CC BY
34
PROPOSALS FOR JUSTIFICATION OF REQUIREMENTS TO HARDWARE CHARACTERISTICS OF PACKET TRANSPORT NODES OF THE COMMUNICATION NETWORK
Alexandr V. Anufrenko, DOI 10.24411/2072-8735-2018-10031
Military academy of communications
named after S.M. Budenny, St. Petersburg, Russia,
Andrey K. Kanaev,
St.petersburg state transport university of the emperor Keywords: aggregate node, transport
Alexander I, St. Petersburg, ^üü^ [email protected] network, simulation, GPSS STUDIO, QoS.
Proposals on substantiation of requirements to the characteristics of equipment of packet TN nodes are based on the developed models of traffic aggregation and distribution in the TN and methodological recommendations for the calculation of quality of service indicators in the planning and design of the TN.
With the help of the developed models, the influence of structural, functional and load parameters of the AN and a fragment of the TN on the characteristics of the aggregated traffic of the communication network is investigated. The models allow to set adequate requirements for the AN and the TN in compliance with the specified level of quality of service for all categories of network traffic.
The developed guidelines for the calculation of quality of service indicators contribute to improving the efficiency and accuracy of the calculation and evaluation of quality of service indicators. In the process of developing models and guidelines applied the laws of queueing theory, graph theory, probability theory, decision making theory, methods of analytical planning, fractal geometry, device simulation, etc. The novelty of the developed models is to take into account the protective mechanisms of communication networks with packet switching, as well as the pulsating nature of aggregate traffic. The novelty of the developed recommendations is the use of an integrated quality indicator to evaluate quality of service indicators in the analysis of the characteristics of aggregated traffic.
When using the calculated data obtained, the adequacy of decisionmaking in the process of planning and design of both ANs and the TN in terms of meeting the requirements for quality of service increases.
Information about authors:
Alexandr V. Anufrenko, Military academy of communications named after S.M. Budenny, Research fellow of the research center, St. Petersburg, Russia
Andrey K. Kanaev, St.Petersburg state transport university of the emperor Alexander I, Head of the department "Electrical Communication", St. Petersburg, Russia
Для цитирования:
Ануфренко А.В., Канаев А.К. Предложения по обоснованию требований к характеристикам оборудования узлов пакетной транспортной сети связи // T-Comm: Телекоммуникации и транспорт. 2018. Том 12. №2. С. 47-54.
For citation:
Anufrenko A.V., Kanaev A.K. (2018). Proposals for justification of requirements to hardware characteristics of packet transport nodes of the communication network. T-Comm, vol. 12, no.2, pр. 47-54.
r I Л
Introduction
Telecommunication technologies in recent years are experiencing large-scale changes associated with the construction of a new generation of networks (Next Generation Networks), which in the first placc displays the issues of the quality of the network (Network Performance) and quality of service (QoS).
However, in departmental networks, little attention is paid to aggregate node (AN), although they occupy a key place in the network, being an intermediate link between the transport network (TN) and the access network. The quality of functioning of the TN and access network depends on the quality of functioning of the AN 111.
Despite the fact that the approaches to the design of various departmental communication networks largely depend on the global trends in the development of telecommunications infrastructure and are guided by modern hierarchical models of networking, departmental telecommunications networks are often represented as a two-level planar model. In this case, the aggregation level functions are included in the functions of the TN, Therefore in this article wc will focus on the TN, terminating nodes which perform the functions of AN.
TN is a large and complex object of research. Accurate classification of structures TN is absent. Fig, 1 shows the developed version of this classification. Taking this option into account, the study considers a regional TN with a linear topology based on iithernet/DWDM technologies.
Structure of transport networks
Software F 1IDClioU.il Topological Physical
Basic software Middleware OTN Ethernet Point-point Hardware ATS Multiplexers
ATM MPLS-TP Linear chain Hardware DTS Switches
SDH (MG-SDH) Radial Hardware OTS Bridges
Ring Electrical components Routers
Territorial Logical Double ring Electronic components PBX
Backbone OSI ATM The radial-hub upioelectïonie components Gates
TCP IP OTN Fully connected Optical component Servers
Regional SDH Ethernet Cellular Software components
Fig, I. Classification of transport network structures
Also it should be noted that the previously created models and algorithms of functioning of the TN do not take into account the impact of protective mechanisms for packet networks, the QoS parameters under varying user load. However, the protective mechanisms are widely used in packet communication networks and include a lot of different options (Fig. 2).
For the formulation of proposals required development of models of aggregation and distribution of traffic in a TN communication, including models of the AN of a communication network and a model ofa fragment of the TN. The proposals are also based on methodological recommendations for the calculation of QoS indicators in the planning and design of the TN, including calculations obtained using the developed models.
Pnrtfrttvc nicc linnisms of I hi' aggi-cgnlr no [It- ami lrau%)jo!1
rouiuiiiDlrilioii llf tWttk ±
The availability of the aggregate qodf TUc effectivenm of the aggregate node Network security mechanisms
Link Aggregation Control Protocol (LACP) Hot Standby Router Protocol (HSRP) Spanning Tree Protocol (STP) h UlplUUTU
Port Aggregation Protocol (PAgP) \ muai Router Rcdundaitev Piotocol (VRRP) Media Redundancy Protocol (MRP)
Nortel MuliiLmk trunking Hipei ring
EtherChannel (rooking Gaiewny Load Balancing (GLBP) High AvflLlabtlttv Scande» Ring (HASR)
Adaptec ' i Duralink trunking Common Address Redundancy Protocol (CARP) Ethernet Ring Piotection (ERP)
Bidirectional Forwarding Detection (BFD)
Label Switched Path (LSP)
Fig. 2. Classification ofdefense mechanisms
On the basis of the developed variant of classilication of protective mechanisms, all of them can be divided into two large groups: network and node. Nodal, in turn, are divided into mechanisms that arc responsible for the availability of AN and mechanisms that arc responsible for the effectiveness of AN [2]. The belonging of some technologies to the appropriate class is shown in Fig. 2.
In view of the complexity of the analysis of the elements of the TN and objective difficulties associated with the possibility of rapid justification of decisions in the planning and design of TN should be considered relevant to the development of proposals for the selection of a set of nodal and network protective protocols for the appropriate structure of the AN.
The models of aggregation and distribution of traffic
in a transport network
The structure of the modeled fragment of the TN is presented in Fig, 3 and includes terminal nodes-AN (UA) and transit nodes (TC) of the TN.
A model of the AN include six self-developed fault-tolerant structures:
one switch;
one switch and one router connected in series and located on the same node;
one router;
two switches connected in parallel and located on the same node;
a switch and router connected in scries and backed up by a switch and router also having a serial connection. All elements are on the same node;
a switch and router connected in series and backed up by a switch and router also having a serial connection. The backup elements are on a different node [2-4J.
The models take into account the self-similar properties of aggregated traffic and node security mechanisms of packet switching communication networks [5].
where
A ~d+l+m+w+ z\ (9)
B=p- (d+m +z)+>f (d+l+in +zj ii ■ (m +z)+m z-(I-P )-m-zJ (10) C=/wfl-(d+m+z)+d(m+z)+mz-(l-%)-m-z]+d-(m-z-(l^)-,j (} ■mz-rl Id- (m -z) ->-mz-(l-% )m-zj]
D^fw [d-(m-z- (1-%)-m-z)+l [d-(m+z)+mz-(l-%)-m-z]j 2) +<i-l ■(m-z-(l-P)-m-z)]
E=d-l-w-fm-z-d-PJ-m-z) (13)
The average time spent on a computer attack is as follows:
w-m!p--d(z+^>---
it ' k'
15 V +4+3 +2 C+D (sj
According to F(t), i,, is presented in [9, 10]. As input data use the following values of time and probability, the
corresponding profite model of a computer attack:
2 min, f™" - 2 min/"1"™ = 2 min, " P.,
— 2 min,
= 2 min,
'« = 0,1. ..0,9.
Presented in the Fig. 6. the results show that the variance and the average time of the implementation of computer attacks on the channel and network levels, OSI not significantly different from each other.
Melhocl of implementation of computer attacks Result of calculation t„(iwu) at p = 0.8 Result of calculation F(l) (mm) at P — 0,S T^mto) o (min)
Hie analysis of data pickets at 1 lie data ¡ink layer 25 0,B 25 8,3
The analysis of data packs is at the network level 25 0.8 25 8.3
Fig. 6. The results of calculating the time spent on a computer attack
Model the impact of the en emy on a fragment of the transport network connection
Given that variety of computer attacks affect on the different levels of ISO of communication networks, one can assume that with the aim of reducing the time to implement them, the enemy will be affected in several ways. At the same time, one of the most vulnerable places that the enemy will affect is the monitoring system and the management system. On this basis, in the process of functioning of the TN fragment model the influence of equivalent computer attacks at the channel level is taken into account. The enemy implements computer attack type "Scanning the network and its vulnerabilities" in the following sequence:
direct the Ethernet frame with the Continuity Check Message with probability p| 2.| for the average time ti2, with the time
distribution function D(t);
direct Ethernet frame with message Loopback Message with probability p¡ii for average time / with time distribution
function L(l)\
direct Ethernet frame with Linktrace Message message with probability p|for average time /|21 with time distribution
function M(t)\
to implement the computer attacks enemy initially launches the so ft ware-hardware complex for the average time tlamell with the
function of time distribution W(t)\
in addition, during the implementation of the ocomputer attack difficulties, suggesting that they re-run for the average time t with the distribution function of time lit),
1 rTJKOf
Mathematical model of equivalent of a computer attack type "Scanning of network and its vulnerabilities": in the form of stochastic network is presented in Fig 7:
-On
t',2.2
"O
Fig. 7. Mathematical model of equivalent computer attack in the form ofstochastic network
Given the transformation rules of the relevant models [9, 10], calculation of the expression (3.26 and 3.27) for the integral distribution function of probability F(t), and the average time
/ impact of equivalent computer attack is following: wd-P,j , '(s+sltJ \~exp(Ak t)
F(t)-
-il.
3 w-t-Pl2: ■ (z + s2t) I -e.xpf£li t)
<f?-(s2k) ~s2k
(15)
t3 w-m l] 2J <(z+s3t) 1 — exp(Sk -i)
<p'-(s3k)
*VA =
E
w-d• !\,, -(z+s \k) 1 -exp(s\k t)
-Ù w-l-n,,-(z+s2) ) \~exp(sli I) <p'-(s2t) ' (s2ty
tv ■ m ■ /',, • (z + s3( ) \—exp(sik ■ t) 1 <p>(slt) (s\f
(16)
r
As input it uses the parameter values of the computer attack realization time given in [9, 10]: 2 min,^, = 5 min,
/¡77 = 2 min, t~ = 2 min, P]2i = 0,1.. .0,9, Pl22= 0,1...0,9, /»,,-0,1...0,9.
Given the fact that the order of calculations of the distribution function and the mean time for computer attacks other species, essentially, the same solutions discussed above [9, 10], we give only the results (Fig. 8).
The form ofa computer attack Method of implementation of computer attacks fca(min) F(t)
Network traffic analysis The analysis of data packets at the data link layer 25 8,3
Network traffic analysis The analysis of data packets at the network level 25 8,3
The parameters of traffic sourccs (load parameters)
Bil rate .kbps
The algorithmic delay of the coding, ins
Encapsulation delay, ins
The distribution of the delay of encapsulation
The number of bytes in the [Ethernet frame
W1 =( encoding sample duration in scconds * codcc bandwidth) / S bytes
W2 (the RTP head er) .by tes
W3 (the LIPP he ad er), by tes
W4 (the IP header), by to s
The intensity of the frame transmission, fps
W5 (frame service information) bytes
The law of distribution of intensity frames
Duration of ON-period, sec
Duration OFF- period, sec
The law of distribution of ON/OFF periods
The Coefficient Of Hurst
Structural and functional parameters
Number of inputs/outputs of the AN elements
Performance of the AN element, frames/sec
The capacity of the input buffer of the AN elements, frames_
The law of distribution of the intensity of frames processing by the element of the AN
The capacity of the output buffer of the AN, frames
The rate of frame transmission via the communication channels, fps The lime interval o I'failure of an element of the AN, sec
The time interval of failure of the main channel,sec
The lime interval oftlie recovery element, sec
The lime interval of the recovery channel, sec
The rate of frame transmission via the communication channels, fps The performance of the TN node, operation /sec_
Bil rate, bps
The capacity of the inpul buffer of the node of ilie TN, bit
The variant of the operation of the TN
Number of optical channel
The buffer capacity of the direction, bit
Delay of the protective mechanism, sec
The lime to failure of a node of TNs, sec
The recovery time ofa node of TNs, see
The lime to failure of optical channel, sec
The recovery time of optical channel, sec
Simulation models have a software implementation in the form of two programs: "Models of the network aggregate node", "Models of the transport network fragment" (Fig. 9-11).
In programs, you specify parameters according to the source data (Fig. 9):
Fig. 8. The results of calculating the time spent on a equivalent computer attack
Models of the network AN and a fragment of the TN are constructed using a discrete-event approach based on the simulation program "GPSS STUDIO". All functional modules of the developed models are described using the language General Purpose Simulation System World [2-4]. Each class of objects of the models has a certain set of functions and parameters that together describe the logic and regularities of their behavior.
Parametric tuning elements AN and TN occurs by means of description operators of program "GPSS STUDIO". The input parameters taken into account in the models are summarized in the Table 1:
Table t
Input parameters of the models
Models of the network Qggrefffe nodi
Planning Modeling
- O
Traffic generators Aggregate node General GT1 GT2 GT3
voice (Poisson) On Off mode
Coding
Encapsulation
DiraBon of
ON-period ( 5«)..
Dilation of Off-period (set)
Number of samples
Encoding delay
(sec) Encoding speed (kbps)
Fig. 9, Program window "Models of the network aggregate node". Tab "Traffic generators"
The values of structural and functional parameters of the AN are entered (Fig. 10):
" Made Is of the network aggregate node
Planning Modeling
a
Data entry
Results
Traffic gene rarars Aggregate node General
AN 1 AN 2 AN 3 AN 4 AN 5 AN 6
Switch 1
Frame process delay (sec)
The capacity of the input buffer (bits)
Router 1
Switch 2
AN 1
Number of AN channels
Fig. 10. Program window "Models of the network Tab "Aggregate node"
aggregate node".
You enter the values of the structural-functional parameters of the nodes of TN, The simulation time is set (Fig, 11),
The purpose of each tab of the programs is described in detail in the textbook "Simulation studies in the simulation environment GPSS STUDIO".
The developed models of aggregation and distribution of traffic in the TN allow to identify the impact of protective mechanisms on the basic parameters of the quality of service (frame delay time of aggregated traffic, jitter, loss of personnel) and justify the choice of the most appropriate mechanism, based on the requirements for the level of quality of service and reliability requirements.
Model of the transport network fragment
Planning Modeling
Data entry
Results
Traf fie genera rors Transport network Genera! TNetl TNet2
TNI TN 2 TN3 TN4 TNS TN6 TN7
TN MBinCti ResCh
Perfcnnance (bps) Biffer capacité' (bits)
Work time (sec)
Reccverv dm e (sec)
TNetswrtch
TNetl
f
■
Fig. II. Program window "Model of the transport network fragment". Tab "Transport network"
The Fig. 12 shows the effect of defense mechanisms on the latency of the aggregate traffic and jitter, demonstrating the fact that these parameters of nodes in different effects on the service quality indicators.
OjOi
r Frame del y
AIM defense
mechanism
3 Sfl
The formation of the integral quality index goes through the stages [11, 12] presented in Fig. 13:
Collection of network quality parameters
I
Single quality indicators 1
Generalized performance indicators
The integral indicator of quality
Fig. 12. The effect of defense mechanisms on the latency ofihe aggregate traffic and jitter
Methodical recommendations about calculation of quality of service indicators at planning and design of a transport network
Existing methods for assessing the quality of service in the analysis of the processes of aggregation and distribution of aggregated traffic in the TN communication to varying degrees have certain disadvantages:
the need to involve a large trained team of operators and auditors;
complexity and duration of measurements and processing of the results;
the need for specially equipped rooms with low noise levels; the need for a communication interruption at the time of testing, etc.
The developed guidelines for the calculation of service quality indicators allow rapid assessment in real time [11, 12].
Since the number of indicators that evaluate the quality of communication to users is quite extensive, the formation of an integrated indicator greatly simplifies the task of assessing the quality of communication services.
Fig. 13. The formation stages of the integral quality index
Collection of network quality parameters is made by measuring them at fixed points. Single quality indicators are formed from the corresponding parameters. Generalized performance indicators arc calculated directly using the quality parameters or unit quality indicators for the selected groups. The integral indicator of quality is calculated from the generalized indicators of quality of work. For the calculations can not be used in all summary measures, and some of the most important pointers of each group [11, 12].
A proposal for the selection of complex node and network protective protocols for the corresponding structure of the aggregate node
On the basis of the developed models and methodical recommendations the problem of substantiating the complex of nodal and network defense protocols for the corresponding structure of the AN is solved.
When calculating and assessing the required values of service quality indicators input load parameters were taken into account the variant of information load of the AN, presented in Tabie 2:
Table 2
The variant of aggregate informational loud
Category 1 subscribers Category 2 subscribers Category 3 subscribers The total number of subscribers
Telephony 16 21 64 101
Video telephony 16 22 63 101
Email 16 55 30 101
Access to the data- 16 55 30 ¡01
base
Videoconferencing 16 55 30 101
Fax 16 21 64 ¡01
Telegraph commu- 18 66 17 101
nication
Data transmission 16 55 30 101
Thus, 202 voice traffic generators, 505 data traffic generators and 101 video traffic generators functioned in the modeling process.
The structural and functional parameters were set taking into account the fact that the first variant of the AN model structure was chosen and the characteristics of the telecommunication equipment presented in the Table 3 were set. In the simulation included the effects of computer attacks of the type "Network traffic analysis".
Table 3
Estimates of the latcncy of the protective protocols
Type of hardware Defense mechanisms Delay (ms) Type of hardware Defense mechanisms Delay (ms)
RT-3806 «Nate*» HSRP, VRRP 34 Dionis-NX series 4000 LACP 46
Juniper series SRX240 JSRP 45 Dionis-NX scries 3000 LACP 45
Juniper series SRX650 JSRP 44 Microlink series IPSW 3300 HSRP, VRRP 31
Juniper series MIOi JSRP 47 Cisco WS C3650-24PS-E LACP 46
Huawei series AR 2200 RRPP, VRRP 36 Huawei Quid way S5700 RRPP. VRRP 30
Microlink series SL IPSW-SYS-UR2000 HSRP, VRRP 33 QTech QSW series 3300 VRRP, ERRP 34
RKSS RSQS6850 LACP, VRRP.BFD 35
The simulation results showed that Huawei Quidvvay S5700 series protective mechanisms bring the lowest delay, and the largest - Juniper MIOi series. Taking into account the simulation results, the use of Huawei Quidvvay S5700 series telecommunication equipment is proposed for the AN consisting of one telecommunication device processing a load similar to the one presented in the Table 3.
Conclusion
The account in models of aggregation and distribution of traffic in a TN of protective mechanisms for packet network facilitates a more accurate selection of fault-tolerant structures such as AN and TN. The use of an integrated indicator of quality of service in the developed guidelines contribute to a more accurate assessment of QoS indicators. The proposals formulated on the basis of models and methodological recommendations increase the validity of the
choice of nodal and network security mechanisms in the design and planning of AN and TN, taking into account the provision of QoS indicators in the process of processing of aggregated traffic on AN and its distribution in the TN.
References
1. Anufrenko A.V. Volkov D.V. and Kanaev A.K. (2015). The organizing principle of AN multiservice communication networks. Collection of scientific papers by IV international science.-technical, and science.-method. conference "Actual problems of infbtelecommunication in science and education", vol. 2, pp. 203-206,
2. Aiuifrenko A.V., Barantsev A.V.. Kanaev A.K. (2015). Fault-tolerance communication networks based on packet-oriented technologies. Proceedings of jubilee. I'm 70. scientific.-technical, conference dedicated to radio Day, pp. 304-306.
3. Anufrenko A.V,, Vasiliev E.L., Kanaev A.K. and Subbotin D.V. (2017), Models of the AN in a telecommunication network. Proceedings of the I! interuniversity scientific-practical conference "Problems of technical support of troops in modern conditions ", pp. 63-65.
4. Anufrenko A,V„ Bennetta E.V. and Kanaev A.K. (2017). A Simulation model of the AN to study the effect of structural parameters of the AN, the aggregate traffic parameters. Transport of Ural, no. 2 (53), pp. 47-50.
5. Shelukhin O.I., Osin A.V. and Tcnacshen A.M. (2003), Fraktal-nie protcessi v telecomunikaciakh [Fractal processes in telecommunications], Radiotekhnika, Moscow.
6. Boev V.D. (2004). Moclelirovanie si.stem Instrumentalnie sred-stva GPSS World [Modeling of systems. Tools GPSS World], Bkhv-Peierburg, St. Petersburg.
7. Aliyev T.l. (2009). Osnovi modelirovania diskretnikh system [Bases of modeling of discrete systems], SPbGU ITMO, St. Petersburg.
8. Boev V.D. (2009), Komputernoe modelirovanie. Elementi teorii i praktiki [Computer simulation. Elements of theory and practice], VAS, St. Petersburg.
9. Kocinac M.A., Osadchv A.I., Kocinac M.M., Of Louth O.S., Dementyev V.E. And Vasyukov D.Y. (2014), Obespechenie otkazoys-toichivosti informatsionno-telekommynikacionnikh setei v ysloviakh informacionnogo protivoborstva [Sustailiability of information and telecommunications networks in conditions of information warfare], St. Petersburg.
10. Privalov A.A. (2001), Metod topologicheskogo preobrazovania stohasticheskikh setei i ego ispolzovania dla analiza setei cvazi VMF [ Method of topological transformation of stochastic networks and its use for analysis of communication systems of the Navy], VMA, St. Petersburg.
11. Saali T. and Kerns K. (1991). Analiticheskoe planirovanie. Organizacia sistem [Analytical planning. Systems organization], Radio i cvaz, Moscow.
12. Saali T. (1989). Prinatie reshenii. Metod analiza ierarhii [Decision-making. Method of analysis of hierarchies], from English. Radio i cvaz, Moscow.
ПРЕДЛОЖЕНИЯ ПО ОБОСНОВАНИЮ ТРЕБОВАНИЙ К ХАРАКТЕРИСТИКАМ ОБОРУДОВАНИЯ УЗЛОВ ПАКЕТНОЙ ТРАНСПОРТНОЙ СЕТИ СВЯЗИ
Ануфренко Александр Викторович, Военная академия связи им. СМ. Буденного, Санкт-Петербург, [email protected]
Канаев Андрей Константинович, Петербургский государственный университет путей сообщения Императора Александра I,
ФГБОУ ВО ПГУПС, Санкт-Петербург, [email protected]
Дннотация
Предложения по обоснованию требований к характеристикам оборудования узлов пакетной транспортной сети связи основываются на разработанных моделях агрегации и распределения трафика в ТС и методических рекомендациях по расчету показателей качества обслуживания при планировании и проектировании транспортной сети. При помощи разработанных моделей осуществляется исследование влияния структурных, функциональных и нагрузочных параметров узла агрегации и фрагмента транспортной сети связи на характеристики агрегированного трафика сети связи. Модели позволяют задавать адекватные требования к узлу агрегации и транспортной сети связи в условиях соблюдения заданного уровня качества обслуживания для всех категорий трафика сети. Разработанные методические рекомендации по расчету показателей качества обслуживания способствуют повышению оперативности и точности расчета и оценки показателей качества обслуживания. В процессе разработки моделей и методических рекомендаций применены законы теории массового обслуживания, теории графов, теории вероятности, теории принятия решений, методы аналитического планирования, знания фрактальной геометрии, аппарат имитационного моделирования и т. д. Новизна разработанных моделей заключается в учете защитных механизмов сетей связи с коммутацией пакетов, а также пульсирующего характера агрегированного трафика. Новизна разработанных рекомендаций заключается в использовании интегрального показателя качества для оценки показателей качества обслуживания при анализе характеристик агрегированного трафика.
При использовании полученных расчетных данных возрастает адекватность принятия решений в процессе планирования и проектирования как узла агрегации сети связи, так и транспортной сети связи в условиях выполнения требований к качеству обслуживания.
Ключевые слова: узел агрегации, транспортная сеть связи, имитационное моделирование, GPSS STUDIO, QoS. Литература
1. Ануфренко А.В., Волков Д.В., Канаев А.К. Принцип организации узла агрегации мультисервисной сети связи / Актуальные проблемы инфотелекоммуникаций в науке и образовании. IV Междунар. науч.-технич. и науч.-метод. конференция: сб. науч. ст.: в 2 т. Спб.: СПб ГУТ, 2015. С. 203-206.
2. Ануфренко А.В., Баранцев А.В., Канаев А.К. Обеспечение отказоустойчивости сетей связи, функционирующих на базе пакетно-ориентированных технологий / Юбил. 70-я Всерос. Научн.-технич. конференция, посвященная Дню радио. СПб.: ЛО ЦНИИС, 2015. С. 304-306.
3. Ануфренко А.В., Васильев Е.Л., Канаев А.К., Субботин Д.В. Модели узла агрегации в телекоммуникационной сети / Труды II межвузовской научно-практической конференции "Проблемы технического обеспечения войск в современных условиях". Спб. 2017. С. 63-65.
4. Ануфренко А.В., Бенета Э.В., Канаев А.К. Имитационная модель узла агрегации для исследования влияния структурных параметров УА на параметры агрегированного трафика // Транспорт Урала. Екатеринбург. 2017 г. № 2 (53). С. 47-50.
5. Шелухин О.И., Осин А.В., Тенякшев А.М. Фрактальные процессы в телекоммуникациях. М.: Радиотехника, 2003. 480 с.
6. Боев В.Д. Моделирование систем. Инструментальные средства GPSS World: Учеб. пособие. СПб.: БХВ-Петербург, 2004. 240 с.
7. Алиев Т.И. Основы моделирования дискретных систем. СПб.: СПбГУ ИТМО, 2009. 363 с.
8. Боев В.Д., Сыпченко Р.П. Компьютерное моделирование. Элементы теории и практики. СПб.: ВАС, 2009. 436 с.
9. Коцыняк М.А. Обеспечение устойчивости информационно-телекоммуникационных сетей в условиях информационного противоборства / М.А.Коцыняк, А.И. Осадчий, М.М. Коцыняк, О.С. Лаута, В.Е. Дементьев, Д.Ю. Васюков. СПб.: ЛО ЦНИИС, 2014. 126 с.
10. Привалов А.А. Метод топологического преобразования стохастических сетей и его использования для анализа систем связи ВМФ. Спб.: ВМА, 2001. 186 с.
11. Саати Т., Кернс К. Аналитическое планирование. Организация систем: Пер. с англ.: Радио и связь, 1991. 224 с.
12. Саати Т. Принятие решений. Метод анализа иерархий: Пер. с англ. М.: Радио и связь, 1989. 316 с.
Информация об авторах:
Ануфренко Александр Викторович, Военная академия связи им. С.М. Буденного, ФГКВОУ ВО ВАС научный сотрудник научно-исследовательского центра, Санкт-Петербург, Россия
Канаев Андрей Константинович, Профессор, д.т.н., Петербургский государственный университет путей сообщения Императора Александра I, ФГБОУ ВО ПГУПС заведующий кафедрой "Электрическая связь", Санкт-Петербург, Россия
T-Comm Том 12. #2-2018
