenuation. This research introduces a technique that accounts for
rainfall rate while calculating symbol error rate. This differs from traditional methods that employ
average values. This work also provides a closed-form formula for error probability. Numerical
analysis shows that the density distribution function of rain rate greatly increases the symbol error
rate. Changing the power transmitter, modulation constellation, and/or coding intensity may
ensure the desired bit error rate. This research sought to add the rain rate probability—a substantial
signal transmission attenuator—in the symbol error rate estimate. Clouds, atmospheric gases, dust,
and sandstorms dampen. At higher frequencies, wavelength and particle size cause signal error
significantly. These attenuation techniques vary. Small raindrops' wave dispersion may increase
 PS.04/08/2017

attenuation. DSD-aware attenuation models should be used in higher-frequency research. Future

studies may add more atmospheric restrictions to our estimate. Research should consider
troposphere scintillation. Wind turbulence in the troposphere mixes the largely horizontal layers
with similar refractive indexes. Troposphere mixing causes rapid refractive index variations on
microscopic scales and short timescales. These rapid index changes affect wave amplitude and
phase. Dry tropospheric scintillation follows. Rain is a slower kind of tropospheric scintillation.
Scintillation does not absorb because the signal mean level is unaltered. The influence is frequency-
dependent since scale size causes larger oscillations at shorter wavelengths.

Given the coming advent of 6G and the communication, it will enable between humans and life-
saving devices, an accurate symbol error rate estimate is crucial. 6G's high frequencies make it
particularly susceptible to rain-induced attenuation and distortion. Thus, communication
technology requires accurate symbol error estimates.

4. Rain Rate and Attenuation Measurements Over 26 GHz Microwave 5G Link System in
Malaysia

This research examines how rain affects 26 GHz millimetre wave propagation. A fifth-generation
microwave radio connection system at Universiti Teknologi Malaysia Johor Bahru helped measure.
1.3 kilometres. Radio, rain gauge, and data logger comprised the system. Ericsson CN500 compact
E-link. With minute-by-minute temporal resolution, measurements were made and recorded
everyday for a year. After that, the MATLAB code processed and analysed the annual rain rate and
attenuation, including for the wettest month. Based on the findings of the research, it was
determined that the rain rate was 120 millimetres per hour for 0.01% of the time, while the specific
rain attenuation was 26.2 decibels per km and the overall rain attenuation for 1.3 kilometres was
34 decibels. In addition, the data that was gathered from the measurements conducted during the
worst month showed results that were worse than those projected by the model developed by the
International Telecommunication Union (ITU). The rain rate (51%) and attenuation (34%) were
similar. When it came to the rain rate and rain attenuation, the computed average percentage of
error between the measurements and the projected values was 143% and 159%, respectively.
Averaging both data sets determined this. Therefore, it is possible to draw the conclusion that the
figures for the worst month in Malaysia are better than what the ITU model projected would be the
case.In this article, the rain rate and rain attenuation of the 26 GHz frequency band connection
were evaluated and discussed based on actual measurements carried out using microwave 5G radio
link systems in Malaysia. These measurements were taken over a period of three months in
Malaysia. Based on the findings that were provided and analysed, it was discovered that the rain
rate at 0.01% was 120 millimetres per hour, the specific rain attenuation was 26.2 decibels per km,
and the overall rain attenuation across 1.3 kilometres was 34 decibels. In addition, the worst month
statistic that was derived from the actual observations was lower than what was anticipated by the
ITU model; specifically, the rain rate and rain attenuation were lower by about 51% and 34%,
respectively. The computed average percentage of error between the measured and anticipated
values for the rain rate and rain attenuation was 143% and 159%, respectively. These figures are
expressed as a percentage of the overall value. As a result, the conclusion that was reached was
that the number for the worst month in Malaysia is less than what was projected by the ITU model.
 PS.04/08/2017

4. Rain attenuation prediction for 5G communication links in Minna, Nigeria

In the near future, millimetre-wave systems will provide mobile data rates that are faster than the
current communication systems in Nigeria, however, at these high frequencies, severe signal fading
brought on by rain is a significant drawback. Accurately estimating rain-induced attenuation is one
of the major issues facing 5G millimetre wave technology. In this work, rain attenuation and 1-
minute rain rate statistics for 5G communication lines propagating at 26 GHz and 38 GHz along a
path-length range from 1 km to 5 km are presented. The data was recorded using a MadgeTech
Rain101A rainfall data recorder and a tipping bucket rain gauge type. The research produced a rain
rate of 89 mm/hr that surpassed 0.01% of the time. The route attenuation was calculated with the
ITU-R P.530-17 model. Horizontal polarisation attenuation values at 26 GHz are 19.85 dB, 28.09 dB,
35.28 dB, 41.64 dB, and 47.58 dB for 1 km, 2 km, 3 km, 4 km, and 5 km path lengths, respectively,
while vertical polarisation attenuation values are 15.94 dB, 22.73 dB, 28.43 dB, 33.59 dB, and 38.
At 38 GHz, the horizontal polarisation expected attenuation values were 24.25 dB, 34.49 dB, 43.05
dB, 50.79 dB, and 58.03 dB for the same path length. The vertical polarisation anticipated
attenuation values were 66.97 dB, 28.09 dB, 39.91 dB, 49.76 dB, and 58.66 dB, respectively. The
research area's prospective 5G mm-wave network builders will find these results useful.
In order to calculate the 1-minute rain rate cumulative distribution function and forecast rain-
induced attenuation for 5G communication connections for the research region, two years of 1-
minute rainfall data obtained from 2019 to 2020 were employed in this work. At a time exceedance
of 0.01%, the recorded rain rate is 89. mm/hr. ITU-R P530-17 model suggestion for horizontal and
vertical polarisations at 26 GHz and 38 GHz was used to obtain the assessment of rain attenuation
at 0.01% of time exceedance. For a 1 to 5 km channel length, the path attenuation values for
horizontal polarisation at 26 GHz are 19.85 dB, 28.09 dB, 35.28 dB, 41.64 dB, and 47.58 dB, whereas
the values for vertical polarisation are 15.94 dB, 22.73 dB, 28.43 dB, 33.59 dB, and 38.45 dB. At 38
GHz, the horizontal polarisation expected attenuation values were 24.25 dB, 34.49 dB, 43.05 dB,
50.79 dB, and 58.03 dB for the same path length. The vertical polarisation anticipated attenuation
values were 66.97 dB, 28.09 dB, 39.91 dB, 49.76 dB, and 58.66 dB. The findings demonstrate that
the polarisation type, frequency, rain rate, and route length all affect attenuation. The findings from
this study will be beneficial for the effective and efficient design of 5G communication systems in
the research domain.

5. A Survey of Rain Attenuation Prediction Models for Terrestrial Links—Current Research

Challenges and State-of-the-Art

Millimeter-wave (30–300 GHz) frequency is a good option for 5G and future wireless networks, but
radio links at this frequency band are limited by things in the air. Rain is the most important weather
factor that causes the wave to weaken, which is why the fade minimization method (FMT) needs to
be able to predict it. Estimating rain absorption has been done with a lot of different methods.
Different models have different sets of input values and different ways to estimate. This study has
shown that there are a number of ways to get data for rain absorption models. In this study, the
theories for how rain affects land are looked at in detail. As far as we know, there is no study of
models for reducing rain on land. In this piece, the needs of this poll are talked about first, along
with different ways to make a collection. The terrestrial links models are put into groups, and then
qualitative and numeric comparisons are made between these terrestrial rain attenuation models.
Also, a set of methods for evaluating mistake performance is shown. There is also talk about open
research problems and challenges, like how important it is to make a rain reduction model for the
short-range link in the E-band for 5G networks and beyond.
 PS.04/08/2017

6. Deep Learning for Rain Fade Prediction in Satellite Communications

Rain is a big problem for line-of-sight satellite systems, robotic aerial vehicles, high-altitude
platforms, and microwave lines that work on frequency bands like Ka-band or higher. So, rain fade
predicting is important for these systems because it lets them switch between ground ports before
a rain fade event to keep service running smoothly. Empirical, statistical, and fade slope models can
predict rain fade to some extent, but they usually need statistical readings of how rain behaves in
a certain area and can't be used to predict rain fade on a large scale. Also, these models can usually
predict rain fade events in the near future, but they can't predict far into the future. This makes it
harder to handle resources in a responsible way. In this study, a deep learning (DL)-based design is
suggested that can predict future rain fade using data from satellite and radar images as well as
measures of link power. Also, the preparation of data and the design of the architecture have been
described in detail, and many tests have been done. Experiments show that the suggested DL design
does a better job of predicting rain fade in the short and long run than the best machine learning-
based systems on the market today. Also, the results show that radar data with information about
weather conditions is better for short-term predictions, while satellite data with information about
cloud movement is better for long-term predictions. In this study, we suggest a DL-based design for
predicting rain fade in airborne communication systems that work at Ka-band or higher frequencies
(Q-band and V-band), such as satellite systems, UAVs, or HAPs. To do this, we have written down
all of the steps for cleaning and preparing the data, as well as the building design. We have done a
number of tests to figure out how to predict long-term and short-term rain fade and compared our
results to the most cutting-edge ML-based methods. The results show that the suggested DL-based
design does better than state-of-the-art models in terms of accuracy, memory, precision, and f1-
score, especially for long-term forecasts. Also, the tests show that radar data with information
about weather conditions is better for short-term predictions and satellite weather data with
information about cloud movement is better for long-term predictions.

7. A Deep Learning-Based Model for Prediction of RF Wave Attenuation Due to Rain

Every year, there is a big rise in the demand for wireless data bandwidth for communication links,
so mobile providers must be getting ready for a big jump in mobile traffic in the future. This makes
experts and scientists want to learn more about the higher-range bands. The main problem for
these bands is that they are more sensitive to things like rain, dust, clouds, and so on. There are
many scientific and physical methods that can be used to predict rain absorption. Those models
aren't as exact or complicated as they seem. As new technologies like Machine Learning and AI
come out, they can make any system better. These are the kinds of technologies that telecom
experts are looking for in order to make telecom systems better. Artificial Neural Networks (ANN)
are used to build a model for this study. This model was trained with data from the AMSER2 satellite
that was collected in real time. Compared to the study that has already been done, the suggested
model will be more correct. This model method (6G) will help with the building of communication
tools for the next generation.

In this study, a model for machine learning that is based on artificial neural networks is made. In
this study, real-time data from the AMSER 2 satellite is used. From space, scientists saw how much
rain fell over a year. The benefit of the suggested model is that it only needs to know how much
and how often it rains to predict how much the clouds will reduce the light. The suggested model
shows that it is accurate 99.7% of the time, which is pretty good.
Machine learning methods for reducing the effects of rain have their pros and cons. The pros of the
model being proposed are:
 PS.04/08/2017

Improved accuracy: Compared to traditional models, proposed Machine Learning models have
made rain reduction predictions more accurate by learning from large datasets and finding
complicated patterns that are hard for people to see.

Flexibility: The proposed machine learning models can be easily changed to different types of
communication systems and can handle multiple input factors, which can improve the accuracy of
forecasts in different situations.

Speed: The proposed machine learning models can make guesses quickly, which can be important
in real-time uses. Some of the problems with the plan being proposed are:

Data needs: For suggested machine learning models to train well, they need a lot of data, which can
be expensive and take time to get.

Machine learning models that have been offered can be hard to understand, which makes it hard
to understand how they make their predictions. This can be a problem when it's important for
people to be open and answerable.

8. Learning-Assisted Rain Attenuation Prediction Models

The signal-to-noise ratio (SNR) of an earth-space or ground radio link is affected by how much rain
falls. So, with the help of fade minimization techniques (FMTs), it needs to control the power of the
broadcast signals based on how much rain is predicted. But because rain is a chance event, it is hard
to guess how much rain reduction a particular link might experience. There are a lot of scientific,
physical, and mixed models based on nature that can be used to predict the predicted rain
reduction. Also, since the learning-assisted (LA) method was made, many optimisations and
decision-making tasks have become easier. In this work, the LA rain attenuation (LARA) model was
sorted by the factors that were put into it. Also, for a comparison study, each of the accepted
frequency components of LARA models was tallied, and the exact share of each model was found.
The precision and association of input-output factors are shown, which is different from all other
LARA models that are currently available. Also, it sums up and talks about open-study topics and
problems.

METHODOLOGY

In this section, a few methods are carried out to gather information for this thesis. First of
all, the study and information gathered from the literature review are been analyzed. As
proof for this thesis, two methods will be carried out to prove the result.

1. MATLAB Simulink

a. Collect the rain rate data.

b. Data in time series form
c. Add path the data in MATLAB Simulink
d. Collect the result and analyze

2. Deep Learning
 PS.04/08/2017

a. Collect data: dataset includes measurement of rain attenuation and

corresponding 6G performance metrics.
b. Preprocess the data: preprocess the data to ensure its quality and compatibility
with the deep learning model.
c. Design the model: Appropriate deep learning architecture for rain attenuation
prediction. Involve convolution neural networks (CNNs) and recurrent neural
networks (RNNs).
d. Train the model: train using pre-processed data. Involves feeding the input rain
attenuation data into the model and adjusting the parameters to minimize the
prediction error.
e. Evaluate the model: assess the performance of the trained model using the
testing dataset. Measure metrics such as accuracy, precision and evaluation.

OUTCOMES: EXPECTED and/or TO DATE

1. MATLAB Simulink

- Rain Data collected from UTHM

-Collected Rain data should be in time series
-One sample in every DeltaT seconds
 PS.04/08/2017

a. Input Parameters

DeltaT: the interval between two samples of the rain rate data in seconds.
Phi:latitude of the ground station, North (positive) and South (negative), in
degress.
Hs: altitude of the ground station above sea level.
Freq: frequency of the link in GHz (6G)
Taur: E.M carrier wave polarization in degress
Theta: slant path elevation of link in degrees.

Final output: A variable will contain rain attenuation time series in dB and the t
variable will contain the corresponding time axis in seconds.

2. Deep learning

There are few models for rain attenuation in 6G using deep learning. The artificial neural
network is used in this thesis. An artificial neural network is widely used in deep learning
methods that can apply to rain attenuation prediction. This is because it can be trained
using massive datasets related to the parameter as frequency, rainfall rate, path length
and humidity.

CONCLUSIONS (TO DATE)

For the creation of high-speed wireless communication networks, research on rain attenuation in
6G systems is essential. Higher frequency signals are more affected by rain, and it is difficult to
forecast attenuation because of how rain behaves during a single episode. Communication
engineers can create more precise models and construct more durable communication systems
with the use of experimental research and investigations characterising the attenuation of
terahertz radiation in various weather situations. By necessitating closer proximity between
transmission stations, joint communication and sensing in 6G networks can also lessen the effects
of rain attenuation.

*NOTE:
1. Please provide adequate information for a fair evaluation of your research progress thus far.

2. Save as PDF file before uploaded in e-LKP system