=Paper= {{Paper |id=Vol-2845/Paper_13.pdf |storemode=property |title=Development of distributed Multi-Segment Wireless Networks for Determining External Situtions |pdfUrl=https://ceur-ws.org/Vol-2845/Paper_13.pdf |volume=Vol-2845 |authors=Andriy Dudnik,Volodymyr Kvasnikov,Olexander Trush,Tetyana Domkiv |dblpUrl=https://dblp.org/rec/conf/iti2/DudnikKTD20 }} ==Development of distributed Multi-Segment Wireless Networks for Determining External Situtions== https://ceur-ws.org/Vol-2845/Paper_13.pdf

Development of Distributed Multi-Segment Wireless Networks
for Determining External Situations
Andriy Dudnika, Volodymyr Kvasnikovb,Olexander Trusha, and Tetiana Domkivb
a
Taras Shevchenko National University of Kyiv, 60 Volodymyrska Street, Kyiv, 01601, Ukraine
b
National Aviation University, ave. Cosmonaut Komarov 1, 03058, Ukraine

Abstract
The object of the research is the development of a distributed multi-segment wireless
network of the IEEE 802.11 and ZigBee standards at 2.4 GHz, which in this study is
proposed to be used to establish the fact of fire. The main areas of application are conducting
observations for detecting fires, conducting chemical and radiation reconnaissance of objects
and territories. Also, it is equally important to monitor objects that are in continuous
production, in particular in areas with no line of sight. Where the use of video or infrared
systems for detecting emergency situations is not possible for two reasons:
- these systems require more frequent maintenance, therefore, their installation when
deploying continuous production, where there will be no access later, will cause great
inconvenience during maintenance;
- Mechanical obstacles, like trees or shrubs, are insurmountable obstacles to the optical or
infrared beam compared to the radio signal.
The sensors can be made both on a flexible basis, for example, for a more reliable fixing on a
tree, or on a solid one. Such sensors can be placed even from an airplane in places where it is
problematic to place them by ground methods. Placing video or infrared systems for
detecting emergency situations in this way is impossible. They require high-quality
installation work and are placed only from the ground. They require more frequent
maintenance than wireless sensors in a distributed network. This, in turn, greatly complicates
the maintenance of such systems in hard-to-reach places of the natural landscape. Also, the
components of distributed wireless networks are much cheaper. One of the most problematic
places is such phenomena as scattering and absorption during the passage of radio waves
through a forest, as well as other objects of the natural landscape.\In the course of the study,
methods were developed for measuring distance by means of multi-segment distributed
wireless networks using an algorithm for bypassing sensors in indirect visibility zones in
order to localize the fire point. Determined not only the dependence of the radio signal power
with increasing distance, but also on the density of trees in the forest, which affects the
accuracy of localization of the fire source. It has been found that the drop in signal power
between open and mixed areas is almost a third, with the same indicators of increase in
distance. Corrections for these indicators are introduced into the system, which ensures a
decrease in the error in determining the coordinates of the fire source. Thanks to this, it is
possible to operate a distributed network in order to measure the distance and angle between
objects in the radius of which a fire has occurred, in areas with no line of sight. This ensures
sufficient accuracy (error δ≈2 m), which is impossible when using traditional measuring
instruments. It is shown that the influence of the flame on the signal power leads to a drop in
power (- 1 dBm). This change does not reduce the accuracy of determining the position of the
fire, but makes it possible to determine that this is a flame.

Keywords 1
Emergency, signal, radio waves, open space, flame, IEEE 802.11, ZigBee.

IT&I-2020 Information Technology and Interactions, December 02–03, 2020, KNU Taras Shevchenko, Kyiv, Ukraine
EMAIL: a.s.dudnik@gmail.com (A. 1); kvp@nau.edu.ua (A. 2); Trush.viti@gmail.com (A. 3); tdomkiv@gmail.com (A. 4)
ORCID: 0000-0003-1339-7820 (A. 1); 0000-0003-3888-772X (A. 2); 0000-0002-4188-2850 (A. 3); 0000-0002-1625-7754 (A. 4)
©️ 2020 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)

127
1. Introduction
The question of studying the propagation of radio signals in ordinary areas is of great scientific
and technical interest, especially recently. Significant relevance of these works came with the
improvement of wireless network technology of the IEEE 802.11 standard, which uses the
frequencies of 2.4 and 5 GHz. The reason for this is that forest fires have serious consequences for the
environment and the economy, both in Ukraine and around the world.
The use of distributed multi-segment networks of the IEEE 802.11 standard provides advantages
in prompt transmission for the exchange of necessary information between emergency services, their
location relative to the point of occurrence of an emergency situation. This has a significant impact on
the radio, both in the open and in the average (forest area). Radio waves, propagating through trees,
are distorted due to scattering and absorption. Due to the fact that the signal power decreases, this
method of propagation was effective only at short distances. But when using distributed networks
(more than 2 devices in a segment) and connecting the required number of segments, it is possible to
deploy a network of up to tens of kilometers.
Delays in locating an object that is part of a fire detection system cause incorrect data to appear on
the computer network for emergency detection. This shortcoming, in turn, can lead to, for example,
late detection of the source of ignition. Such effects are most likely in environments where there is a
significant amount of radio emission. Among the many standards of computer networks are those
whose data transmission system is based on the use of wireless communication channels. Such
systems are used as means of signal propagation of computer networks for emergency detection
(wireless sensor networks).
The relevance of this work is that in the process of deploying wireless computer networks to detect
emergencies, the ability of the system to work in a mixed area is essential. One of the ways to solve
this problem is to improve the existing methods of "bypassing" the sensors that make up the system.
Delays in sending packets, as well as errors in sending them, are also a significant disadvantage of
existing wireless computer networks. This is also the cause of radio overload and significantly
increases the occurrence of errors in the localization of coordinates, within which there was an
emergency situation, such as fire.
But the most urgent task to be solved in this work is to identify signs based on changes in a
number of parameters that characterize the occurrence of an emergency situation (fire). Also, these
features make it possible to distinguish fire and smoke from other natural and man-made phenomena.

2. Analysis of literature sources and problem statement

Article [1] is devoted to the use of the Internet to control the measuring head, but the processing
and adjustment of the obtained indicators via the Internet remained unresolved. Instead, in [2] the
research of the issues of development of analog means of adjustment of measuring equipment is
considered, but it does not investigate the issue of determining the position of the object. However,
the work [3] is devoted to the elimination of localization errors by means of systems for finding the
coordinates of the object, but for this purpose a cable communication is used.
This shortcoming has been eliminated in [4], which is devoted to a general overview of various
protocols of wireless networks, but only for the purpose of analyzing their bottlenecks, without giving
methods for their elimination. This issue is addressed in [5], which are devoted to positioning
algorithms that can bring positive changes in the process of localization of objects, but they are not
able to identify signal transmission media.
Another area of solving this range of problems includes works [6], which are devoted to current
issues of integration of sensor networks and methods of their improvement. A separate area should
include works [7], which are devoted to methods of determining the position used by satellite
navigation systems. An equally important area related to the long life of equipment is the work [8, 9]
on energy-saving technologies for computer networks to detect emergencies. As a confirmation of the
need for more charge is the work [11], which states that the signal level in the open is higher than in
the forest.

128
An option to overcome the above shortcomings may be work [12]. It presented the results of
experimental studies of the influence of species and structural properties of forest vegetation on the
peculiarities of the propagation of electromagnetic waves in the meter range.
However, wireless computer networks do not work in the meter range. Therefore, the results of
experiments are described in [13], which show a significant difference in the change of the spectra of
pulse signals during propagation in different types of forest vegetation.
This issue is largely addressed in [14], which shows experimental data on the attenuation of radio
waves by the crowns of individual trees. No less relevant is the work [15], which reviewed
electrodynamic models and methods of analysis of radio wave propagation in forests at different
frequencies and distances.
This approach, with a greater focus on the specifics of forest vegetation, is presented in [16, 17]. In
these works, as well as in works [18-23] the data on weakening of power of radio waves at simulation
of a forest fire, and also in laboratory conditions are given, thus metal salts were poured on a burner
flame, emitting burning of leaves in the forest containing meadow.
All this suggests that this study is appropriate. It discusses models and methods for deploying
distributed multi-segment wireless computer networks to measure distance and angle between objects.
These models and methods are adapted to areas of no direct vision, which allows determining the
position at which the emergency occurred (fire.) Also described an experiment, the results of which
reflect the effect of flame and smoke on signal strength, which will identify them among others types
of interference.

3. The purpose and objectives of the study of the effect of flame on signal
strength
The purpose of the study is to develop a distributed multi-segment wireless network with the
improvement of existing technological solutions to identify emergency situations (sources of fire) in
the forest area, which is an area of lack of direct visibility,
To achieve this goal it is necessary:
1. To propose a method of measuring the distance between the objects of wireless computer
(sensor) networks, in order to calculate the position of the emergency situation (ignition
sources);
2. to offer the equation of distribution of an error of calculation of position of objects of
wireless computer (sensor) networks (𝛿𝑛 ), for the purpose of introduction of adjustments
in process of localization of a site of ignition;
3. to investigate the algorithm of "bypassing" the devices of distributed wireless computer
(sensor) networks, which allows to measure the distance to objects in areas of no direct
visibility, which is almost impossible when using conventional measuring equipment;
4. Conduct an experiment to investigate the effect of flame and smoke on the signal strength
of distributed wireless networks, which will reveal the features that distinguish fire and
smoke from other types of interference.

4. Materials and methods for studying the effect of flame on the signal
strength of distributed multi-segment wireless computer networks
Neural network positioning methods are based on measuring signal levels from all nearby access
points or base stations whose coordinates are known. After creating a software or hardware solution
of the neural network, it is necessary to create a mathematical model of positioning and perform
network training.
Many mathematical models are known that can describe the relationship between the distance to
an object and the signal level, and a new case model can be developed.
Therefore, we can conclude that the methods of positioning a mobile station using artificial neural
networks can provide a fairly high accuracy.

129
The proposed method of positioning is based on the values of the RSSI, which is determined as
follows [23]:
𝑅𝑆𝑆𝐼 = −10𝑛 ∙ 𝑙𝑜𝑔𝑑 + 𝐴, (1)
where d is the distance, A is the transmitter power, n is the signal propagation constant, in the free
space n = 2.
The method, based on the nesting of neural networks, requires training of the network on a fairly
large sample, which is obtained on the basis of preliminary analysis. After that, if the training is
successful, the neural network can perform an assessment of the location of the mobile station. The
obtained estimate will be formed as an estimate of the coordinates of the location of the station.
The generalized algorithm of operation of the location system based on the neural network can look
like this [39-43]:
 preliminary analysis of the localization area is carried out;
 the structure of a neural network and algorithm of its training is chosen;
 neural network training is performed on the basis of collected data;
 The trained neural network is tested;
 The adequacy of the received estimations is checked.
Usually as a neural network for spatial localization use multilayer perceptrons with one hidden
layer. The number of input neurons is proportional to the number of antennas in the system, and the
number of output neurons is proportional to the dimension of space. As a neural network learning
algorithm, the backpropagation algorithm is most often used.
The RSSI value from several nearest base stations (at least three) to the desired mobile station
enters the neural network, at two outputs of which, after appropriate processing, the coordinate value
(latitude and longitude) of the mobile station appears.
The proposed neural network (Fig. 1) is a multilayer perceptron - one of the easiest ways to
combine neurons into a network. The input layer consists of k ≤ 6 neurons xi, i = 1… k. The hidden
layer consists of m neurons yj, j = 1… m, and m = 2 (k + p) + 1.
The original layer consists of p = 2 neurons zn, n = 1…p. Input and latent neurons are
interconnected by synapses with weights wij, latent and output neurons are interconnected by synapses
with weights vjn. The inputs of the neurons of the input layer receive signals si, which are RSSI values
from base stations. Each input neuron xi transmits the obtained value, multiplied by the corresponding
weights, to all neurons of the hidden layer.

Figure 1: Neural network

130
The proposed neural network functions as follows. Each latent neuron yj adds the obtained
weighted values of xiwij and calculates the value of yjS by formula (2)
𝑘 (2)
𝑦𝑗𝑆 = − ∑ 𝑥𝑖 ∙ 𝑤𝑖𝑗 + 𝑤0𝑗 ,
𝑖=1
where w0j is the displacement of the latent neuron.
The value of the activation function for each hidden neuron is calculated by the formula (3)
𝑦𝑖 = 𝑓(𝑦𝑗𝑆 ). (3)
In this case, the activation function is a function of the hyperbolic tangent calculated by the
formula (4)
𝑒 𝑎 − 𝑒 −𝑎 (4)
𝑓 (𝑎 ) = .
𝑒 𝑎 + 𝑒 −𝑎
Each original zn neuron adds the obtained values and calculates the values of znS by the formula (5)
𝑘 (5)
𝑧𝑛𝑆 = ∑ 𝑦𝑖 ∙ 𝑣𝑖𝑗 + 𝑣0𝑗 ,
𝑖=1
where v0n is the displacement of the original neuron.
The value of the activation function, which is a function of the hyperbolic tangent, for each source
neuron is calculated by the formula (6)
𝑧𝑛 = 𝑓 (𝑛𝑛𝑠 ). (6)
The output values of the output neurons are the values of the latitude and longitude of the location
of the mobile station, i.e.
𝑧1 = С𝑠ℎ ; 𝑧2 = 𝐶𝑑 .
Both output neurons calculate the error taking into account the reference output value tn by the
formula (7)
𝛿𝑛 = (𝑡𝑛 − 𝑧𝑛 ) ∙ 𝑓`(𝑧𝑛𝑆 ), (7)
weight adjustment: ∆𝑣𝑗𝑛 = ℎ ∙ 𝛿𝑛 ∙ 𝑦𝑗 and offset adjustment: ∆𝑣0𝑛 = ℎ ∙ 𝛿𝑛 , where h – learning rate.
Hidden neurons calculate the total error of the original neurons, which enters the j-hidden neuron
𝑝

𝛿𝑗𝑆 = ∑ 𝛿𝑛 ∙ 𝑣𝑗𝑛 ,
𝑖=1
error value 𝛿𝑗 = 𝛿𝑗𝑆 ∙ 𝑓`(𝑦𝑗𝑆 ); adjustment of weights ∆𝑤𝑖𝑗 = ℎ ∙ 𝛿𝑗 ∙ 𝑥𝑖 and offset adjustment
∆𝑤0𝑗 = ℎ ∙ 𝛿𝑗 .
The values of the weights are updated according to the formulas (8)
𝑤𝑖𝑗 = 𝑤𝑖𝑗 + ∆𝑤𝑖𝑗 ; 𝑣𝑖𝑗 = 𝑣𝑖𝑗 + ∆𝑣𝑖𝑗 (8)
Neural network training continues until the total standard error is minimized
𝑝 𝐿
1
𝐸𝑠 = ∑ ∑(𝑧𝑛𝑞 − 𝑡𝑛𝑞 )2 ,
2
𝑛=1 𝑞=1
where L is the dimension of the training sample.
The network has three classes of devices: FFD-routing devices that use the development of the
authors [1-4, 44] (Full Function Device – a device with a full set of functions – R), coordinating
devices that use developments [1, 4], and additional uses a GPS signal to determine its position
(anchor nodes,) (Coordinators – FFD with additional system resources depending on the complexity –
G) and RFD-end devices that use the development [23] (Reduced Function Device - a device with a
limited set of functions – S).
Multi-segment networks are formed on the basis of distributed, having more than 2 devices, by
connecting to the devices R or G adjacent networks (segments), which can increase their range by
many kilometers. In addition, each R device connects to the same segments with S and G devices, as
in the right and left parts of the figure. (Fig. 2).

131
Figure 2: Sensor network model

In a multi-segment distributed wireless network for monitoring and detection of emergency
situations, it is possible to use the distance between adjacent anchor devices, which is found in areas
with the best signal level, as a reference. The reference distance can be used both within one segment
and within all network segments, correcting for a drop in the radio signal level [23].
In R, it is proposed to use, in order to determine the distance to neighboring nodes, both the radio
signal and the beam of the laser rangefinder, which is aimed at the object on which the neighboring
node is fixed.
The SNDWAY SW-M80 laser rangefinder was used. Its connection to network devices is based on
the methods presented in [1-4], which were previously performed by the authors. This is done in order
to be able to adjust the obtained distance values on the basis of two alternative values, in order to
increase accuracy.
The second purpose of such duplication is the ability to obtain at least one of the indicators, if the
other, due to an emergency situation on the route of the signal (beam), it is physically impossible to
obtain [23].
To conduct an experiment to determine the distance between the objects of a distributed multi-
segment wireless computer (sensor) network, in order to identify the point of occurrence of an
emergency situation (fire), an algorithm for "bypassing" network devices in areas of no direct
visibility [44]:
1: positionsi=ϕ;
2: msgi=nil;
3: if ni ∈ U then {if it is an anchor node};
4: Start Walking();
5: start posTimer.
6: end if
7: posTimer timeout.
8: (xi, yi):=getGpsPosition();
9: Send position(xi, yi) to all nj ∈ Ni.
10: Restart posTimer.
11: msgi=position(xk, yk)such that distk=distanceEstimitation(msgi);
12: positionsi:= positionsi ∪ {( xk, yk, distk)};
13: if size (positionsi)<=3 then { if there are enough positions };
14(xi, yi):= positionComputation(referencesi);
15: end if.
This algorithm was modeled in the MATLAB system, on the example of a network within the
radius of which is a rectangular building (Fig. 3).

132
Figure 3: The result of modeling the deployment of the network in the area of no direct visibility in
the area of 100x100 m

To conduct an experiment to obtain data on the signal power drop, depending on the range, in the
area of indirect visibility, the application Lizard Systems WiFi Scanner is proposed (Fig. 4).

Figure 4: WiFi Scanner application window

133
R devices are shown in red, and Member Node and G are shown in blue.
This application is a very powerful system. It allows you to monitor more than 10 parameters of
wireless computer networks, including the decrease in signal strength with increasing distance.

5. The results of the study of the effect of flame on signal power in multi-
segment distributed wireless networks
The task of direct scientific research of this work is experimental and practical detection of
attenuation of signal power and data transmission in the open field, as well as in the forest at different
distances. Experimental evaluation of signal attenuation was performed in three stages.
The first part of the experiment was to measure the signal strength and bandwidth of the channel in
the open at the stadium, with no obstacles to the signal. Measurements were performed with a
distance change every five meters.
The second part of the experiment was conducted in a deciduous (oak) forest, medium density,
with flat terrain with densely growing shrubs. The average height of the trees was 15 m, the average
diameter of the trunks was 0.3 m, and the bushes were 2 m high.
Measurements were made every ten meters. The transmitter and receiver were located at a height
of 1.2 m from the ground and were optimally directed at each other. The third part of the experiment
was performed when exposed to the signal of the flame of fire, with the distance between the
receiving and transmitting device was 6 m
In this part of the experiment, the forest fire of the underlying surface was simulated. Data
transmission was performed using Wi-Fi IEEE 802.11 technology in the frequency range of 2.4 GHz
and a signal power of 100 mW at an antenna gain of not more than 6 dB, which is the allowable range
of the standard [23]. Measurement of signal strength was performed using a personal computer, WiFi
Scanner software version.
The signal intensity was measured in dBm and displayed on the oscillogram, which is a graph of
the power of the signal coming to the receiver from the time of its arrival (Fig. 4).

Figure 4: Oscillogram of signal power dependence on the time of its arrival

The peak values of the signal power are visible on the oscillogram. The signal was recorded, on
average, 4 minutes on each section of the distance. The results thus obtained were subsequently
averaged.

134
A comparative evaluation of the measurement of the signal power dependence on the distance in
the open space (straight line) and in the forest area (dashed line) was performed. The graph of this
assessment is shown in Fig. 5.

Open space Forest
-40
-45
-50
Power drop (dBm)

-55
-60
-65
-70
-75
-80
-85
-90
-10 10 30 50 70 90
Distance (m)
Figure 5: Graph of signal power dependence on distance in different environments

The signal strength decreases as the distance between the transmitter and receiver increases. In the
open at a distance of 90 meters, the signal decreased by 15 dBm. The amount of information
transmission ranged from 1.85 Mbps to 1.65 Mbps, and as the distance increased, the volume
decreased accordingly.
In the forest, the signal is significantly weakened, this is due to the passage of the signal through
the thick bush in the first ten meters of the experiment and already at 70 meters the signal becomes
much lower, approaching the threshold sensitivity.
This is quite enough, because the maximum distance between neighboring devices of the
distributed network, within one segment, is 50 meters.
When the flame affected the transmitted signal, a decrease in the signal power c - 46.7 dBm to -
47.6 dBm was observed, respectively, a decrease in the power ≈– 1 dBm.
This is primarily due to the fact that in the flame, as a result of a chemical reaction, mobile positive
ions and negative particles - electrons are formed.
Thus, the concentration of charged particles in the flame plasma is 10 12 ions/cm3. These particles
in turn affect the propagation and signal strength.

6. Discussion of the results of studies of the effect of flame on signal power
in multi-segment distributed wireless networks
The disadvantage of this study is that as a result of the experiment, a wireless computer was
deployed in a heterogeneous area. The received signal power is 30% less than in the open.
The advantage is that despite this, the network continues to operate at the minimum sensitivity
limit and measure the distance to neighboring objects with a tolerance of δ≈2 m.
These indicators allow to determine the position of the emergency situation (fire) with sufficient
accuracy. It should also be noted that when simulating a leaf fire, the display of the laser rangefinder
SNDWAY SW-M80 displayed "ERROR".
This error is due to the fact that flames and smoke are environments that do not reflect the laser
beam. In rare cases, if reflected, it reaches the rangefinder receiver with significant distortion that
makes it impossible to determine the distance to the object with sufficient accuracy.

135
Rain and fog are also such environments, but they would cause such errors within the range of the
entire network, not just a single area.
A useful result of this study is the identification of parameters that allow the separation of flames
and smoke from other natural and man-made phenomena that may occur in the forest.
Therefore, it can be concluded that if the movement of network devices in the study area did not
occur, and the rangefinders transmit the signal "ERROR" only in a separate area, which eliminates the
possibility of rain or fog. In this case, if the measurement of the signal power in this area for more
than 2-3 minutes shows a drop of ≈ – 1 dBm, then with a high probability we can conclude the
occurrence of an emergency situation (fire).
After all, if a non-transparent object appears between adjacent devices, the laser rangefinder will
send the distance to this object, not the "ERROR" signal.
This study is a continuation of a series of experiments to determine the linear angular dimensions
between objects of wireless computer networks.
On the other hand, it is the first in terms of studying the effects of flame and smoke on the
parameters of the localization system. In the following researches it is planned to analyze the
dependence of network parameters on different stages of combustion.

7. Conclusions

1. The method of measuring the distance between the objects of distributed wireless
networks is proposed, in order to calculate the position of the emergency situation
(ignition source), which allows the deployment of the network and assess the position of
its objects. This allows you to monitor the occurrence of fires, by means of multi-segment
distributed wireless networks, expanding the coverage area, within each segment, up to
100x100m.
2. The equation of the distribution of the error in calculating the position of the objects of
wireless computer (sensor) networks, in order to make adjustments to the process of
localization of the fire area. This equation allows us to estimate the occurrence of an
emergency situation (ignition source) with an error of δ≈2 m, which is sufficient accuracy.
3. The algorithm of "bypassing" devices of wireless networks is investigated, which allows
to measure the distance in the zones of absence of direct visibility, which allows
determining the position of the neighboring object when the signal power drops by 30%.
This result is almost impossible when using conventional measuring techniques.
4. An experiment was performed which showed that flame and smoke are such environments
that cause a signal of error of the laser rangefinder, cause a drop in radio signal power ≈ –
1 dBm. It is substantiated why other environments and objects, such as rain, fog or opaque
objects, cannot cause such indicators in the aggregate.

8. References
[1] Sharma, G., & Rajesh, A. (2018). Localization in Wireless Sensor Networks Using TLBO. i-
Manager's Journal on Wireless Communication Networks, 7(3), 32.
[2] Kannadasan, K., Edla, D. R., Kongara, M. C., & Kuppili, V. (2019). M-Curves path planning
model for mobile anchor node and localization of sensor nodes using Dolphin Swarm Algorithm.
Wireless Networks, 1-15.
[3] Ali, M. F., & Shah, M. A. (2018, September). Adaptive Transmission Power-Geographical and
Energy Aware Routing Algorithm for Wireless Sensor Networks. In 2018 24th International
Conference on Automation and Computing (ICAC) (pp. 1-5). IEEE.
[4] Lin, C. H., Chen, L. H., Wu, H. K., Jin, M. H., Chen, G. H., Gomez, J. L. G., & Chou, C. F.
(2019). An indoor positioning algorithm based on fingerprint and mobility prediction in RSS
fluctuation-prone WLANs. IEEE Transactions on Systems, Man, and Cybernetics: Systems.
[5] Liu, C., Wang, X., Luo, H., Liu, Y., & Guo, Z. (2019). VA: Virtual Node Assisted Localization
Algorithm for Underwater Acoustic Sensor Networks. IEEE Access, 7, 86717-86729.

136
[6] Bae, H. J., & Choi, L. (2019, May). Large-Scale Indoor Positioning using Geomagnetic Field
with Deep Neural Networks. In ICC 2019-2019 IEEE International Conference on
Communications (ICC) (pp. 1-6). IEEE.
[7] Mukherjee, S., Amin, R., & Biswas, G. P. (2019). Design of routing protocol for multi-sink
based wireless sensor networks. Wireless Networks, 1-17.
[8] Yang, J., Cai, Y., Tang, D., & Liu, Z. (2019). A Novel Centralized Range-Free Static Node
Localization Algorithm with Memetic Algorithm and Lévy Flight. Sensors, 19(14), 3242.
[9] Chaudhary, N., Alves, L. N., & Ghassemblooy, Z. (2019, April). Feasibility Study of Reverse
Trilateration Strategy with a Single Tx for VLP. In 2019 2nd West Asian Colloquium on Optical
Wireless Communications (WACOWC) (pp. 121-126). IEEE.
[10] Mohammadzadeh, P., Fard, S. M., Azarnia, G., & Tinati, M. A. (2019, April). Location
Estimation of Sensor nodes in 3-d Wireless Sensor Networks Based on Multidimensional
Support Vector Regression. In 2019 27th Iranian Conference on Electrical Engineering (ICEE)
(pp. 1725-1729). IEEE.
[11] Gui, L., Zhang, X., Ding, Q., Shu, F., & Wei, A. (2017). Reference anchor selection and global
optimized solution for dv-hop localization in wireless sensor networks. Wireless Personal
Communications, 96(4), 5995-6005.
[12] Rai, S., & Varma, S. (2017). Localization in wireless sensor networks using rigid graphs: A
review. Wireless Personal Communications, 96(3), 4467-4484.
[13] Giri, A., Dutta, S., & Neogy, S. (2020). Fuzzy Logic-Based Range-Free Localization for
Wireless Sensor Networks in Agriculture. Advanced Computing and Systems for Security, 3-12.
[14] Aslan, Y. E., Korpeoglu, I., & Ulusoy, Ö. (2012). A framework for use of wireless sensor
networks in forest fire detection and monitoring. Computers, Environment and Urban Systems,
36(6), 614-625.
[15] Giri, A., Dutta, S., & Neogy, S. (2020). Fuzzy Logic-Based Range-Free Localization for
Wireless Sensor Networks in Agriculture. Advanced Computing and Systems for Security, 3-12.
[16] Liang, J., Yu, X., Liu, X., Mao, C., & Ren, J. (2019). Target Detection, Localization, and
Tracking in Wireless Sensor Networks. Mission-Oriented Sensor Networks and Systems: Art and
Science, 309-361.
[17] Y. Kravchenko, O. Starkova, K. Herasymenko, A. Kharchenko, “Peculiarities of the IPv6
implementation in Ukraine”, 4th International Scientific-Practical Conference Problems of
Infocommunications Science and Technology, PIC S and T 2017 – Proceedings. pp.363–369.
[18] S. Korotin, Y. Kravchenko, O. Starkova, K. Herasymenko, R. Mykolaichuk “Analytical
determination of the parameters of the self-tuning circuit of the traffic control system on the limit
of vibrational stability”, International Scientific-Practical Conference Problems of
Infocommunications Science and Technology, PIC S&T`2019 – Proceedings, pp. 471–476.
[19] M. Rakushev, O. Permiakov, S. Tarasenko, S. Kovbasiuk, Y. Kravchenko, O. Lavrinchuk
“Numerical Method of Integration on the Basis of Multidimensional Differential-Taylor
Transformations”, International Scientific-Practical Conference Problems of
Infocommunications Science and Technology, PIC S&T`2019, Proceedings. pp.675–678.
[20] O. Pysarchuk, A. Gizun, A. Dudnik, V. Griga, T. Domkiv, S. Gnatyuk Bifurcation Prediction
Method for the Emergence and Development Dynamics of Information Conflicts in Cybernetic
Space. 1st International Workshop on Cyber Hygiene & Conflict Management in Global
Information Networks. Kyiv. Ukraine 2019 – Procedings. pp.692-709.
[21] Rokochinskiy, A., Volk, P., Kuzmych, L., Turcheniuk, V., Volk, L.,& Dudnik, A. (2019,
December). Mathematical model of meteorological software for systematic flood control in the
carpathian region. In 2019 International Conference on Advanced Trends in Information Theory
(ATIT) (pp. 143-148). IEEE.
[22] Y.Kravchenko, O. Leshchenko, A. Trush , K. Dukhnovska, O. Kovtun. Synergetic Approach To
The Study Of Control Systems. Scientific Journal of Astana IT University. 2020. №2, pp. 44-52
[23] A. Dudnik, L. Kuzmych, O. Trush, T. Domkiv, O. Leshchenko and V. Vyshnivskyi, "Smart
Home Technology Network Construction Method and Device Interaction organization Concept,"
2020 IEEE 2nd International Conference on System Analysis & Intelligent Computing (SAIC),
Kyiv, Ukraine, 2020, pp. 1-6, doi: 10.1109/SAIC51296.2020.9239220.

137