=Paper=
{{Paper
|id=Vol-2588/paper48
|storemode=property
|title=Multifractal Properties of Traffic Generator Based on Markov Chains
|pdfUrl=https://ceur-ws.org/Vol-2588/paper48.pdf
|volume=Vol-2588
|authors=Volodymyr Simakhin,Serhiy Bondar,Hanna Drieieva,Oleksandr Kovalenko,Oleksandr Drieiev,Mereke Zhumadilova
|dblpUrl=https://dblp.org/rec/conf/cmigin/SimakhinBDKDZ19
}}
==Multifractal Properties of Traffic Generator Based on Markov Chains==
https://ceur-ws.org/Vol-2588/paper48.pdf
Multifractal Properties of Traffic Generator Based on
Markov Chains
Volodymyr Simakhin 1 [0000-0003-4497-0925], Serhiy Bondar 1 [0000-0003-4140-7985],
Hanna Drieieva2[0000-0002-8557-3443], Oleksandr Kovalenko 2 [0000-0001-9297-0650],
Oleksandr Drieiev 2 [0000-0001-6951-2002] and Mereke Zhumadilova [0000-0001-7998-6711]
1
International Research and Training Center for Information Technologies and Systems,
Ukraine, Kyiv
2
Central Ukrainian National Technical University, Ukraine, Kropyvnytskyi
3
Yessenov University, Aktau, Kazakhstan
sima@irtc.org.ua
Abstract. The problems of determining the fractal dimension of a time series
obtained using a self-similar traffic generator based on Markov chains with a
controlled fractal dimension are stated. Based on numerical experiments to de-
termine the fractal dimension of the generated numerical sequences, statistically
significant changes are shown at different scales. The insufficient development
of high-performance algorithms and methods for generating self-similar numer-
ical sequences for procedure of traffic simulation in telecommunication systems
and networks is indicated. The directions of further research on the manage-
ment of the multifractality in generators based on Markov chains are proposed .
Keywords: Traffic Modeling, Self-Similar Traffic, Multifractal.
1 Introduction
Mathematical models of a self-similar time series are used to describe telecommuni-
cation processes. In the graphs showing the load of the computer’s network channel,
self-similarity is determined by the presence of outliers, the amount of which exceeds
the classical statistical theory predictions (see Fig. 1). The horizontal axis shows time
in arbitrary units, and the vertical shows network load to the maximum throughput
ratio.
In most cases, analytical expressions for self-similar traffic predicted QoS parame-
ters cannot be formed, or such transformations can be formed for individual specific
situations, therefore, analytical calculations are inadvisable. For this reason, to deter-
mine the main indicators of QoS, such as jitter, lateness, average number of failures
and others, the simulation with self-similar traffic generators is used. This leads to
reducing and simplifying the number of calculations for generators of self-similar
traffic with controlled fractal properties, which would allow numerical sequences to
Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attrib-
ution 4.0 International (CC BY 4.0) CMiGIN-2019: International Workshop on Conflict Management in
Global Information Networks.
have properties as close as possible to the properties of the telecommunication net-
work’s real traffic, which is being studied.
Fig. 1. Self-similar traffic example [3]
Given the relevance of managing the generated traffic’s fractal properties, this paper
is devoted to determining the dependence of the traffic model’s fractal properties on
the used scale.
2 Background and related work
The analysis of research and publications on the subject revealed the following. In [1],
general methods of fractal and multifractal analysis of time series are considered.
Methods for determining the main indicators of numerical sequences usage for traffic
analysis in telecommunication systems and networks are described. In [4], basic defi-
nitions and concepts of the fractal measurement theory and fractal analysis are also
formulated.
Importance of multifractal analysis for information exchange processes in comput-
er networks is described in [2]. An analysis of the Internet traffic, which was collected
for more than fourteen years, is shown. The development of a global network, which
has changes in fractal properties at all levels of scaling, is described by the authors.
The article contains proofs of a different nature self-similarity existence on separate
time scales.
In [3], the main attention is devoted to the usage of a trained neural network to au-
tomate the classification of traffic according to its fractal and multifractal properties.
Presented in the paper results are successfully used to determine DDoS attacks. This
proves the existence of difference in traffic’s multifractal metrics for different types
of data. The idea of a significant impact on the fractality of traffic is confirmed in [5],
where information on the successful use of fractal analysis to identify flows of P2P
and gaming traffic, transmission information into the clouds services, scanning of
ports or botnet actions is also given.
In [6], multifractal modeling of numerical sequences is used to restore lost frag-
ments of time series. Multifractal interpolation gives better results than random filling
and classical interpolation methods. In [8] fractal interpolation is used to restore traf-
fic with known fractal properties on different time scales.
The results obtained in [7] show that the modeling of complex networks becomes
possible when hierarchical self-similarity is taken into account in the adjacency ma-
trix. In fact, this approach makes it possible to classify large networks and to carry out
modeling. The theory is also applicable for broader content networks, for example, to
hierarchical relationships between structural units of different orders or scaling.
Subsequent works [8-17] are related to the modeling of information exchange pro-
cesses in computer networks aiming to recover lost data, to the simulate telecommu-
nications network at different scales and with various data types, to analyze network
traffic for several applications. All these tasks require a simulated source of multifrac-
tal traffic with controlled properties. The number of publications [17-23] describing
simulating the multifractal traffic emphasize the relevance of the development and
improving both the accuracy of reproducing given properties and reducing the compu-
tational complexity of multifractal traffic generators. Developing of such methods and
algorithms will make it possible to increase the speed or potential complexity of pro-
cess modeling without using systems that are more expensive in telecommunication
networks [24-26].
3 Problem statement
The main task of the paper is to analyze the properties of a binary numerical series,
which is obtained using a generator based on a graph model of states and the transi-
tions’ probability between them. Such a traffic generator is a typical Markov chains
based generator (see Fig. 2).
Fig. 2. Fractal Traffic Generator Model
The model contains a number of states that set the initial generated value. The next
value is obtained by random transitions, where 𝜆0 and 𝜆1 are the probabilities of
changing the state to the next time quantum, and the probabilities 1 − 𝜆0 and 1 −
𝜆1 define chance of staying in the current state. The specified generator does not use
heavy-tail distributions and does not require complex calculations. Any standard veri-
fied pseudorandom number generators in the interval [0; 1) with uniform distribu-
tioncan be used. Reduced number of operations for calculation is a significant ad-
vantage for practical application in simulation modeling systems of telecommunica-
tion networks and allows saving computation time for traffic formation [27-30].
The fractal dimension of a numerical series can be described by various metrics,
which can lead to getting different misleading values. Using a separate metrics on the
model and real data also allows comparing their fractal properties. Therefore, for a
generator model that is shown in Fig. 2, a separate numerical series metric is con-
structed. The coverage width is considered unitary if all 𝑛 consecutive values are not
equal, and zero if all 𝑛 consecutive values are equal to 1 or 0. It is easy to obtain an
analytical form for calculating the probability of encountering a zero-span segment
with 𝑛 elements, which is equal:𝜆1 (1 − 𝜆0 )𝑛 + 𝜆0 (1 − 𝜆1 )𝑛 .
Based on statistical modeling of generator’s operation, the expected value of the
random walk’s magnitude with 𝑛 steps is obtained, which allows determining the
fractal dimension of the series analytically. Moreover, the fractal dimension depends
on the resulting numerical sequence length 𝑛as given by (1).
𝜆 (1−𝜆0 )𝑛 ln(1−𝜆0 )+𝜆0 (1−𝜆1 )𝑛 ln(1−𝜆1 )
𝑑(𝑛, 𝜆0 , 𝜆1 ) = 2 + 1 (1)
𝜆0 +𝜆1 −𝜆1 (1−𝜆0 )𝑛−𝜆0 (1−𝜆1 )𝑛
For the transition probability values 𝜆0 = 𝜆1 = 0.5, which corresponds to the clas-
sical random process, a graph of the fractal dimension 𝐷dependence on the generated
series number of elements 𝑛 obtained by formula (1) is shown in Fig. 3.
The presence of different values of fractal dimension in mathematical objects at
different scaling levels is called multifractality. The presence of multifractality in the
traffic of computer networks is shown in [3].
Fig.3. Theoretically determined fractal dimension 𝐷(𝑛) with the probability of state
change𝜆0 = 𝜆1 = 0.5.
Additional research is needed to use the time series fractal analysis methods based on
the Minkowski–Bouligand dimension or R/S analysis.
4 Methods of calculating fractal dimension
The definition in Minkowski's interpretation can be used to determine the fractal di-
mension [4], as shown by (2).
ln(𝑁)
𝐷 = lim𝑠→0 (2)
− ln(ε)
where 𝜀 – size or diameter of the subset, that is covering the set which dimension is
being determined;
𝑁 – the minimum number of subsets𝜀, that should be used to cover the whole set,
which dimension is being determined.
It is not possible to directly apply (2) to a discrete set because a minimum ε exists.
Therefore, fixed values of ε should be used.
Let coverings with sizes 𝜀 and 𝑘𝜀be used to cover the main set (𝑘 - for a discrete
system is a given natural number; 𝜀 is an unknown coefficient characterizing the se-
lected discrete system). Then, according to the selected subsets, the number of such
subsets to cover is𝑁(𝜀)and 𝑁(𝑘𝜀). The method for calculating the metric 𝑁(𝑘𝜀) is
quite arbitrary, and is selected according to the values studied in a particular process.
For a discrete system, if the definition of the limit is rejected,(3) is obtained from
equation (2).
−𝐷 ln(𝑘) − 𝐶 = ln(𝑛(𝑘𝜀)), 𝐶 = 𝐷 ln(𝜀) (3)
where ln(𝑘), ln(𝑁(𝑘𝜀)) are calculated for several values of 𝑘; 𝐷and 𝐶 are obtained as
a result of applying linear regression to the points (ln(𝑘) , ln(𝑁(𝑘𝜀))).
Given the randomness of the numerical series, 𝑁(𝑘𝜀)is also a random variable,
with a certain expected value and dispersion. Hence, the obtained points
(ln(𝑘) , ln(𝑁(𝑘𝜀)))will approach the line, but not necessarily belong to it.
In order to reduce additional calculations and transformations, the process of gen-
erating “0”is changed to generate “-1”.As a result, for 𝜆0 = 𝜆1 , the expected value of
the generated sequence becomes 0 and the standard deviation equals 1. In addition,
the created sequence {𝑎1 , 𝑎2 , … , 𝑎𝑛 } is replaced by a cumulative series corresponding
to a random walk {𝑏1 , 𝑏2 , … , 𝑏𝑛 |𝑏𝑖 = 𝑎1 + 𝑎2 + ⋯ + 𝑎𝑖 }. For such series, the covering
width 𝑁(𝑘𝜀) for a segment with length of 𝑘 discrete elements can be calculated by
(4):
𝑛/𝑘−1
𝑁(𝑘𝜀) = ∑𝑖=0 (max(𝑏𝑖∙𝑘+1 … 𝑏𝑖∙(𝑘+1) ) − min(𝑏𝑖∙𝑘+1 … 𝑏𝑖∙(𝑘+1) )) (4)
We propose using the power of two as 𝑛, because then 𝑘 can also be doubled when
constructing reference points for linear approximation. The indicated process for
𝑘 = 8,16,32,64,128 made it possible to construct five points for which a linear ap-
proximation determines the Hurst exponent of the test sequence at 0.53 as the inclina-
tion angle of the straight line (see Fig. 4):
Fig. 4. Diagram for Hurst exponent𝐻 = 0.53 with corresponding fractal dimension 𝐷 = 1.47
and probabilities to change state𝜆0 = 𝜆1 = 0.5.
Theoretical calculations for a random sequence shows that the values should be
𝐻 = 0.5with𝐷 = 1.5. This is close to the obtained results.
Classical theory states that for curves on the plane the covering width 𝑁(𝑘𝜀)
should be expressed by the number of squares with sidesize equals𝑘𝜀, which cover the
whole random walk curve. However, the adopted measure (4) is asymptotically equal
to the Minkowski–Bouligand dimension. Moreover, with the same values being cho-
sen for𝑘, the Hurst exponent matches with the one obtained using R/S analysis, for
which the relation to the fractal dimension is already known.
5 Impact analysis of scale selection to the fractal
dimension
For the purpose of experimental confirmation eleven generated sequences containing
1024 samples (“-1” and “1”, as described in previous paragraph) were used to show
the dependence of the numerical series’ fractal dimension on the selected scale. For
each of those, the Hurst exponent was calculated forsets of 𝑘
ues:{1024, 512, 256, 128, 64}; {512, . . . , 32}; {256, . . . , 16}; {128, . . . , 8}. The results
of these experiments are shown in Tables 1-3.
An independent measurement of the Hurst exponent on an appropriate scale is
formed for each of the generation modes, which allows evaluating the statistical pro-
cessing of the result. The mean of the Hurst exponent and the standard deviation 𝜎 of
that mean, which is inversely proportional to the root of the sample length, are found
for that result. Based on the standard deviation, a confidence interval of 99% reliabil-
ity is calculated, which is provided for deviations from the mean ±3𝜎. The bounda-
ries of the calculated reliability interval are respectively shown in the table columns
−3𝜎and +3𝜎.
The results from all three tables indicate a change in the Hurst exponent while
shifting the scale, which confirms the assumption made in the previous paragraph
based on the analytical formula (1) for a simplified metric for calculating fractal di-
mension.
Table 1. Hurst exponents for generating a sequence with 𝜆0 = 𝜆1 = 0.95
k 1 2 3 4 5 6 7 8 9 10 11 Mean -3σ +3σ
1024..64 0.38 0.23 0.48 0.44 0.34 0.29 0.42 0.46 0.40 0.26 0.40 0.37 0.30 0.45
512..32 0.40 0.33 0.52 0.50 0.34 0.38 0.45 0.44 0.41 0.31 0.48 0.41 0.35 0.48
256..16 0.42 0.35 0.53 0.48 0.38 0.41 0.41 0.45 0.45 0.39 0.47 0.43 0.38 0.48
128..8 0.42 0.35 0.46 0.40 0.40 0.39 0.41 0.42 0.43 0.41 0.43 0.41 0.39 0.44
Table 1 shows that, with a high probability of changing the previous state value to the
opposite, traffic modulation occurs almost constantly, which leads to a significantly
lower probability of outliers. This is reflected in the Hurst exponent values, which
differs from 0.37 to 0.41 for various scales. Fig. 5 shows the signal generated with
such parameters.
As a result of random walk based on the obtained sequence, the distance from the
beginning of the movement will change much slower, because for each step to the
side there will be a significantly higher probability of getting the opposite direction in
the next time quantum.
Fig. 5.Histogram of the signal obtained by generator with 𝜆0 = 𝜆1 = 0.95.
The following Table 2 contains results of a numerical experiment with the generator
being set to keep or change the current state with an equal probability 𝜆0 = 𝜆1 =
0.50. In this mode the generator must correspond to the classical random process with
the Hurst exponent𝐻 = 0.5.
Table 2.Hurst exponents for generating a sequence with 𝜆0 = 𝜆1 = 0.50
k 1 2 3 4 5 6 7 8 9 10 11 Mean -3σ +3σ
1024..64 0.41 0.46 0.39 0.41 0.50 0.37 0.40 0.54 0.35 0.41 0.42 0.42 0.37 0.47
512..32 0.48 0.48 0.53 0.47 0.54 0.48 0.45 0.60 0.47 0.52 0.46 0.50 0.46 0.54
256..16 0.52 0.55 0.58 0.52 0.52 0.55 0.52 0.56 0.56 0.59 0.55 0.55 0.52 0.57
128..8 0.55 0.59 0.61 0.57 0.57 0.57 0.56 0.59 0.60 0.66 0.59 0.59 0.56 0.61
It is noted from Table 2 that four times change of the parameter 𝑘gives estimated
values of the Hurst exponent with a reliability of 99%. Which is equal to the numeri-
cal series on 1024 samples giving lower values for the Hurst exponent than on 256
samples with a reliability of 99%. This indicates the difference in the fractal proper-
ties of the numerical sequence at different scales. Using the fact, the appropriate
queue length of the service system for which the Hurst exponent will have a value
close to 0.5can be chosen and the theory of random Poisson flow can be used without
taking into account self-similarity using classical statistics to determine the character-
istics of such a service system.
Fig. 6 shows a fragment of the obtained sequence with the generation parameters
𝜆0 = 𝜆1 = 0.50.
Fig. 6. Histogram of the signal obtained by generator with 𝜆0 = 𝜆1 = 0.50.
The results of the last experiment with the state change probabilities𝜆0 = 𝜆1 =
0.05are shown in Table 3. From the obtained results, it can be concluded that the
numerical series has low probability of changing the trends.
Table 3.Hurst exponents for generating a sequence with 𝜆0 = 𝜆1 = 0.05
k 1 2 3 4 5 6 7 8 9 10 11 Mean -3σ +3σ
1024..64 0.48 0.58 0.45 0.54 0.61 0.58 0.51 0.37 0.64 0.39 0.64 0.53 0.44 0.61
512..32 0.52 0.63 0.56 0.60 0.66 0.61 0.58 0.49 0.69 0.45 0.69 0.59 0.52 0.66
256..16 0.66 0.77 0.72 0.77 0.67 0.69 0.62 0.65 0.79 0.58 0.76 0.70 0.64 0.76
128..8 0.78 0.83 0.83 0.84 0.73 0.82 0.73 0.79 0.86 0.73 0.80 0.80 0.75 0.84
The numerical sequence is persistent and is able to store trends for some time. How-
ever, at the same time, the observed mean value of the Hurst exponent is close to
𝐻 = 0.5 at intervals of 1024 samples. This means that at large distance, the range of
the cumulative series does not differ from the classical random sequence, where the
difference between the maximum and minimum values increases on average in pro-
portion to the root of the number of steps taken.
At short distances, the Hurst exponent is very different, and reaches mean value
of 𝐻 = 0.8. In accordance with this, there is a scale for which self-similar traffic, as in
the previous case, will have the properties of a classical random process. A fragment
of the obtained signal is shown in Fig. 7.
Fig. 7. Histogram of the signal obtained by generator with 𝜆0 = 𝜆1 = 0.05.
6 Conclusion and further work
This paper presented self-similar traffic generators using Markov chains, which differ
from their analogues in lower computational power requirements for simulation sys-
tems, which makes it possible to increase the performance of information movement
modeling in telecommunication systems and computer networks. Further develop-
ment and study of such systems is relevant.
Based on the simplified metric 𝑁(𝑘𝜀), an analytical expression is formed for calcu-
lating the fractal dimension of the generated binary numerical series based on the
Markov chain. The dependence of the fractal dimension on the length of the interval
on which it is calculated is noted and the assumption is made that the multifractality
properties are repeated on classical metrics, such as numerical dimensions based on
R/S analysis or Minkowski-Bouligand dimension.
In order to verify the assumption, an experiment part was carried out which con-
firmed the assumption of a multifractal numerical sequence obtained by generators on
Markov chains with a reliability over than 99%.
Potential future work includes developing methods for controlling multifractality
parameters, or opposite task of eliminating the appearance of multifractality.
References
1. Jan W. Kantelhardt, Fractal and Multifractal Time Series Institute of Physics, Martin-
Luther-University Halle-Wittenberg, 06099 Halle, Germany April 4, 2008 42 p. URL:
https://arxiv.org/pdf/0804.0747
2. Fontugne, Romainand Abry, Patrice and Fukuda, Akiraand Veitch, Darry land Cho, Ken-
jiroand Borgnat, Pierreand Wendt, Herwig Scaling in Internet Traffic: a 14 year and 3 day
long itudinal study, with multiscale analyses and random projections. (2017) IEEE/ACM
Transactionson Networking journal, 25 (4). 2152-2165. ISSN 1063-6692. URL:
https://ieeexplore.ieee.org/document/7878657
3. Lyudmyla Kirichenko, Tamara Radivilova, Vitalii Bulakh Machine Learning in Classifica-
tion Time Series with Fractal Properties. December 2018. URL:
https://www.researchgate.net/publication/329973801_Machine_Learning_in_Classificatio
n_Time_Series_with_Fractal_Properties
4. Aleksandrov P.S., Pasynkov B.A. Introduction to the theory of dimension. Moscow: Sci-
ence, 1973. 576р.
5. Mackenzie Haffey; Martin Arlitt; Carey Williamson, Modeling, Analysis, and Characteri-
zation of Periodic Trafficon a Campus Edge Network. 2018 IEEE 26th International Sym-
posiumon Modeling, Analysis, and Simulation of Computer and Telecommunication Sys-
tems (MASCOTS), pр. 170- 182, 2018.
6. Lai Simin, Wan Li, Zeng Xiangjian. Comparative Analysis of Multi-fractal Data Missing
Processing Methods. Applied and Computational Mathematics. Vol. 8, No. 2, 2019,
pp. 44-49. doi: 10.11648/j.acm.20190802.14.
7. Mahdi Barat Zadeh Joveini, Javad Sadri аnd Hoda Alavi Khoushhal. Fractal Modeling of
Big Data Networks Conference: International Conference on Pattern Recognition and Arti-
ficial Intelligence (ICPRAI 2018) At: Center for Pattern Recognition and Machine Intelli-
gence (CENPARMI), Concordia University, Montreal, Canada, pp. 1-4, 2018.
8. D. Jiang, L. Huoand Y. LiFine-granularity inference and estimation network traffic for
SDN. PLoS ONE 13(5), 2018. Doi.org/10.1371/journal.pone.0194302
9. K. Xie, C. Peng, X. Wang, G. Xieand J. WenAccurate recovery of internet traffic data un-
der dynamic measurements, inProc. of INFOCOM’17, pp. 1–9, 2017.
10. C. Wang, S. T. Maguluri, and T. Javidi Heavy traffic queuelength behavior in switches
with reconfiguration delay, inProc. of INFOCOM’17, pp. 1–9, 2017.
11. G. Xie, K. Xie, J. HuangWang X, Chen Y andWen J. Fastlow-rank matrix approximation
with locality sensitive hashing for quick anomaly detection, in Proc. of INFOCOM’17,
pp. 1–9, 2017.
12. T. M. Tatamikova and O.I. Kutuzov, “Evaluation and comparison of classical and fractal
queuing systems”, XV International Symposium Problems of Redundancy in Information
and Control Systems, pp.155 - 157, 2016.
13. Michał Czarkowski, Sylwester Kaczmarek and Maciej Wolff, “Influenceof Self –Similar
Traffic Type on Performance of QoS Routing Algorithms’, INTL Journal of electronics
and telecommunications, vol. 62, no. 1, pp. 81-87, 2016
14. Lakhmi Priya Das, Sanjay Kumar Patra and Sarojananda Mishra, “Impact of hurst parame-
ter value in self-similarity behaviour of network traffic”, International Journal of Re-
searchin Computerand Communication Technology, Vol 5, No 12, pp.631-633, 2016.
15. K.V. Ushanev, "Imagination models of the mass service system type Pa / M / 1, H2 / M / 1
and study in gon their basis the quality of service of traffic with complex structure", Con-
trol systems, communication and security. №4, p.217-251, 2015.
16. G.A. Kuchuk, O.O. Mozhayev and O.V. Vorobeyov. The method of prediction of fractal
traffic. Radioand Computer Systems, No. 6, pp. 181–188, 2006. http://nbuv.gov.ua/
UJRN/recs_2006_6_34
17. G. A. Kuchuk, O. O. Mozhayev, and O. V. Vorobeyov. Traffic prediction for congestion
management integrated telecommunications network. Radio-electronic and computer sys-
tems, № 8, pp. 261-271, 2007. http://nbuv.gov.ua/UJRN/recs_2007_8_48
18. G.A. Kuchuk, O.O. Mozhayev, and O.V. Vorobeyov. Analiz samopodіbnogo traffic. Aer-
ospace and technology. No. 9, p. 173–180, 2006. http:
//nbuv.gov.ua/UJRN/aktit_2006_9_35
19. Smirnov A.A., Kuznetsov A.A., Danilenko D.A., Berezovsky A., The statistical analysis
of a network traffic for the intrusion detection and prevention systems, Telecommunica-
tions and Radio Engineering, Vol 74, Issue 1, Begel House Inc, 2015, pp. 61-78.
20. Smirnov, O., Kuznetsov, A., Kiian, A., Zamula, A., Rudenko, S., Hryhorenko, V., Vari-
ance Analysis of Networks Traffic for Intrusion Detection in Smart Grids, 2019 IEEE 6th
International Conference On Energy Smart Systems (2019 IEEE ESS), Kyiv, Ukraine
April 17-19, 2019, pp. 353-358.
21. Smirnov, O., Kuznetsov, A., Kavun, S., Babenko, B., Nakisko, O., Kuznetsova, K., Mal-
ware Correlation Monitoring in Computer Networks of Promising Smart Grids, 2019 IEEE
6th International Conference On Energy Smart Systems (2019 IEEE ESS), Kyiv, Ukraine
April 17-19, 2019, pp. 347-352
22. A.A. Kovalenko, G.A. Kuchuk and A.A. Mozhaev. Construction of exponential time
scales in the analysis of multiservice network queues. Radio and Computer Systems. No.
7, p. 257–262, 2010, http://nbuv.gov.ua/UJRN/recs_2010_7_52
23. E.V. Dobrovolsky and O.L Nechyporuk, Modeling of Network Traffic Using Context
Methods, Scientific Papers ONAS them. O.S. Popova, No. 1, pp.24-32, 2005.
24. Odarchenko R., Abakumova A., Polihenko O., Gnatyuk S. Traffic offload improved meth-
od for 4G/5G mobile network operator, Proceedings of 14th International Conference on
Advanced Trends in Radioelectronics, Telecommunications and Computer Engineering
(TCSET-2018), pp. 1051-1054, 2018.
25. M. Zaliskyi, R. Odarchenko, S. Gnatyuk, Yu. Petrova. A.Chaplits, Method of traffic moni-
toring for DDoS attacks detection in e-health systems and networks. CEUR Workshop
Proceedings, Vol. 2255, pp. 193-204, 2018.
26. Al-Azzeh J.S., Al Hadidi M., Odarchenko R., Gnatyuk S., Shevchuk Z., Hu Z. Analysis of
self-similar traffic models in computer networks, International Review on Modelling and
Simulations, № 10(5), pp. 328-336, 2017.
27. Fedushko S., Benova E. Semantic analysis for information and communication threats de-
tection of online service users. The 10th International Conference on Emerging Ubiquitous
Systems and Pervasive Networks (EUSPN 2019) November 4-7, 2019, Coimbra, Portugal.
Procedia Computer Science, Volume 160, 2019, Pages 254-
259. https://doi.org/10.1016/j.procs.2019.09.465
28. Fedushko S., Davidekova M. Analytical service for processing behavioral, psychological
and communicative features in the online communication. The International Workshop on
Digitalization and Servitization within Factory-Free Economy (D&SwFFE 2019) Novem-
ber 4-7, 2019, Coimbra, Portugal. Procedia Computer Science. Volume 160, 2019, Pages
509-514. https://doi.org/10.1016/j.procs.2019.11.056
29. Fedushko S., Trach O., Kunch Z., Turchyn Y., Yarka U. Modelling the Behavior Classifi-
cation of Social News Aggregations Users. CEUR Workshop Proceedings. 2019. Vol
2392: Proceedings of the 1st International Workshop on Control, Optimisation and Analyt-
ical Processing of Social Networks (COAPSN-2019). p. 95–110. http://ceur-ws.org/Vol-
2392/paper8.pdf
30. Fedushko S., Ustyianovych T. (2020) Predicting Pupil’s Successfulness Factors Using
Machine Learning Algorithms and Mathematical Modelling Methods. In: Hu Z., Pe-
toukhov S., Dychka I., He M. (eds) Advances in Computer Science for Engineering and
Education II. ICCSEEA 2019. Advances in Intelligent Systems and Computing, vol 938.
Springer, Cham. pp 625-636. DOI 10.1007/978-3-030-16621-2_58