Review

Received October 25, 2020, accepted November 6, 2020, date of publication November 10, 2020,
date of current version November 20, 2020.

Digital Object Identifier 10.1109/ACCESS.2020.3037143
A Review of the Methods Used to Model Traffic

Flow in a Substation Communication Network
MATTHEW LIAM DE KLERK AND AKSHAY KUMAR SAHA , (Member, IEEE)
Discipline of Electrical, Electronics and Computer Engineering, University of KwaZulu-Natal, Durban 4041, South Africa
Corresponding author: Matthew Liam De Klerk (mattydeklerk@gmail.com)
This work was supported in part by the University of KwaZulu-Natal.
ABSTRACT A significant component in electrical power systems are substations, in which the substation
communication network is essential. Due to the evolving nature of communication protocols, and the
introduction of the IEC 61 850 standard, substation communication networks can provide fast and reliable
information transmission. Such improvements in transmission allow for improved protection and control of
the power system network. A well-designed substation communication network can be achieved with the
use of appropriate modeling, and as a result, this paper aims the review the methods in which parts of, or the
entire substation communication network can be modeled. Modeling methods aim to determine various
parameters of a substation communication network design, which include the traffic load distribution, and
various message delays under different network conditions that may be experienced. In order for modeling
to be carried out, a general structure of a substation communication network must be defined, which is
laid out by the IEC 61 850 standard. The findings of the review in this paper show that there are various
methods in which a substation communication network can be modeled, which include analytical modeling,
software-based modeling, hardware/experimental modeling, and time-series based modeling. However, each
method enables for different parameters to be found, with some of the methods having drawbacks including
a limited network model size, and inability to find certain necessary parameters. The methods reviewed in
this paper can be used and expanded on to create useful and accurate models of a substation communication
network.
INDEX TERMS Ethernet networks, IEC 61850, network modeling, SCADA, substation automation system
(SAS), substation communication network (SCN), SCN structure and architecture.
ABBREVIATIONS Long-Range Dependence LRD

Manufacturing Message Specification MMS
Active Distribution Network ADN Merging Unit MU
Auto Correlation Function ACF Optimized Network Engineering Tool OPNET
Auto Regressive Moving Average ARMA Partial Auto Correlation Function PACF
Breaker Intelligent Electronic Device BIED Precision Time Protocol PTP
End-to-End ETE Priority Queuing PQ
Finite State Machine FSM Protection and Control P&C
First in First Out FIFO Rapid Spanning Tree Protocol RSTP
Fractional Auto Regressive Integrated Sampled Value SV
Moving Average FARIMA Simple Network Time Protocol SNTP
Generic Object-Oriented Substation Event GOOSE Substation Automation System SAS
Intelligent Electronic Device IED Substation Communication Network SCN
Supervisory Control and Data Acquisition SCADA
Switch Intelligent Electronic SIED
System Integrity Protection Schemes SIPS
The associate editor coordinating the review of this manuscript and Transmission Control Protocol TCP
approving it for publication was Rentao Gu . Virtual Local-Area Network VLAN
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
VOLUME 8, 2020 204545
M. L. De Klerk, A. K. Saha: Review of the Methods Used to Model Traffic Flow in a Substation Communication Network
I. INTRODUCTION In another paper, M. Aftab et al. [9] review the communica-

Electrical power systems are essential in modern society, tion technologies that can be employed in active distribution
as they enable the continuous and reliable availability of networks (ADNs). Notably, one section of the technologies
electrical power, which is of paramount importance to many employed falls under the IEC 61850 standard. As a result,
industries requiring electricity to operate. Substations are the research scholars in [9] provide an overview of the stan-
critical nodes in an electrical power system. The substation dard, as well as a discussion around performance evaluation
automation system (SAS) uses information retrieved from the metrics in ADNs. Methods for performance evaluation are
power system in order to ensure proper operation and man- very briefly discussed, but only software-based tools are
agement of the power system network [1], [2]. The functions mentioned [9].
of the SAS include the protection, monitoring and control of V. Mathebula et al. [10] provide a reliability review of
both the primary equipment in substations, as well as their mission critical safety functions in an IEC 61850 based SAS.
associated feeders [2]–[4]. Protection is a major component, The paper provides a comprehensive overview of substa-
as failure of protective equipment in a substation to operate tion automation systems that utilise the IEC 61850 standard,
under fault conditions can lead to severe damage to equip- as well as a discussion around their reliability [10]. However,
ment, danger to human life and instability [5]. the paper does not discuss the performance evaluation of IEC
Software-based substation automation systems, devel- 61850 based SCNs or methods in which traffic flow in the
opments in microprocessor-based intelligent electronic networks can be modeled to evaluate the performance of the
devices (IEDs) and the introduction of the IEC 61850 stan- network [10].
dard for communication systems in substations have led As a result, this paper aims to provide a comprehensive
to a decrease in the use of rigid parallel copper wiring, review, evaluation, and comparison of the methods in which
as it is replaced with an Ethernet-based communication the traffic flow in a substation communication network can be
network [1], [3]. The large changes in the communica- modeled. Although these methods have been briefly reviewed
tion networks employed have led to an increasing depen- in other papers, a comprehensive review and analysis has not
dency of the SAS on the substation communication network yet been presented. A brief overview and description of the
(SCN) [2], [3], [6]. As a result, improved operation and IEC 61850 communication standard is provided in section
management of the SAS can be achieved through the correct II. Section III of this paper reviews the general structure
design and maintenance of an SCN, allowing for fast and of an SCN, which relates to the models reviewed. Addi-
reliable information transmission [3]. Hence, the modeling of tionally, a review of redundancy and architectures that can
SCNs in a simple and reliable manner is essential, allowing be used in an SCN, as well as the use of SCADA systems
for the evaluation of network performance under different is also presented in section III. As previously mentioned,
network conditions and the determination of whether or not the traffic load distribution and maximum message delays
the SCN meets the network performance requirements [3]. are essential parameters in the analysis of an SCN; hence,
The traffic load distribution, and the maximum message section IV aims to briefly discuss these parameters and their
delay under different network schemes and equipment selec- relevance in a communication network. The review of the
tions can be used to analyze the performance of an SCN various modeling methods, which include analytical models,
design [3]. The IEC 61850-5 standard specifies that the software-based models, hardware/experimental models, and
requirements for message transmission time for a SAS must time-series based models is contained in section V, with a
be maintained under any possible operating condition or con- comparison of the modeling methods detailed in section VI.
tingency inside the substation [7]. Modeling of the commu- Finally, a conclusion, as well as suggestions for the expansion
nication network allows for the dynamic performance of the of the literature, is provided in section VII.
communication in the SAS to be determined and any issues
that may occur to be found and rectified. II. THE IEC 61850 COMMUNICATION STANDARD
Due to the developing nature of substation automation, and The IEC 61850 standard was developed to allow for inter-
the importance of the subsequent communication network, operability between devices made by different manufactur-
there are various survey and research papers that investigate ers. The guidelines allow for peer-to-peer communication
the IEC 61850 standard. M. Aftab et al. [8] provide a holistic between IEDs in the substation [11]–[13]. One of the main
overview of the IEC 61850 standard, with discussions on the aims of the standard is to solve interconnectivity issues
background of the standard, the functions and requirements a that arise from the use of IEDs from different manufactur-
system must have to meet the standard, performance evalua- ers [10], [13]–[15]. This is achieved through the adoption of
tion of communication networks using the standard and appli- an abstract architectural construct, which is in the form of data
cations fields in which the standard may be applied. Although objects and services that do not depend on the underlying pro-
the paper [8] does consider performance evaluation of SCNs, tocols [10], [14], [15]. This means that the IEC 61850 stan-
with methods of modeling discussed briefly, the paper does dard specifies a communication protocol that allows for IEDs
not contain an in-depth and critical review of the methods from different manufacturers to communicate using a stan-
available in literature. dardized and beneficial method of communication. There are
204546 VOLUME 8, 2020

various benefits and features which the standard provides,

which include [4], [10], [11], [14]–[16]:
• Fast transmission of Generic Object-Oriented Substa-
tion Events utilizing the peer-to-peer mechanism which
the standard enables.
• The data definition in the communication network is
based on object-oriented modeling.
• Standardized data modeling.
• The system scalability is independent of the manufactur-
ers of the system’s devices.
• Switched Ethernet technology forms the basis of the
standardized high-level communication services.
• The application of the appropriate bus topology allows
for a communication network with increased reliability.
• The specified architectures are simple and cost-
effective.
• Communication extendibility and data integrity. FIGURE 1. Mapping data model and communication services – OSI
layered model [1], [11], [15], [17]–[19].
More specific to the application of this paper, the IEC 61850
standard specifies the medium of communication and pro-
vides a set of functions, formats and layers which define how communication stack, as the stack is structured according to
signals or information is to be transferred between substation the OSI model [11], [13], [18]. The OSI model has 7 layers,
devices [5]. Substation devices generally include IEDS that and as a result, the stack structure consists of Ethernet in
are used for protection, control and monitoring in a substa- layers 1 and 2, TCP/ICP in layers 3 and 4, and MMS in layers
tion.There are various message classes that are specified for 5 to 7. Time-critical messages, which include SV and GOOSE
use in the standard, which include message databases such messages, are mapped to the Ethernet linked layer in order
as; client-server messaging, Generic Object-Oriented Substa- to allow for fast communication of messages in real-time
tion Event (GOOSE), Manufacturing Message Specification using peer-to-peer communication; however, client/server
(MMS), and Sampled Values (SV) [5], [17]. The functionality communication is mapped to layer 7, which is the MMS
of these message classes include [14], [16], [18], [19]: application layer [11], [17], [18]. The time-critical nature
• Client/server messaging is communication between of SV and GOOSE messages means that they are expected
IEDs and the supervisory control and data acqui- to have fast and reliable synchronous message transmission,
sition system (SCADA) and usually consists of meeting the high requirements associated [11], [10].
request/response sequences.
• GOOSE messaging is a service that is commonly used III. THE STRUCTURE OF AN SCN
to transfer time-critical information with a very fast Fig. 2 shows the general structure of an SCN based on the
speed. The time-critical information that can be trans- IEC 61850 standard. The figure indicates that the IEDs in
ferred using GOOSE messaging includes trips between the bay level, used for control, monitoring and protection, are
IEDs, status changes, releases, and blockings. These are connected with the station level devices through the station
examples of events in a substation. GOOSE messaging bus. A single protection & control (P&C) IED can also be
supports peer-to-peer communication. used in the bay level. A P&C IED allows for the integration
• The MMS protocol is generally to transfer data between of the protection and control functionalities required in the
the station level IEDs, and the control and protection bay unit of a substation [7], [18], [24], [25]. Simple network
IEDs, which are in the bay level. time protocol (SNTP) messages are used in order to achieve
• Sampled value messages are used in order to transfer time synchronization of the sequence events in the station bus
current and voltage sample information. The informa- network. In addition to SNTP messages, the station bus also
tion being transmitted is transmitted quickly in a syn- carries GOOSE and MMS messages [3], [26].
chronized stream of samples. Additionally, the bay level IEDs are connected to the
The standard divides the SCN into three levels, which are switchyard devices through the process bus [11], [22], [24].
the process level, the bay level, and the substation, or station Merging units (MUs) are IEDs, used to convert signals from
level [11], [20]–[23]. More detail on the three levels specified instrument transformers to IEC 61850 telegrams [11], [18].
is provided in section III of this report, in which the general The breaker or switch (BIED, SIED) IEDs provide the pro-
structure of an SCN based on the IEC 61850 standard is cess bus functionality for the switchgear. In general, the pro-
discussed in depth. Fig. 1 shows the mapping of various cess bus network carries generic object-oriented substation
message types to the OSI layered communication stack. The events (GOOSE), sampled value (SV), and manufacturing
IEC 61850 standard uses mainstream technology for the message specifications (MMS) messages and protocols [3].
VOLUME 8, 2020 204547

FIGURE 3. Star type architecture - SCN [10], [21], [23], [30].
FIGURE 2. General Structure of an SCN [3], [11], [18], [19], [25], [26].
impact on the performance and cost of the SAS [10], [21].
The reliability, survivability, availability, and response time
Additionally, the sampled values in the process bus net- are important features of an SCN, which can be affected
work are synchronized by precision time protocol (PTP) by the design of the communication network architec-
messages. GOOSE, SV and PTP messages all make use ture. The SCN architecture is not standardized in the IEC
of connectionless multicast transmission, the conversion of 61850 standard; however, cascade star, star, ring, star-ring
many signals to one [3], [27]. For this purpose, it is use- and redundant ring architectures are recognized as the gen-
ful to be able to limit the broadcast domain and separate eral SCN architectures by the Institute of Electrical and
different types of traffic; this can be done through the use Electronic Engineers/Power System Relaying Committee
of virtual local-area networks (VLANs) [3]. However, con- (IEEE/PSRC) [10], [28], [29].
nectionless multicast transmission is not used for the trans- Fig. 3 shows the implementation of the star architecture
mission of MMS messages. Instead, connection-oriented in an SCN. The figure shows that the bay level switches
transmission control protocol (TCP) is utilized as the MMS are all connected to the station level by individual links
protocol is usually used to transfer data between the bay and that the station level consists of a single centralized
level IEDs and the station bus. The TCP protocol provides switch [21], [22], [29]. Each bay level switch is further
inter-network, connection-oriented end-to-end packet deliv- connected to the IEDs in that bay. However, as there is
ery, ensuring reliability [3], [19]. only a single central station switch, it becomes a point of
In general, the station and process buses in an SCN con- failure for the entire SCN [21], [22], [29]. As a result,
sist of an Ethernet switch. These allow for the VLANs the star architecture has no redundancy and offers the
in the SCN to be set up and enables the SCN to support least reliability [10], [22], [29], [30]. Therefore, despite
the features of the SAS, which are distributed to multiple the fact that the star architecture is easy to maintain and
IEDs. The logical architecture of the SCN is designed with offers the lowest network latency, it is only suitable for
consideration for reliability and fast deterministic delivery monitoring [10], [22].
requirements of messages used in the control and protec- Fig. 4 shows the implementation of the ring architec-
tion functions of the substation [1]. The structure of the ture in an SCN. The ring architecture provides inherent
SCN shown in Fig. 2 allows for optimal communication redundancy using the Rapid Spanning Tree Protocol (RSTP)
between devices in the network, as well as communica- or priority protocols as it is connected in a ring of
tion to the control centre. A SAS based on the IEC 61850 switches. This means that there is no single point of fail-
standards uses server/client configuration, which allows for ure [10], [21], [22], [25], [26]. However, despite the inherent
shorter transmission times of substation messages on event redundancy that the ring architecture provides, when the
occurring or request [10], [16]. As a result of this, time- failure of one link or switch occurs, time for network recon-
critical applications such as protection functions benefit figuration is required [10], [22].
from the use of IEC 61850 based SAS’s as they require The SAS’s reliability, survivability and availability can
messages to be transmitted immediately when an event be enhanced by redundancy in the communication system.
occurs [10], [16]. Many redundant SCN architectures can be developed from
As already discussed, the IEC 61850 standard divides an the star and ring architectures. Additionally, IEDs which have
SCN into a station level, a bay level, and a process level. parallel communication ports also improve redundancy in
These levels are connected by the communication network, the communication network [21]. The IEC 61850 standard
and as a result, it is essential to consider the architecture specifies that inoperability can be achieved with a commu-
of the communication network. As the communication net- nication architecture design that has no single point of fail-
work is a major part of the SAS, the architecture has an ure [10], [22], [25], [28]. As a result of the importance of
204548 VOLUME 8, 2020

FIGURE 5. Transmission time of a GOOSE message between IEDs [11],

FIGURE 4. Ring type architecture - SCN [10], [21], [23], [30]. [10], [34].
the SCN architecture, the design of the architecture should across a network at a given point in time [33]. When there
consider [10], [21], [22]: is more traffic in the network, larger queuing delays are
• Interoperability. experienced [33]. The traffic load distribution provides an
• Maintainability. indication of how the traffic is distributed in the network
• Dependability. and allows for the identification of switches with high traffic
• Latency delays. load [3].
• Redundancy in the design. The substation architecture, as well as the processing time
• Bandwidth. of the communication devices, determines the message trans-
• Network segmentation. mission time in an SCN [10], [11], [34]. The factors affecting
• Network convergence. the transmission time are shown in Fig. 5. In Fig. 5, tb is
• Scalability and expandability. dependent on the network architecture. This network archi-
A supervisory control and data acquisition (SCADA) sys- tecture delay gives an indication of the total latency caused
tem enables the monitoring and control of systems and pro- by the network queuing and processing time at each sub-
cesses that are geographically dispersed [31]. The SCADA station communication device along the message path. The
system is a centralized system, from which information for architecture latency is representative of the end-to-end (ETE)
an entire substation can be obtained [31], [32]. Increased performance of the SCN. Fig. 5 reiterates the importance of
productivity and cost-saving are some of the benefits that can the SCN in a SAS [10], [11], [34].
be obtained from a SCADA system, as it allows for the early
detection and analysis of faults, also enabling improved main- V. METHODS TO MODEL TRAFFIC FLOW IN AN SCN
tenance [31], [32]. The substation SCADA system, coupled As SCNs are an extremely important part of a substation,
with the SAS creates a comprehensive method in which to and significantly affect the SAS, the design of SCNs must be
monitor and control the system, providing protection, control, reviewed to ensure that they produce a fast and reliable form
automation, monitoring and communication abilities [32]. of transmission for messages required in the SAS. The analy-
As discussed previously, a substation network often consists sis of the performance of an SCN can be carried out through
of IEDs, transformers (from which merging units collect the modeling of the communication network. The distribution
analogue values), switchgear, and cables and communication of the traffic load, as well as the maximum message delay,
links [27]. On account of the benefits a SCADA system are notable parameters in the analysis of the communication
provides, modern substations often use a SCADA system in network’s performance. This section of this paper aims to
conjunction with the SAS. However, the focus of this paper investigate the methods that can be used to model SCNs and
is specifically to review the modeling methods of an SCN in obtain various parameters.
order to assist in the development of a fast and reliable SAS.
As a result of this, section V presents the methods to model A. ANALYTICAL MODELING – NETWORK CALCULUS
the communication networks, with a more in depth analysis AND PATH FLOW MODEL
of the use of SCADA systems falling out of the scope of this Reference [3] provides an analytical method that can be
paper. used to model the traffic flow in an SCN. The analytical
traffic-flow model proposed can be applied to find various
IV. SCN TRAFFIC LOAD DISTRIBUTION AND MESSAGE network parameters, which include the traffic load distribu-
TRANSMISSION TIME IN AN SCN tion, as well as the maximum message delay. The research
The traffic load in a portion of a communication network scholars in [3] initially propose a physical connection model
gives an indication of the rate at which packets are transmitted for the ports in the SCN, as well as logical connection mod-
in the proximity of a given node. In simpler terms, network or els for both the VLAN and TCP mechanisms in the SCN,
data traffic considers the amount of data which is transmitting generated from a model representing the general structure
VOLUME 8, 2020 204549

element C ij is equal to a 1 if there is a logical connection

(dashed arc in Fig. 7 and Fig. 8) directed from port j to port
i, and is 0 otherwise [3]. The C matrix is formed with the
use of another two matrices in which the VLAN IDs and
port-based VLAN IDs are considered [3]. Lastly, as part of
the logical connection model for the TCP mechanism, it is
proposed that a P × P matrix C a,b can be found to denote the
logical connection for the message sent from port a to port
b [3]. The C a,b matrix is found using other matrices, in which
the TCP established communication connection relationship,
and the subordinate relationship between ports and switches
are considered [3].
In addition to the physical and logical connection models
generated, the analytical modeling that [3] proposes also
requires the formation of a source and a service model of
the traffic flow in the SCN. Ports which are intended to
send messages are traffic-flow sources, for which a source
model is developed. The source flow model allows for the
FIGURE 6. General Structure of an SCN used for analytical modeling [3]. properties of the traffic-flow sources to be described [3],
[35], [36]. The triggering relationships between GOOSE
flows, the message rate, message length, the data rate of
traffic flows and correspondence between a port and the
messages it sends are some of the properties that the source
model aims to describe [3], [35], [36]. Conversely, the service
model aims to describe the service that traffic flow receives
from the switches [3], [35], [36]. The network calculus the-
FIGURE 7. Flow diagram of the process bus network [3]. orem is used to develop both the source and service models
of the SCN.
The traffic flow can be described by a cumulative function
using the network calculus theorem. The basic concepts of
the network calculus theorem are the arrival and service
curves [3]. Reference [3] proposes that a matrix S can be
used in order to create a source model for the SCN. Matrix
S is a P × D matrix, in which P is the total number of ports,
FIGURE 8. Flow diagram of the station bus network [3]. and D is the total number of messages in an SCN [3]. The
corresponding relationship between messages sent and their
of an SCN [3]. Fig. 6 shows an example of the model of sources is described by matrix S. When considering the S
the general structure used. This is analogous to the general matrix, the element Sij is equal to aj if port i is the source
structure described in section III. However, the ports and of message j. The research carried out in [3] suggests that
physical connections of the components and levels of the the transmission characteristics of the various messages in
substation communication network are labelled, which aids the source model should also be considered. These consid-
in the development of the matrix models for the analytical erations include the fact that SV flows have a fixed data
method proposed, as well as the flow diagrams for both the rate, fixed intervals are used when sending PTP and SNTP
process and station buses generated from the matrices, seen messages (meaning the SV source model is the same as the
in Fig. 7 and Fig. 8. PTP and the SNTP messages source model) and that only in
The physical connection model proposed uses a matrix A the steady-state period does the GOOSE flow i have a fixed
to represent the physical connections, in which A is a P × P data rate [3]. The steady-state period represents the period in
matrix, where P is the number of physical ports in the SCN. which no fault or warning event occurs. In the steady-state
In the matrix, element Aij is equal to a 1 if there is a physical period, the lowest data rate of GOOSE flow is utilized; how-
connection directed from port j to port i, and is 0 otherwise ever, when an event occurs, the related GOOSE flow changes
[3]. The formation of such a matrix can be represented by the to the highest data flow rate [3]. The data rate then declines
diagrams in Fig. 7 and Fig. 8, in which the solid arcs represent exponentially; however, the change in rate creates a traffic
the physical connections. burst for a very short period of time [3]. Lastly, time-driven
Similarly, the logical connection model for the VLAN MMS flow also has the same source model as that of SV
mechanism proposed uses a matrix C to represent the logical flow, as MMS messages are configured by a time-driven or
connections, in which C is a P × P matrix. In the matrix, an event-driven mode [3].
204550 VOLUME 8, 2020

The service model proposed by [3] suggests that the queue scenario can be used in order to set the network monitoring
scheduling policies must be taken into account in order for threshold. Traffic load lower than this threshold at some point
the model to be developed. The priority queuing (PQ) and in the network could indicate that a packet loss fault or link
the first in first out (FIFO) queue scheduling policies are disconnection has occurred [3].
the two most important that must be considered [3]. The PQ The maximum load scenario occurs in the opposite case,
policy indicates that packets are served in order of prior- when all of the GOOSE flows are in bursting period, indi-
ity (higher priority packets first), whereas the FIFO policy cating the maximum traffic load distribution of an SCN. The
indicates that packets are served in order of arrival. These maximum traffic load that a link may withstand can be found
queue scheduling policies are commonly used to improve the using the maximum load scenario; however, this scenario may
quality of service as well as to achieve congestion control [3]. not exist practically. The maximum load scenario is essential
The service policy of packets depends on the priority of the for equipment selection and the design of the network [3].
packets received. In the case of packets of different priorities, Lastly, when there is an occurrence of events or faults in a
the service policy between packets is PQ, whereas, when primary system, it is defined to be the typical event scenario.
considering packets that have the same priority, the FIFO The network performance under conditions in which faults
service policy is used [3]. or alarms occur can be analyzed with the application of
Having developed the connection, supply, and service the traffic-flow model and calculation algorithm to a typical
models, [3] proposes a traffic-flow calculation algorithm, event scenario. [3]. The analysis, as previously mentioned,
which can be used to obtain the distribution of messages, is carried out with the prediction of the maximum message
the traffic load and the maximum message delay. The mes- delay and the traffic load distribution.
sage delay is found using an algorithm developed, which
makes use of the various matrices found in the models devel- B. ANALYTICAL MODELING – ADAPTED NETWORK
oped. The source model matrix, S, is used in the traffic CALCULUS AND PATH FLOW MODEL
load distribution algorithm. The traffic load distribution can Reference [38] aims to extend and simplify the analytical
be calculated under different conditions by considering the model proposed and developed by [3]. The proposed adap-
GOOSE triggering relationship as part of the development tation of the network calculus and path flow model also
of the load distribution calculation algorithm. An example of discusses the minimum load, maximum load, and typical
such a condition is a single-phase short circuit [3]. The delay event scenario; however, a fourth scenario, called the network
of messages in a switch can be divided into four types, which abnormal flow calculation scenario is added. This scenario
are the transmission delay, the packet receiving delay, the pro- considers abnormal states of the network due to anomalies
cessing delay, and the queuing delay [3]. These four types of that are caused by the communication network itself. Cor-
delay are important when considering the maximum message responding flow characteristics are seen from many of the
delay. The packet receiving delay, and the processing delay abnormal states, which include network jams and switch
are both in the range of a few microseconds and can be failure [38]. Calculations performed based on this scenario
considered to be approximately 3 µs [3], [37]. Additionally, allow for thresholds to be set in the protection being used to
the fixed arrival and service curves proposed in the network monitor network anomalies [38].
calculus theorem can be used in order to calculate the upper The method proposed by [38] requires three characteris-
bound of the transmission and queuing delay, which is also tics of the message source to be defined when developing
referred to as the maximum message delay [3]. As with the the model, which are, the number of each type of message
traffic load distribution, the GOOSE triggering relationship packet the source sends, the length of each message package
can be applied to the algorithm to calculate the maximum type, and the transmitting frequency of each message packet
message delay under certain conditions [3]. When finding type [38]. The format of the packets of each message type
the maximum message delay an assumption, that all triggered allows for two vectors to be formed, Fd∗1 and Sd∗1 . These
GOOSE flows burst at the same time, is made; however, this vectors consider the transmitting time interval and the packet
is not the case due to the network delay. This assumption leads length of each packet respectively. They are arranged accord-
to a longer delay than is seen in practical simulations, and as a ing to the sequential order of the message source in a digital
result, the calculation provides a conservative analysis of the substation in which the total packet types is d [38].
network performance [3]. Initially, [3] proposed the formation of the physical and
The analytical modeling method proposed by [3] also logical connection models, as well as the supply and service
allows for the consideration of the network performance models to find the traffic load distribution as well as the
under various scenarios. These scenarios are the minimum message delay. Reference [38] also defines the physical and
load scenario, the maximum load scenario, and the typical logical connection models and in doing so, defines a mes-
event scenario. sage transmission path model. This does not differ from the
The minimum load scenario occurs when the SCN is in method developed in [3]; there is merely a separation of the
steady-state, and the GOOSE flow has a fixed minimum data calculation carried out into an extra model. However, [38]
rate [3]. This scenario is the most common operating state for later defines a message transmission path matrix, as well
the SCN. The traffic load distribution in the minimum load as a message flow distribution model, and a message flow
VOLUME 8, 2020 204551

calculation flowchart, which differ from the work done in [3]. mentioned models. This theory has two key concepts, which
The method proposed by [38] allows for the message flow are the Stochastic Arrival Curve and the Stochastic Service
to be calculated; however, the method does not propose a Curve. From this, an arrival curve of traffic flow, and a service
way to find the message transmission time delay, which is curve of traffic flow can be developed for the SCN being
an important aspect in SCN design and modeling. investigated. GOOSE and SV messages are the main mes-
sage types considered in the arrival curve proposed by [39]
C. ANALYTICAL MODELING – STOCHASTIC NETWORK when calculating and analyzing the network’s communica-
CALCULUS AND TRAFFIC FLOW MODEL tion performance. The method proposed allows for the traffic
P. Xie et al. [39] propose another analytical model, from distribution in the SCN to be analyzed in various scenarios.
which the parameters of an SCN can be analyzed. The method These include steady-state conditions, conditions in which
uses the mapping relationship between the input and output there are primary faults, and finally, conditions in which there
nodes in order to describe the transmission process of traffic are secondary faults, which include low traffic load scenar-
flow in a branch. Additionally, the transmission performance ios and message content faults. Additionally, the analytical
of an SCN is analyzed with the use of a stochastic network model and network calculus theory proposed allows for the
calculus theory [39]. transmission delay of the traffic to be analyzed [39].
Various models are used to form the final analytical model
of the network. The first model is the network topology D. A BOUNDED MODEL OF COMMUNICATION DELAYS
model, in which the corresponding relationship between the IN SIPS AND NETWORK CALCULUS THEORY
nodes and branches is expressed using matrix A. Matrix A is C. Huang et al. [40] investigate a method that can be used
a Nn × (K + M ) matrix, where Nn is the total number of input to analyze the data latency of system integrity protection
nodes, K is the number of single input branches, and M is schemes (SIPS). The method investigated involves the use of
the number of multi-input branches [39]. The value of each a bounded model for communication delay. SIPS are required
element in matrix A is as follows: to be developed to allow for both time-critical and reliable
• aij = 1 when Ni is the input node of bj . operation, and as a result, the consideration of delays is essen-
• aij = −1 when Ni is the output node of bj . tial. Idealistic parameters are, in most cases, unfavourable
• aij = 0 in all other cases. for SIPS design considerations [40]. The research scholars
The arrival curve of the input traffic flow, as well as the in [40] consider the division of system integrity protection
service curve that is provided by the nodes in the branch are into two categories based on different protection scopes,
used in order to determine the output of branch i [39]. which are wide-area protection and substation-area protec-
The second model proposed is an analytical traffic-flow tion [40]. The wide-area network considers the protection
distribution model. The traffic load of a specific layer or level of the power system against system-wide disturbances [40].
in the substation communication network can be modelled A bounded model is used in order to model the commu-
making use of matrix A previously defined, as well as the nication delay over regional and backbone networks in the
departure curves of the traffic flow, and the service curves wide-area protection network.
of the nodes [39]. However, various matrix operations are It is worth noting that the modeling of the substation-area
required in order to develop the model using the method protection discussed in the paper is of relevance to this review.
proposed by [39]. The researchers [40] make use of the network calculus theory
The third and final model developed in the proposed in order to model the SCN network traffic. The choice of
method is the analytical fault event model. This model inves- the network calculus theory for modeling was informed by
tigates three different matrix configurations depending on the the concept that the network calculus theory allows for focus
fault scenario which could occur. This first scenario is a break on performance guarantees, rather than the average values
in the path of traffic flow transmission. When this scenario that are dealt with in classical queuing theory [40]. The
occurs, the output node of the link can no longer receive data research scholars in [40] use the arrival and service curves,
from the sending node. As a result of this, the -1 element the fundamental aspects of the network calculus theorem,
in matrix A (defined in the network topology model) corre- in order to determine the delay and back-log bound of the
sponding to the output on which the fault has occurred, is no network traffic. For this, the constraints on the arrival and
longer included [39]. The second fault condition considered is service processes, considering the average traffic rate, maxi-
the dislocation of the branch output node. This fault scenario mum instantaneous burst, minimum service rate and latency
occurs when the output of a branch no longer corresponds to parameter, are used [40].
the correct output node. Matrix A is adjusted to represent the The paper uses the network calculus theorem in order to
fault conditions that occur in this scenario [39]. The final fault propose a method that allows for delay bound analysis to
scenario accounted for in the model is a fault that occurs due be performed on SCNs in which priority-based queuing is
to transmission data mistakes. The output of the branches are used in order to schedule different types of messages in the
adjusted using a vector εk to represent this fault scenario [39]. network [40]. The researchers suggest that various results
Reference [39] also proposed the use of the stochastic or properties can be obtained from the network calculus
network calculus theory in conjunction with the previously theory employed, from which delay bound analysis can be
204552 VOLUME 8, 2020

performed. The properties that are readily available from the The modeling of the MU IEDs proposed by [7] is based
network calculus theory are as follows [40]: on IEC 61850-9-1. The default destination address is deter-
1. The superposition property. mined by the Ethernet broadcast address; however, the model
2. The leftover service property under priority scheduling. does support both unicast and multicast transmission. The
3. The leftover service property under FIFO scheduling. packet size, sample rate, address, start time, stop time and
4. The output property. multicast group address can be specified to configure the
5. The concatenation property. model of the MU IED proposed by [7]. The communi-
The delay and back-log service guarantee analysis property cation stack of the proposed MU IED model contains an
is also essential for delay bound analysis of the SCN traffic. application layer, Ethernet layer and physical layer. The
As would be anticipated, the arrival and services curves of various layers allow for the configuration of the model to
GOOSE and SV messages are given the highest priorities meet the requirements of the system as well as the various
in the model developed [40]. Additionally, the lengths and standards [7]. The use of the bus or star topology in the
frequencies of the messages, as well as the switch port model allows for different methods of connection between
rate, are also considered in the proposed method [40]. The the MU IEDs and the P&C IEDs. The process bus is used
model allows for SV message delays in a substation, as well to connect the MU IEDs to the P&C IEDs in the bus
as GOOSE message delays between two substations to be topology [7].
analyzed under different network traffic loads [40]. The The circuit breaker IED has important functionalities in the
research scholars in [40] see the use of the network calculus SCN, which are the receiving of the trip messages, the send-
theorem which provides bounded analysis as an essential ing of a GOOSE event to the other protection IEDs, as well
method, due to the requirement that all time-critical messages as the station PC, and the calculation of the end-to-end (ETE)
must be delivered within the maximum allowable message delay, [7]. The ETE delay can be defined as the amount of
delays [40]. delay between the creation of the message, which is at the
application layer of the source IED, and the arrival of the
E. OPNET MODELING OF IEDS message at the receiving IED’s application layer. As with
T. Sidhu et al. [7] proposes a method of modeling IEC the MU IED model, the GOOSE message package size, the
61850-based IEDs and setting up a platform for SCN per- address, and the transmission type can be configured in the
formance studies using the optimized network engineering circuit breaker IED model [7]. The GOOSE message is sent
tool (OPNET) modeler. The creation of IED models allows automatically when a trip or control command is received,
for the specific characteristics of a SAS network to be rep- and as a result, it is not necessary to set the start time of the
resented and modeled. As mentioned previously, the general message. As GOOSE messages are time-critical, the prior-
structure of an SCN in a SAS has three main levels, two of ity of the messages must be considered when creating the
which are the station and the process levels [7]. The research model [7]. Additionally, the fact that the breaker IED has to
scholars in [7] propose the modeling of three different types exchange messages with both the protection IEDs, as well
of IEDs that are seen in the general structure of an SCN as the station PC must also be taken into account, meaning
described previously, which are merging unit IEDs, breaker the breaker IED must be modeled to support client-server
IEDs, and combined protection and control (P&C) IEDs. The communication [7].
IEC 61850 guideline specified communication stack is used The modeling of the P&C IED proposed using the OPNET
in order to develop the model proposed by reference [7]. This modeler is very similar to that of the breaker IED. How-
involves the classification of messages into seven categories, ever, it should be noted that when connected in the bus
following which, the performance requirements of the mes- topology, the P&C IED model must have two bus commu-
sages are used in order to map the messages into different nication ports as it must communicate with both the pro-
communication stacks [7]. cess and station bus. Two ports are also required in the
An object-oriented modeling approach is used by the star topology [7]. Background traffic flow to the station PC
OPNET modeler. As previously discussed, IEDs and switches or server can also be generated by the P&C IED model,
are examples of network devices in an SCN and are called if it is configured to provide this function. In fault mode,
node models when modeled using the OPNET modeler. the IED model sends multicast trip messages at a speci-
Modules that are connected through packet streams or static fied time to the corresponding breaker IEDs [7]. The model
wires make up the node models, and each of the modules is also requires the address, destination address and multicast
assigned to a process module. This configuration allows for group address to be configured for the model to operate
the required behaviours to be achieved [7], [21]. Addition- correctly.
ally, a finite state machine (FSM) approach is used by the With the creation and configuration of the IED models in
OPNET process model, which allows for the implementation OPNET, the software allows the SAS communication net-
of resources, protocols, algorithms, applications, and queuing work being researched to be modeled. The research carried
policies [7], [21]. The states and transitions of the FSM are out by [7] focuses on the ETE delay for critical messages.
used to graphically define the progression of a process in The ETE delay is a critical statistic when evaluating the
response to events. performance of the SCN. Reference [7] shows that the ETE
VOLUME 8, 2020 204553

delays can be found using different operating conditions, delay must be estimated when assessing the performance of
including different LAN speeds, different sampling rates and, the process bus. This propagation delay is usually less than
priority tagging. 20 ns for cables with a length of less than 20 m. The error in
The research scholars in [21] also proposed the OPNET the frame arrival time-stamps of the Ethernet card, as well
simulation software as a method of modeling a substation as the absolute time error, must also be considered when
communication system in order to analyze the performance. conducting the experimental testing [41]. Two experiments
Paper [21] reviews various different architectures that can be were proposed using the method. In the first experiment
employed in a substation automation system, which include proposed, three ports on the Ethernet capture card were used
the star architecture, the ring architecture and various adapta- in order to directly capture SV frames from three merging
tions of the star and ring architectures, which add redundancy unit cards. With one of the merging unit cards generating
and as a result, reliability [21]. Paper [21] also discusses the one pulse per second clock for the other merging unit
the object-oriented modeling approach that is used by the cards as well as the Ethernet capture card [41]. The second
OPNET modeling software, as well as the fact that the process experiment proposed used two Ethernet switches between
model in the modeler uses an FSM approach. Ultimately the the merging units and the Ethernet capture card, allowing
research scholars in [21] proposed a method in which to find the network latency under this condition to be found. The
the ETE time-delay of an SCN under different conditions, second experiment allows for a larger substation network to
which included various LAN speeds, priority tagging, and be represented [41].
different VLAN sampling rates.
The research provided in paper [1] also proposed a method G. HARDWARE AND EXPERIMENTAL MODELING
to find the ETE time-delay of a specific SCN architecture Reference [37] proposes the use of a process bus test-bed
using the OPNET modeler in order to analyze the perfor- to create a hardware/experimental model for the process
mance of the SCN architecture. The research is carried out bus of an SCN, which differs from the analytical and
by the same authors that proposed the work in paper [21]; simulation-based models previously discussed. Although
however, [1] investigates additional aspects in the design of simulation models allow for much larger substation com-
an SCN architecture, which include, redundant paths, pro- munication networks to be modelled, [37] suggests that
vision for local data concentrators, no single point of fail- the quality of the simulation models largely influence the
ure, the absence of gateways for inter IED communication, results. Detailed-event based models may not be available
enhanced reliability and finally, cost. With a design carried for use in simulation software as industrial Ethernet switches
out focused on these aspects, the SCN architecture proposed rated for substation use have not been widely used in
by [1] was simulated using the OPNET modeler, focusing the past [37].
on the ETE delay performance of the proposed network. The The process bus model formed includes the exchange of
performance analysis under various configurations was con- the measured transformer data between the process and bay
sidered, including different transmission rates, normal and levels, as well as the exchange of control data between the
heavy background network traffic, normal and fault condi- process and bay levels in the SCN [37]. A very similar method
tions, priority tagging and different VLAN configurations [1]. to that proposed in [41] was used to find the latency of
These conditions include those used in [21], as well as addi- frames. This method again involved the use of an Ethernet
tional ones in order to review the performance of the SCN capture card; however, an Ethernet tap was also used. The
architecture designed. Ethernet tap is placed between the message source and the
first Ethernet switch. As a result of this arrangement, any
F. USING ETHERNET CARDS TO MEASURE PROCESS source of Ethernet traffic can be used, and the frame latency
BUS PERFORMANCE is the time difference between the frame being received from
As mentioned previously, it is essential to analyze the per- the tap, and the frame received from the switch. The setup
formance of an SCN in order for the successful design proposed is shown in Fig. 9.
and implementation to be carried out. The dynamic per-
formance of the communication in the SAS can be deter-
mined through modeling, allowing for any issues that may
occur to be found and rectified. Reference [41] proposed a
different method, using experimental modeling to find the
latency introduced by Ethernet switches in the sample value
process bus.
The method involves the use of an Ethernet capture card FIGURE 9. Measurement of the frame latency [37].
to measure the delay of SV messages. The sampling of the
merging unit, as well as the time-stamping unit in the Ethernet Additionally, [37] proposes the use of a much larger test
capture card are synchronized by the same one pulse per sec- network than was seen in [41]. Reference [37] also proposes
ond source [41]. As the synchronizing signal is transported hardware/experimental model setups that allow for interac-
through cables and media or level converters, the propagation tion testing to be performed. The testing of these models
204554 VOLUME 8, 2020

considered two types of interaction, the effect that The Box-Jenkins methodology is applied to the ARMA mod-
high-volume SV traffic has on management and GOOSE eling in five steps, namely [42]:
signalling, as well as the complement, the effect that man- 1. Data preprocessing.
agement and GOOSE messages have on the delivery of SV 2. Model-identification.
messages [37]. 3. Estimation of the model parameters.
4. Diagnostic verification of the model.
H. TIME-SERIES MODELING
5. Possibility forecasting.
R. Feizimirkhani et al. [42] discuss a manner in which
GOOSE messages between IEDS in smart grids can be The research scholars in [42] progress to discuss the
modeled, making use of a time-series modeling method. application of the ARMA model proposed to the specific
The paper, mentioning the randomness that exists in data case study described. This deals with the implementation
behaviour in power grids, indicates that a stochastic or of the ARMA model over the data traffic obtained from
probability model is useful for the representation of such the co-simulation. The application of the ARMA model is
data [42]. As a result, data-driven Auto Regressive Mov- discussed in five steps [42]:
ing Average (ARMA) models can be used as a time-series 1. Data preparation – The ARMA model must only be
method for characterization and modeling of the traffic [42]. applied to stationary time series, which are assumed to be
The research conducted is discussed through a case study in a specific form of statistical equilibrium. The statistical
to which the modeling method is applied. The case study equilibrium implies that the time series vary over time in
scenario utilised is of some importance as the reactive power a stable manner about a fixed mean. As non-stationary
absorbed or injected into the grid can be independently characteristics are almost always present in practi-
controlled by different nodes, with a centralized supervisor cal time series, the non-stationary behaviour must
not required [42]. In a similar manner, the voltage and/ or be removed [42]. The paper discusses various meth-
reactive power can be controlled within specified boundaries ods (tools) which can be used for this purpose, including
in a smart grid, making use of distributed and cooperative detrending, dissimilarity and deseasonality [42]. These
algorithms [42], [43]. As a result of the benefits of the tools remove characteristics such as dependence on time,
voltage and reactive power control, the implementation of long-range dependence (LRD) and seasonal patterns [42].
the algorithms required is expected to continue. As a result, 2. Model Identification – The estimated Auto Correlation
the communication network and connected devices must be Function (ACF) and estimated Partial Auto Correlation
able to adapt, allowing for the implementation of the algo- Function (PACF) can be used to measure the statistical
rithms. The case study discussed by the research scholars [42] relationships that exist within a series, allowing for the
consists of two renewable energy generating units (used in correct type of ARMA model with the best orders to
order to compensate the reactive power of a load unit), be determined [42]. Identification of the most applicable
and each unit has an IED, which communicates through a model and orders is important as the application of these
VLAN. Additionally, MATLAB/Simulink is utilised in order provides a stochastic data model, which has a simple
to form a co-simulation platform allowing for the electrical polynomial form [42].
and communication grid to be modeled. This co-simulation 3. Parameter estimation – The estimated model can be eval-
model allows for the reactive power variation data flow traf- uated for statistical adequacy using various diagnostic
fic, which is transmitted through IEC 61850 based GOOSE tests suggested in the Box-Jenkins method. If the esti-
messaging to be obtained [42], [44]. The paper focuses mated model is deemed unsuitable through the diagnostic
on the modeling of the variation in the number of data tests performed, the identification and estimation stages
packets which flows through the communication channel are repeated. The cycle is repeated until a model per-
over time [42]. forms suitably in the diagnostic tests run, as the future
A sample realization of n observed samples, measured behaviours of the time series being analysed can be fore-
over equal time-spaces, can be measured from an infinite cast through the use of a satisfactory model [42].
population of the samples in question, y1 , y2 , . . . , yn using 4. Model Implementation – The model implementation stage
time-series modeling [42]. The modeling scenario proposed, used by the researchers [42] allows for the ARMA model
which allows the SCN data flow to be modeled, considers proposed to be applied to the sampled data over the simu-
the cumulative number of packets which are passed between lation network developed. An important step in the imple-
the PV and load unit IEDs in the case study used. The linear mentation process is the determination of the correct sam-
relationship which exists through current values, historical pling period [42]. Fig 10. shows a flowchart of the ARMA
data and exogenous factors can be investigated by making use modeling procedure proposed by the research scholars
of time-series modeling and analysing methods. Exogenous in [42]. The figure provides a graphical representation of
factors refer to external factors, factors that are outside of the the first four steps in the application of the Box-Jenkins
system. The Box-Jenkins model, a widely used method for model that have been discussed.
mathematical representation of time series analysis, is applied 5. Mathematical expression – The originally measured traffic
to the ARMA model proposed by the researchers [42]. can be reconstructed through the use of the ARMA model.
VOLUME 8, 2020 204555

of the SCN data are essential and are utilised in the model
developed [45]–[47]. SCN data traffic can be considered with
the following characteristics [45], [48]:
1. Self-similarity and long-range dependence.
2. Smoothness.
3. Multi-fractal.
4. Periodic.
5. Irregularity.
The Fractional Auto Regressive Integrated Moving Aver-
age (FARIMA) model proposed by the research scholars
in [45], is an alternative model structure to the ARMA model
used by the researchers in [42]. As a result, similarities
between the methods can be noted, such as the use of the
Auto Correlation and Partial Auto Correlation Functions.
Additionally, the Hurst parameter is found, which can be used
to indicate the degree of auto correlation [45]. The paper
briefly proposes the formulation of an optimized FARIMA
model, as well as a threshold model for different confidence
levels. A threshold model for the communication network
traffic flow can be developed under different confidence lev-
els with the use of different experimental data obtained from
the communication network in an intelligent substation [45].
With the development of the final threshold model, SCN
data traffic flow under normal operating conditions can be
characterized [45].
FIGURE 10. Flowchart of the ARMA modeling procedure [42].

J. RIVERBED (OPNET) MODELING OF SAS COMPONENTS
AS WELL AS A METHOD FOR EXPERIMENTAL VALIDATION
This is done through the application of the inverse opera- OF RESULTS
tions, to those that were applied to the originally measured T. Ustun et al. [49] propose a method for both modeling and
traffic [42]. experimental validation of the results obtained, for the per-
Finally, the research scholars in [42] also investigate the formance analysis of different types of messages utilised in
effect that three main communication channel parameters; an active distribution system (ADS) substation. The different
packet loss, channel speed and cable length, have on the types of messages considered are defined by the IEC 61850-5
traffic model. Various values were used for each parame- standard, and are as follows [49]:
ter, allowing for an investigation into the sensitivity of the 1. Type 1 – fast messages
ARMA models to variations in some of the communication 2. Type 2 – medium speed messages.
network parameters [42]. The number of packets with respect 3. Type 3 and 7 messages are grouped, which include low
to time in the communication network is presented in the speed and command messages respectively.
paper, giving an indication of traffic flow and traffic load 4. Type 4 – raw data messages.
distribution [42]. 5. Type 5 – file transfer messages.
6. Type 6 – time synchronization messages.
I. FARIMA BASED THRESHOLD MODEL FOR DATA An in-depth description of the communication links and
CHARACTERIZATION data transfer between devices in a SAS is provided by the
W. Hao and Q. Yang [45] propose a method of characterizing research scholars in [49]. Additionally, a software approach
data traffic in intelligent SCNs with the use of a FARIMA to the modeling of the traffic flow in an SCN is proposed,
based threshold model, based on real data measurements. with the use of the Riverbed Modeler being discussed in
Additionally, the paper [45] aims to predict the data traffic the paper [49]. The Riverbed Modeler, which was previously
behaviour with the bounding thresholds against time, with the OPENT modeler, allows for the modeling of the various
the use of a filtering algorithm. Initially, a brief review of components contained in the SAS, including the P&C IEDs,
intelligent substations based on the IEC 61850 standards, MU IEDs, breaker IEDs, Ethernet switches and fiber optic
as well as a discussion of the characteristics of SCN data traf- links. The simulation requires that appropriate nodes are
fic is provided. As SCNs are required for the intelligent func- selected to model the IEDs being investigated. As a result,
tionality of electrical substations, and diverse communication a node which accommodates the communication stack for
patterns are exhibited by SCN data traffic, the characteristics MMS based messages is used for the modeling of MU IEDs,
204556 VOLUME 8, 2020

and a node which has dual capability of two communication nature of cyclic data allows it to be mathematically modeled
stacks is utilised for the P&C IED as well as the breaker considering the size of the data, the amount of cyclic data
IED [49]. This is due to the fact that the P&C and breaker which arrives per unit time (the numeric equivalent to the
IEDs need to be modeled allowing for relation to both the sampling frequency of the IEDs), and the time delay of the
publishing and subscribing of GOOSE messages, as well as messages [50]. The time delay is considered to be made up
client-server type communication, the latter of which requires of the sum of the Ethernet delay, the pretreatment time of the
complete TCP/IP implementation [49]. sender and the postproceeding time of the receiver [50]. The
The various assumptions made in the development of the maximum delay times specified by the IEC 61850 standards
simulation model are also described by the research scholars. can be considered using the mathematical model.
The model of the Ethernet switch selected should be able to The stochastic data considered in the paper is divided into
support a full-duplex of communication links at the desired two types, which are classified as follows [50]:
speeds [49]. The software modeling method proposed allows
1. Type 1 – These messages include trip messages, switch
for the performance evaluation of the SCN with the use of the
operation messages and transformer tap modulation data.
ETE delay metric. This evaluation can be performed under
2. Type 2 – These messages include event log checking and
various network conditions, which include the use of VLANs
protection setting modification data.
and a priority tagging scheme. The difference in results under
the various conditions can be analysed to determine the best The research scholars in [50] propose the modeling of the
configuration for the SCN [49]. arrival of stochastic data traffic with the use of the Poisson
Additionally, the research scholars in [49] also propose a process. The Poisson process is chosen as the packet data
method to produce a hardware replication of a real substation is generated in a random time period with a probability P.
bay in a laboratory. In this setup, the Ostinato software is The amount of packets arriving in two mutually exclusive
used in order to generate a fixed amount of background traffic time periods is independent as packets which arrive one after
and achieve realistic results from the experimental setup [49]. another have no correlation. Additionally, packet size can
As in the experimental setup proposed in [37], the use of a vary with time, or it could be fixed [50]. The nature of the
network (Ethernet) tap is proposed in [49], allowing for the arrival of the stochastic data indicates the suitability of the
duplication and monitoring of signals. Wireshark, a packet Poisson process.
capture software is utilised to capture GOOSE messages in Burst data, during a random time, is dependent on events
the breaker IED. Additionally, the researchers [49] propose that have previously occurred and are also generated with
the further use of software in order to replicate the behaviour a probability λ. When a substation event (fault) occurs,
of P&C and breaker IEDs on computer systems in the labo- the messages change to burst mode, from the original cyclic
ratory. With the correct hardware configuration, performance mode [50]. This change results in burst data flow. The gen-
evaluation of both VLAN and priority tagging configurations eration of burst data causes a large amount of network traf-
can be performed [49]. fic in a short space of time, leaving the network free for
long periods once the transmission of the data packets is
K. MATHEMATICAL AND OPNET SIMULATION BASED complete. As a result of this nature, burst data flow can
MODELING be said to have long-range dependence and self-similarity
Z. Zhang et al. [50] acknowledge the IEC 61850-5 classi- characteristics, presenting the same burstiness at different
fication of the seven different message types; however, for time scales [50], [51]. The research scholars discuss that
the purpose of the modeling method proposed, the reachers self-similarity of data flow in a network can described with
classify substation messages into three types, which are cyclic the use of the ON/OFF model and heavy-tailed distribu-
data, stochastic data and burst data. The data types are classi- tion. The ON/OFF model allows for the representation of
fied as follows [50]: data sources whose state repeatedly change between sending
messages and not sending messages. In the ON state of the
• Cyclic data flow is generated through the periodic sam- model, it is assumed that data is generated at a constant rate,
pling of current and voltage transformers. whereas it is assumed that no data is generated in the OFF
• Stochastic data is data which is generated through state [50], [52], [53]. Due to the described nature of burst
substation events, such as trip messages. It is typical traffic flow, the ON/OFF state model and Pareto distribution
event-driven data. are used for the mathematical model proposed. The amount
• Burst data contains information which belongs to of data for different types of messages can be found using the
GOOSE messages, such as information regarding pro- models, constituting the analysis of data flow for a typical
tection actions and changing the status of breakers. substation presented by the researchers [50]. In addition to
Cyclic data, which includes high priority SV messages, the mathematical models presented, the paper also provides
as well as medium speed GOOSE messages, is typical modeling with the use of the OPNET simulator, which inves-
time-driven data that is of a fixed length. This indicates that tigates Ethernet delays under different network configura-
the packet size can be decided in advance and that the mes- tions (with and without VLANs), as well as different SCN
sages are triggered at the same time [50]. The time-dependent architectures [50].
VOLUME 8, 2020 204557

L. ADDITIONAL MODELING METHODS PROPOSED adapted and simplified, allowing for an additional scenario
IN LITERATURE to be added, which is an abnormal flow scenario. However,
The research scholars in [54] again use the OPNET modeler; although the adapted network calculus and path flow method
however, one of their main goals was to determine whether is very similar to the original method, the simplifications
the real-time demands of a SAS can be met using Ether- and adaptions mean that the maximum message delay can
net. This entails the investigation of Ethernet’s performance no longer be solved for, which is an essential parameter in
characteristics in a substation communication network. Prior communication network modeling. Lastly, the stochastic net-
to the simulation, reference [54] investigated various issues work calculus and traffic flow model method differs slightly
that involved the simulation of communication networks and from the first two analytical methods reviewed. Although one
the measurement of Ethernet’s performance. These issues of the main aims of the method is also to find the traffic
were protocol, disturbances, performance requirements, sub- load distribution and maximum message delay, the method
station topology and message rate, and switched Ethernet also considers the analysis of fault events that may occur in
characteristics. The research conducted using the OPNET the communication network. The fault events considered are
modeler found that the real-time demands of a SAS can breaks in the traffic path/flow of transmission, the dislocation
be met with the use of switched Ethernet, as it has suffi- of the branch output node, and transmission data mistakes.
cient performance requirements to successfully operate in an Additionally, instead of investigating the network parameters
SCN [54]. This was tested through the examination of Eth- under the minimum, maximum and typical event scenarios,
ernet’s capabilities as a common network, handling multiple the method investigates steady-state, as well as primary and
coexisting traffic types. This investigation found that various secondary fault conditions.
substation automation configurations could be successfully The simulation methods reviewed all made use of the
implemented with the use of a switch-based fast Ethernet OPNET modeler. Notably, besides the modeling of SCNs
network [54]. The use of UDP/IP is also investigated in the for design and analysis, the OPNET modeler can also be
research carried out in [54]. used in order to confirm the suitability of Ethernet to the
communication network. One of the main tasks in the sim-
ulation of the communication network using the OPNET
VI. DISCUSSION AND COMPARISON OF THE METHODS modeler is the creation of models for the IEDs in the network.
REVIEWED The software uses an object-oriented modeling approach, and
The analytical modeling methods reviewed in this paper all an FSM approach is used by the process model, allowing
use linear algebra (matrix operations) and the network cal- for the implementation of resources, protocols, algorithms,
culus theorem in order to model substation communication applications and queuing policies. In general, when creating
networks. The main findings of the methods reviewed are an IED model, the packet size, sample rate, address, start
the traffic load distribution, and the maximum message delay. time, stop time and multicast group address must be specified
These network parameters are discussed in section IV of this to configure the model of the IED. Ultimately, the results
paper and are essential to the successful operation of the com- of the simulation model focus on the ETE delay for critical
munication network. In general, when analytical modeling messages (the maximum message delay, as discussed in the
methods are considered, various sub-models are developed analytical modeling). However, the operating conditions in
using the arrival and service curve models for traffic flow in which the ETE delay can be found vary extensively from the
the network. These are used in all of the analytical modeling analytical methods reviewed. The different operating condi-
methods reviewed and form part of the network calculus tions in which the SCN can be simulated include, different
theorem. The largest differences between the methods are the LAN speeds, different sampling rates, different transmission
parameters which the methods are able to model, and under rates, normal and heavy background network traffic, normal
which conditions the parameters can be found. and fault conditions, priority tagging and different VLAN
The network calculus and path flow model aims to create configurations. Additionally, the architecture of the SCN can
a source model which describes the triggering relationships be easily adjusted.
between GOOSE flows, the message rate, message length, Experimental modeling was reviewed as a method to deter-
the data rate of traffic flows and correspondence between mine the essential performance parameters of an SCN. The
a port and the messages. Additionally, the service model main aim of the experimental modeling was to find the
developed in the method takes into account the queuing and latency introduced by Ethernet switches in the SCN. This
scheduling policies in the network. These sub-models allow was done through the use of Ethernet capture cards; however,
for the traffic load distribution, as well as the maximum the error in the frame arrival time-stamps, as well as the
message delay to be found in three scenarios, namely the absolute time error must be taken into account when carrying
minimum load scenario, the maximum load scenario, and out the experimental modeling. The use of an Ethernet tap in
the typical event scenario. However, the method assumes conjunction with an Ethernet card allows for any source of
that all GOOSE flows burst at the same time, providing Ethernet traffic to be used, and the method allows for interac-
a conservative, and likely higher maximum message delay tion testing to be performed. The interaction tests include the
than seen in simulation, and practically. This method can be effect that high-volume SV traffic has on management and
204558 VOLUME 8, 2020

TABLE 1. Modeling methods and applications. TABLE 1. (Continued) Modeling methods and applications.
Lastly, time-series based modeling methods were also

reviewed as a method in which traffic flow in a substation
communication network can be modeled. These make use
of the ARMA and FARIMA models to model traffic flow.
However, the development of such models is an iterative
process if the original model does not meet certain diagnostic
checks. Additionally, experimental or simulated data is also
required for use in these methods. The amount of data (num-
ber of packets) versus time is presented using this method,
providing an indication of traffic flow and traffic load in the
communication network.
GOOSE signalling, as well as the complement, the effect that All four types of modelling methods have certain advan-
management and GOOSE messages have on the delivery of tages; however, certain methods are able to provide addi-
SV messages. tional parameter identification in additional scenarios.
VOLUME 8, 2020 204559

TABLE 2. Modeling methods and advantages and disadvantages. TABLE 2. (Continued) Modeling methods and advantages and
disadvantages.
204560 VOLUME 8, 2020

TABLE 2. (Continued) Modeling methods and advantages and These methods include analytical, simulation, experimental,
disadvantages.
and time-series based modeling methods. Simulation-based
modeling was found to be a very useful tool, as the end-
to-end message delay can be simulated in various network
conditions; however, analytical modeling is more useful for
representing the traffic load distribution.
This paper reviews the current methods that can be used
to model the traffic flow in a substation communication net-
work. However, this work can be used in order to develop
additional models or extend the current ones discussed. The
development of substation communication networks to pro-
duce fast and reliable substation automation systems is likely
to continue, and as a result, the further development of simple
ways in which the communication networks can be modeled
is of paramount importance.
The analytical modeling methods provide accurate models REFERENCES

that allow for the traffic load distribution to be analyzed, [1] M. S. Thomas and I. Ali, ‘‘Reliable, fast, and deterministic substation
which was not a focus in either the simulation or experi- communication network architecture and its performance simulation,’’
mental modeling methods. However, the analytical methods IEEE Trans. Power Del., vol. 25, no. 4, pp. 2364–2370, Oct. 2010.
[2] B. Falahati and E. Chua, ‘‘Failure modes in IEC 61850-enabled substa-
require various matrix and calculus operations, adding to their tion automation systems,’’ in Proc. IEEE/PES T&D, Dallas, TX, USA,
complexity. Simulation methods, using the OPNET modeler, May 2016, pp. 1–5.
are arguably simpler to implement; however, they only allow [3] Y. Zhang, Z. Cai, X. Li, and R. He, ‘‘Analytical modeling of traffic flow in
the substation communication network,’’ IEEE Trans. Power Del., vol. 30,
for the ETE delay to be found. Even so, the simulation no. 5, pp. 2119–2127, Oct. 2015.
modeling method is extremely beneficial, as a large array [4] K. Kaneda, S. Tamura, N. Fujiyama, Y. Arata, and H. Ito, ‘‘IEC61850
of network conditions can be analyzed. One of the major based substation automation system,’’ in Proc. Joint Int. Conf. Power Syst.
considerations of the OPNET simulation method should be Technol. IEEE Power India Conf., New Delhi, India, Oct. 2008, pp. 1–8.
[5] N. B. Sparks and A. K. Saha, ‘‘Applications of goose messaging for breaker
the accuracy of the models, as industrial Ethernet switches fail protection,’’ in Proc. SAUPEC, Johannesburg, South Africa, 2018,
rated for substation use have not been widely used in the pp. 301–306.
past, meaning detailed-event based models may not be avail- [6] G. Liao, K. Hopinson, J. Tang, L. Ding, and X. Wang, ‘‘A simulation study
on the Ethernet communication of a substation automation system based
able for use in the simulation software. While experimental on EPOCHS,’’ in Proc. Int. Conf. Power Syst. Technol., Chongqing, China,
modeling means the substation equipment can be physically Oct. 2006, pp. 1–7.
analyzed, there are limitations on the size of the network [7] T. S. Sidhu and Y. Yin, ‘‘Modelling and simulation for performance evalua-
tion of IEC61850-based substation communication systems,’’ IEEE Trans.
that can be modeled easily. Additionally, experimental mod- Power Del., vol. 22, no. 3, pp. 1482–1489, Jul. 2007.
elling is largely dependent on the accuracy of the equipment [8] M. A. Aftab, S. Hussain, I. Ali, and T. S. Ustun, ‘‘IEC 61850 based
used, and as a result, can be an expensive method to use. substation automation system: A survey,’’ Int. J. Elect. Power Energy Syst.,
Lastly, time-series based modeling allows for the traffic load vol. 120, no. 106008, pp. 1–16, Mar. 2020.
[9] M. Aftab, S. Hussain, and I. Ali, ‘‘ICT technologies, standards and pro-
distribution to be analyzed; however, the possible iterative tocols for active distribution network automation and management,’’ in
nature of the process could make it cumbersome. Table 1 and Advanced Communication and Control Methods for Future Smartgrids.
Table 2 provide summaries of all the modeling methods, London, U.K.: IntechOpen, Nov. 2019, pp. 1–16.
[10] V. C. Mathebula and A. K. Saha, ‘‘Mission critical safety functions in IEC-
displaying applications, key results and the advantages and 61850 based substation automation System–A reliability review,’’ Int. J.
disadvantages of the methods. Eng. Res. Afr., vol. 48, pp. 149–161, May 2020.
[11] I. Xyngi and M. Popov, ‘‘IEC61850 overview–where protection meets
communication,’’ in Proc. 10th IET Int. Conf. Develop. Power Syst. Pro-
VII. CONCLUSION
tection (DPSP) Manage. Change, Manchester, U.K., 2010, pp. 1–5.
The objective of this paper was to review the methods in [12] S. M. Suhail Hussain, M. A. Aftab, and I. Ali, ‘‘A novel PRP based deter-
which the traffic flow in a substation communication network ministic, redundant and resilient IEC 61850 substation communication
can be modeled. Due to the development of software-based architecture,’’ Perspect. Sci., vol. 8, pp. 747–750, Sep. 2016.
[13] K. Brand, M. Ostertag, and W. Wimmer, ‘‘Safety related, distributed func-
substation automation systems, microprocessor-based intel- tions in substations and the standard IEC 61850,’’ in Proc. IEEE Bologna
ligent electronic devices and the introduction of the IEC Power Tech Conf., Bologna, Italy, Jun. 2003, p. 5.
61850 standard for communication systems, Ethernet com- [14] S. Roostaee, R. Hooshmand, and M. Ataei, ‘‘Substation automation sys-
tem using IEC 61850,’’ in Proc. PEOCO, Shah Alam, Malaysia, 2011,
munication systems are being more commonly used in sub- pp. 393–397.
station communication networks, leading to an increased [15] R. E. Mackiewicz, ‘‘Overview of IEC 61850 and benefits,’’ in Proc. PES
dependency of the substation automation system on the com- TD, Atlanta, GA, USA, Oct./Nov. 2006, pp. 623–630.
munication network. The review carried out in this paper [16] B. E. M. Camachi, O. Chenaru, L. Ichim, and D. Popescu, ‘‘A practical
approach to IEC 61850 standard for automation, protection and control
found that there are various methods in which the traffic of substations,’’ in Proc. 9th Int. Conf. Electron., Comput. Artif. Intell.
flow in a substation communication network can be modeled. (ECAI), Targoviste, Romania, Jun. 2017, pp. 1–6.
VOLUME 8, 2020 204561

[17] F. Tilaro, B. Copy, and M. Gonzalez-Berges, ‘‘IEC 61850 industrial com- [40] C. Huang, F. Li, T. Ding, Y. Jiang, J. Guo, and Y. Liu, ‘‘A bounded model of
munication standards under test,’’ in Proc. ICALEPCS, San Francisco, CA, the communication delay for system integrity protection schemes,’’ IEEE
USA, 2013, pp. 427–430. Trans. Power Del., vol. 31, no. 4, pp. 1921–1933, Aug. 2016.
[18] ABB Review Special Report: IEC 61850, ABB Asea Brown Boveri Ltd., [41] D. M. E. Ingram, F. Steinhauser, C. Marinescu, R. R. Taylor, P. Schaub, and
Zürich, Switzerland, 2010. D. A. Campbell, ‘‘Direct evaluation of IEC 61850-9-2 process bus network
[19] A. Khavnekar, S. Wagh, and A. More, ‘‘Comparative analysis of IEC 61850 performance,’’ IEEE Trans. Smart Grid, vol. 3, no. 4, pp. 1853–1854,
edition-I and II standards for substation automation,’’ in Proc. IEEE Int. Dec. 2012.
Conf. Comput. Intell. Comput. Res. (ICCIC), Madurai, India, Dec. 2015, [42] R. Feizimirkhani, A. I. Bratcu, and Y. Bésanger, ‘‘Time-series modelling of
pp. 1–6. IEC 61850 GOOSE communication traffic between IEDs in smart grids—
[20] S. Mohagheghi, J. Stoupis, and Z. Wang, ‘‘Communication protocols and A parametric analysis,’’ IFAC-PapersOnLine, vol. 51, no. 28, pp. 444–449,
networks for power systems-current status and future trends,’’ in Proc. 2018.
IEEE/PES Power Syst. Conf. Expo., Washington, DC, USA, Mar. 2009, [43] M. Manbachi, M. Nasri, B. Shahabi, H. Farhangi, A. Palizban,
pp. 1–9. S. Arzanpour, M. Moallem, and D. C. Lee, ‘‘Real-time adaptive VVO/CVR
[21] I. Ali and M. S. Thomas, ‘‘Substation communication networks architec- topology using multi-agent system and IEC 61850-based communica-
ture,’’ in Proc. Joint Int. Conf. Power Syst. Technol. IEEE Power India tion protocol,’’ IEEE Trans. Sustain. Energy, vol. 5, no. 2, pp. 587–597,
Conf., New Delhi, India, Oct. 2008, pp. 1–8. Apr. 2014.
[22] J.-C. Tan and W. Luan, ‘‘IEC 61850 based substation automation system [44] T. Peirelinck, A. I. Bratcu, and Y. Besanger, ‘‘Impact of IEC 61850 GOOSE
architecture design,’’ in Proc. IEEE Power Energy Soc. Gen. Meeting, communication quality on decentralized reactive power control in smart
Detroit, MI, USA, Jul. 2011, pp. 1–6. distribution grids—A co-simulation study,’’ in Proc. IEEE Electr. Power
[23] X. Yang, N. Das, and S. Islam, ‘‘Analysis of IEC 61850 for a reliable Energy Conf. (EPEC), Ottawa, ON, Canada, Oct. 2016, pp. 1–6.
communication system between substations,’’ in Proc. Australas. Univ. [45] W. Hao and Q. Yang, ‘‘Data traffic characterization in intelligent electric
Power Eng. Conf. (AUPEC), Hobart, Tas, Australia, Sep. 2013, pp. 1–6. substations using FARIMA based threshold model,’’ Energy Procedia,
[24] M. R. D. Zadech, T. S. Sidhu, and A. Klimek, ‘‘Suitability analysis of vol. 145, pp. 413–420, Jul. 2018.
practical directional algorithms for use in directional comparison bus [46] R. Bulbul, P. Sapkota, C.-W. Ten, L. Wang, and A. Ginter, ‘‘Intrusion eval-
protection based on IEC61850 process bus,’’ IET Gen., Transmiss. Distrib., uation of communication network architectures for power substations,’’
vol. 5, no. 2, pp. 199–208, Feb. 2011. IEEE Trans. Power Del., vol. 30, no. 3, pp. 1372–1382, Jun. 2015.
[25] N. Das and S. Islam, ‘‘Analysis of power system communication architec- [47] B. Vaidya, D. Makrakis, and H. T. Mouftah, ‘‘Authentication and authoriza-
tures between substations using IEC 61850,’’ in Proc. 5th Brunei Int. Conf. tion mechanisms for substation automation in smart grid network,’’ IEEE
Eng. Technol. (BICET), Brunei, Asia, 2014, pp. 1–6. Netw., vol. 27, no. 1, pp. 5–11, Jan. 2013.
[26] M. G. Kanabar and T. S. Sidhu, ‘‘Reliability and availability analysis of [48] G. Vaziri, F. Almasganj, and M. S. Jenabi, ‘‘On the fractal self-similarity
IEC 61850 based substation communication architectures,’’ in Proc. IEEE of laryngeal pathologies detection: The estimation of hurst parameter,’’
Power Energy Soc. Gen. Meeting, Calgary, AB, Canada, Jul. 2009, pp. 1–8. in Proc. Int. Conf. Technol. Appl. Biomed., Shenzhen, China, May 2008,
[27] N. Sparks and A. Saha, ‘‘Implementation of multi-protocol substation com- pp. 383–386.
munication and automation network,’’ in Proc. SAUPEC, Johannesburg, [49] T. S. Ustun, M. A. Aftab, I. Ali, and S. M. S. Hussain, ‘‘A novel scheme for
South Africa, 2018, pp. 277–282. performance evaluation of an IEC 61850-based active distribution system
[28] S. Gupta, ‘‘Reliability analysis of IEC 61850 substation communication substation,’’ IEEE Access, vol. 7, pp. 123893–123902, Sep. 2019.
network architectures,’’ Ad. Res. Elect. Electron. Eng., vol. 3, no. 2, [50] Z. Zhang, X. Huang, B. Keune, Y. Cao, and Y. Li, ‘‘Modeling and simula-
pp. 93–98, Jan./Mar. 2016. tion of data flow for VLAN-based communication in substations,’’ IEEE
[29] S. Gupta, ‘‘Performance analysis of substation communication network Syst. J., vol. 11, no. 4, pp. 2467–2478, Dec. 2017.
architectures in OPNET,’’ Int. J. Res. Ad. Eng. Tech., vol. 3, no. 2, [51] S. Song, J. K.-Y. Ng, and B. Tang, ‘‘Some results on the self-similarity
pp. 40–47, May 2017. property in communication networks,’’ IEEE Trans. Commun., vol. 52,
[30] N. Liu, M. Panteli, and P. A. Crossley, ‘‘Reliability evaluation of a sub- no. 10, pp. 1636–1642, Oct. 2004.
station automation system communication network based on IEC 61850,’’ [52] J. Zhang and T. Konstantopoulos, ‘‘Multiple-access interference processes
in Proc. 12th IET Int. Conf. Develop. Power Syst. Protection (DPSP), are self-similar in multimedia CDMA cellular networks,’’ IEEE Trans. Inf.
Copenhagen, Denmark, 2014, pp. 1–6. Theory, vol. 51, no. 3, pp. 1024–1038, Mar. 2005.
[31] S. Mnukwa and A. K. Saha, ‘‘Implementation of substation SCADA [53] T. Kim, D. Lee, and S. Lee, ‘‘Hierarchical multilevel on/off source traffic
and automation systems in the port of durban,’’ in Proc. IEEE PES/IAS modeling for a warship combat system,’’ IEEE J. Ocean. Eng., vol. 39,
PowerAfrica, Cape Town, South Africa, Jun. 2018, pp. 214–219. no. 2, pp. 226–242, Apr. 2014.
[32] S. Mnukwa and A. K. Saha, ‘‘SCADA and substation automation sys- [54] T. Skeie, S. Johannessen, and C. Brunner, ‘‘Ethernet in substation automa-
tems for the port of durban power supply upgrade,’’ in Proc. Int. tion,’’ IEEE Control Syst., vol. 22, no. 3, pp. 43–51, Jun. 2002.
SAUPEC/RobMech/PRASA Conf., Cape Town, South Africa, Jan. 2020,
pp. 1–5.
[33] E. Hyytia and J. Virtamo, ‘‘On traffic load distribution and load balancing MATTHEW LIAM DE KLERK was born in Glen-
in dense wireless multihop networks,’’ EURASIP J. Wireless Commun. wood, Durban, South Africa, in 1998. He is cur-
Netw., vol. 2007, pp. 1–15, Mar. 2007. rently pursuing the B.Sc. degree in electrical engi-
[34] A. Ingalalli, K. S. Silpa, and R. Gore, ‘‘SCD based IEC 61850 traffic esti- neering from the University of KwaZulu-Natal.
mation for substation automation networks,’’ in Proc. 22nd IEEE Int. Conf. His interests are related to power system protection
Emerg. Technol. Factory Autom. (ETFA), Limassol, Cyprus, Sep. 2017, and power electronics.
pp. 1–8.
[35] R. L. Cruz, ‘‘A calculus for network delay. I. Network elements in isola-
tion,’’ IEEE Trans. Inf. Theory, vol. 37, no. 1, pp. 114–131, Jan. 1991.
[36] R. L. Cruz, ‘‘A calculus for network delay, Part 11: Network analysis,’’
IEEE Trans. Inf. Theory, vol. 37, no. 1, pp. 132–141, Jan. 1991.
[37] D. M. E. Ingram, P. Schaub, R. R. Taylor, and D. A. Campbell, ‘‘Network
AKSHAY KUMAR SAHA (Member, IEEE) is cur-
interactions and performance of a multifunction IEC 61850 process bus,’’
rently an Associate Professor and an Academic
IEEE Trans. Ind. Electron., vol. 60, no. 12, pp. 5933–5942, Dec. 2013.
Research Leader with the School of Engineer-
[38] Y. Chen, J. Zhang, Q. Han, C. Guo, and Q. Zhang, ‘‘Message flow modeling
analysis and application for digital substation process layer network,’’ in ing, University of KwaZulu-Natal, Durban, South
Proc. IEEE 7th Annu. Int. Conf. CYBER Technol. Autom., Control, Intell. Africa. He is a registered Professional Engineer
Syst. (CYBER), Honolulu, HI, USA, Jul. 2017, pp. 1351–1355. with the Engineering Council of South Africa and
[39] P. Xie, Z. Cai, C. Guo, P. Liu, Z. Chen, Y. Zhang, Y. Han, and L. Li, Fellow of SAIEE. His research works are related
‘‘Traffic flow calculation method for substation communication network,’’ to advances in power systems.
in Proc. IEEE PES Innov. Smart Grid Technol. Conf. Eur. (ISGT-Europe),
Ljubljana, Slovenia, Oct. 2016, pp. 1–6.
204562 VOLUME 8, 2020

Review

Uploaded by

Copyright:

Available Formats

Review

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Review

Uploaded by

Copyright:

Available Formats

Received October 25, 2020, accepted November 6, 2020, date of publication November 10, 2020,

date of current version November 20, 2020.

A Review of the Methods Used to Model Traffic

ABBREVIATIONS Long-Range Dependence LRD

I. INTRODUCTION In another paper, M. Aftab et al. [9] review the communica-

204546 VOLUME 8, 2020

various benefits and features which the standard provides,

VOLUME 8, 2020 204547

FIGURE 3. Star type architecture - SCN [10], [21], [23], [30].

204548 VOLUME 8, 2020

FIGURE 5. Transmission time of a GOOSE message between IEDs [11],

VOLUME 8, 2020 204549

element C ij is equal to a 1 if there is a logical connection

204550 VOLUME 8, 2020

VOLUME 8, 2020 204551

204552 VOLUME 8, 2020

VOLUME 8, 2020 204553

204554 VOLUME 8, 2020

VOLUME 8, 2020 204555

FIGURE 10. Flowchart of the ARMA modeling procedure [42].

204556 VOLUME 8, 2020

VOLUME 8, 2020 204557

204558 VOLUME 8, 2020

Lastly, time-series based modeling methods were also

VOLUME 8, 2020 204559

204560 VOLUME 8, 2020

The analytical modeling methods provide accurate models REFERENCES

VOLUME 8, 2020 204561

204562 VOLUME 8, 2020

You might also like