See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/358561499

Towards Wi-Fi-based Time Sensitive Networking Using OMNeT++/NeSTiNg

Simulation Models

Conference Paper · December 2021

DOI: 10.1109/ICECET52533.2021.9698580

CITATIONS READS
6 157

3 authors:

Arnold B D Kinabo Joyce Mwangama

University of Cape Town University of Cape Town
3 PUBLICATIONS 21 CITATIONS 70 PUBLICATIONS 326 CITATIONS

SEE PROFILE SEE PROFILE

Albert A Lysko
Council for Scientific and Industrial Research, South Africa
130 PUBLICATIONS 657 CITATIONS

SEE PROFILE

All content following this page was uploaded by Albert A Lysko on 01 March 2022.

The user has requested enhancement of the downloaded file.

 Proc. of the International Conference on Electrical, Computer and Energy Technologies (ICECET)
9-10 December 2021, Cape Town-South Africa

Towards Wi-Fi-based Time Sensitive Networking

Using OMNeT++/NeSTiNg Simulation Models
Arnold B. D. Kinabo Joyce B. Mwangama Albert A. Lysko2,1
1 2
Department of Electrical Engineering Department of Electrical Engineering NextGen Enterprises and Institutions (NGEI)
University of Cape Town University of Cape Town Council for Scientific and Industrial Research
Cape Town, South Africa Cape Town, South Africa Pretoria, South Africa
[email protected] [email protected] [email protected]

Abstract—The Internet of Things (IoT) has varied applications growing of these sectors, and Industrial IoT (IIoT) is expanding
in industry, where its networks are uniquely characterised by a its products and services to fill this demand.
stringent requirement for reliability. Many technologies, such as
PROFINET and EtherCAT, arose to fill this need for reliable IIoT devices exhibit certain characteristics: precise
networks. Developed to run on standard Ethernet components, synchronisation and scheduling, for low latency; dedicated
most of these are proprietary in nature. A recent emergence is the bandwidth resources or their robust allocation; all of which
Time Sensitive Networking (TSN) standard, a collection of result in increased performance and reliability. Above all,
Ethernet standards by the IEEE TSN Task Group. TSN allows reliability is an indispensable requirement for mission-critical
communication between standard-conformant devices, regardless applications in IIoT [2]. More specifically, concerning devices
of the vendor. The resulting technology produces deterministic on the factory-floor, this is mainly provided through use of
networks with low jitter and latency. However (and as factory- standard Ethernet. Ethernet evolved from the office environment
floor layouts become more dynamic), wireless infrastructure is into the industry as a means to satisfy real-time determinism on
favourable compared to wired Ethernet in situations that require the plant level [3]. (Before its inception, industrial computer
increased mobility and range. Carrying over the determinism and networks comprised devices termed fieldbuses which were used
reliability of TSN from wired infrastructure to these wireless as a control platform on the factory floor). Apart from its ability
networks would be advantageous as many of these wireless to cater for the characteristic requirements mentioned, many
networks struggle to offer the same determinism as is present in proprietary fieldbus technologies were developed on Ethernet
TSN. There isn’t as yet any TSN-compliant wireless
because it was mass-produced and inexpensive. Using
implementation. This paper seeks to explore how TSN could be
carried out over the 802.11ac Wi-Fi standard. The paper also looks
fieldbuses based on common hardware, but with specially
at the factors that stand in the way of wireless TSN and develops designed protocols could save a company in excess of 50 %,
a TSN emulation model that tries to operate in Wi-Fi. The compared to the fully-proprietary older configurations that were
resultant model’s performance was studied in iterative in place, all the while maintaining real-time responsiveness
experiments to aid in analysing wireless TSN. The work has which was lacking in these older systems [3] – among the major
revealed that TSN traffic can be supported over an interference- motivations for the use of Ethernet.
free Wi-Fi channel, although the channel use efficiency of such a
IEEE 802.1Q Time Sensitive Networking (TSN) is a relative
network is limited to a few percent. Thus, Wi-Fi may require
newcomer among all these protocols. TSN is a set of Ethernet
further enhancements to its mode of operation to make it better
suited for time-sensitive applications in industry. standards that aim to provide deterministic messaging over the
physical and data link layers of the network [4]. Its main
Keywords—802.11ac, 802.1AS, critical traffic, Fourth Industrial advantage over other protocols and technologies stems from the
Revolution, 4IR, Industrial Internet of Things, IIoT, priority, fact that it is vendor-agnostic, as shall be discussed.
reliability, scheduling, synchronisation, Time Sensitive Networking, The commonality of all above-mentioned industrial
TSN, Wi-Fi.
communication technologies is that they run on Ethernet.
I. INTRODUCTION For IIoT devices, the wireless medium can provide:
The Internet of Things (IoT) is a fast-changing space with extended roaming capabilities for automated guided vehicles
various business use cases and sectors ensuring that more (AGVs); increased mobility for industrial robots; not to mention
devices are produced on a daily which have the ability to connect faster and potentially reduced costs for infrastructure
with each other over the Internet. Sources in [1] evidence that installation. With mobility in applications on the rise, there is a
the global market which stood at $235B in 2017, will reach gap to be addressed as far as determinism for non-wireline
$1.2T in 2022. The manufacturing industry is among the fastest communications is concerned. That is why wireless TSN is

978-1-6654-4231-2/21/$31.00 ©2021 IEEE

worth investigating. However, this opportunistic case of TSN 1) 802.1AS: This allows for tight synchronisation among
communications over a wireless medium is left largely devices. IEEE Standard 802.1AS [11] is based on, and includes
unaddressed as has been evidenced in [2], [5]–[7]. There are a profile of IEEE Std 1588-2008 [12], also termed IEEE 1588v2
several factors and reasons that deter its being employed in PTP (Precision Time Protocol), which is a protocol for
industry. Chiefly, the reliability suffers in a wireless network
synchronising clocks in a network’s nodes. It enables TSN
owing to the greater number of collisions from the many
wireless users. The collisions add to congestion, and this in turn, devices to have a shared understanding of time.
packet loss and longer [8], less deterministic latencies in 2) 802.1Qbv: IEEE Standard 802.1Qbv [13] introduces a
networks. In essence, without reliability, the aspect of stringent scheduling mechanism. Messages are placed into queues, each
timing in industry communications is non-existent. having different priorities. They can only transmit in their
scheduled time windows [14]. Transmission gates are used to
Given this challenge of deterministic timing in wireless
enforce this. The Qbv mechanisms can guarantee a maximum
networks, this paper seeks, firstly, to consider the issues that
hinder the realisation of TSN over a wireless network, and then, bounded latency for the network [14] and herein lies the
to develop an emulation model attempting TSN. Our goal is to property of determinism.
work towards availing the benefits of TSN to an otherwise non- 3) 802.1Qbu: TSN provides a frame pre-emption
deterministic, non-time-sensitive Wi-Fi infrastructure. mechanism through IEEE Standard 802.1Qbu [15]. This allows
a critical message to interrupt a non-critical one mid-flow. It
As such, this paper’s sections are organised thusly: Section
2 shortly describes determinism on the plant floor, and TSN with then allows the non-critical message queue to continue from
its defining features and mechanisms; Section 3 details a review exactly where it left off after it is done [14].
of some popular simulation frameworks and motivates for the Other standards include: 802.1Qcc [16] – Stream
choice in platform for the TSN emulation model; Section 4 Reservation Protocol (SRP) Enhancements and Performance
discusses briefly on factors that affect determinism in Wi-Fi; Improvements, and 802.1CB [17] – Frame Replication and
Section 5 presents the Wi-Fi based TSN prototype and outlines Elimination for Reliability (FRER).
some experiments; Section 6 discusses the results of these In the context of industry, a delay of 1 ms to 10 ms is suitable
experiments to provide a conclusion and the way forward on for most applications, as seen in the requirements in [18]. As for
wireless TSN. channel capacity, industrial applications do not require much
bandwidth. As per Bankov [6], the required channel capacity is
II. TSN AND DETERMINISM ON THE FACTORY FLOOR within 100 Kbps, which can be easily supplied by IE protocols
Machines in the industrial plant are constantly and fieldbuses.
communicating signals and messages to each other. Since it was
discovered that TCP/IP/UDP with the traditional fieldbus Where TSN gets around the other protocols is that it does not
configurations consisting of proprietary cables and IO card are face interoperability issues in the same way the proprietary
suboptimal for providing the deterministic real-time responses protocols do. Where the other IE protocols are not compatible
that these machines require, the industry turned to solutions that with one another in their present state [19], TSN can integrate
utilised CAT5 (Category 5 cable) Ethernet and network with and work alongside them, doing so through various
interface cards (NIC) [3]. As per Hibbard [3], as a result, many interfaces such as gateways and couplers. Not only does TSN
Industrial Ethernet (IE) technologies surfaced, which were offer future-proof real-time capabilities that exceed current IE
designed to use these common hardware in the physical layer in protocols [20], but since TSN is a standard, it can be applied by
combination with special protocols to provide the necessary different vendors so that the technologies can work hand in
determinism. In fact, there are some 30 IE technologies in hand.
existence, of which about five remain prominent: PROFINET, III. REVIEW OF SIMULATION FRAMEWORKS FOR TSN
EtherNet/IP, Sercos III, EtherCAT, and OPC UA. These five are
built on open standards, unlike many others which either require For our work, the OMNeT++ platform [21] was chosen.
proprietary vendor devices or their source code is not available. Other simulation programs are available; the most notable are
We have lent further discussion to these in [9]. NS-2 (and -3 is now available), and the Riverbed Modeler, more
popularly known as OPNET. (The OPNET modeler and OPNET
As mentioned earlier, there is a newer alternative to these Technologies in its entirety has since been acquired by Riverbed
traditional fieldbuses and technologies, described in the next Technologies, so it is not easily found by the same name,
subsection. although the name OPNET is still widely used and understood).
A. Time Sensitive Networking (TSN) Many others exist, but these are often more specialised in their
function and at the time of this review there weren’t any
IEEE 802.1 Time Sensitive Networking (TSN) envelops a encountered that were able to support TSN yet. Unlike
set of standards developed by the TSN task group [10], under OMNeT++, the two do not readily work with all operating
the IEEE 802.1 working group. system (OS) platforms: OPNET works with Windows and
It aims to provide the determinism needed to guarantee Linux OS; NS-2 works with MacOS, Solaris and Linux, though
Quality of Service (QoS) in the end-to-end network flows [4], on Windows, it needs to be installed through the Cygwin
doing so by way of several standards, of which the main ones software.
are: NS is short for Network Simulator. Like OMNeT++, NS-2
has a modular design, and it is fully public-source. According to
[22], among NS-2’s main drawbacks, its object-oriented design IV. DETERMINISM AND WI-FI
limits its scalability for WSNs (Wireless Sensor Networks). There are certain factors that make Wi-Fi not time-sensitive.
Xian et al. [23] have noted how, although it has many protocol According to Khorov [27], the matter in question has to do with
models, they are centred around TCP/IP, which they believe the end delay of packets, something that can be attributed to two
makes these models excessively unitary. NS-3, on the other main sources, namely the delays in queueing and packet
hand, is seemingly a combination of NS-2 and other simulation transmission:
tools [22]. The name, NS-2, leads many to believe it is the
predecessor to NS-3, but this could not be further from the truth. =
NS-3 is not backwards compatible with NS-2, and [22] reports +
that NS-3 offers more realistic models. Like NS-2, it also needs (1)
third-party software, Cygwin, to be installed on Windows. Both
NS-2 and NS-3 programs are free and open-source and so invite
community development for the extension of their models. In relation to TSN, the queueing delay is a matter of how
Similar to OMNeT++, OPNET is structured into models. long the TSN packet should wait in a queue while those ahead
The review in [24] gives an in-depth analysis of OPNET and its of it are being transmitted and their successful receipt confirmed.
suitability for different use cases. For the research case, it is As for the transmission time for packets, this is summed up by
suited firstly because of its feature-rich library, and next to this, the PPDU (PLCP (Physical Layer Convergence Protocol)
its extensive documentation, which is suitable for development, Protocol Data Unit) duration, or packet duration, and the channel
testing and validation. In industry, it is apt as it allows for the access time. The channel access time could be especially
modelling of real-life situations. A comprehensive library of problematic because Wi-Fi makes use of unlicensed spectrum.
pre-existing models leaves less room for errors and supports A brief look into this spectrum helps explain where the problem
reliability; the simulator can be reliably used for testing and lies.
development in commercial and mission-critical applications Insomuch as frequency spectrum is a limited and expensive
[24]. The main drawback comes, however, in its closed nature – resource, many users prefer to deploy their proprietary or other
its source code does not provide access to its simulation kernel solutions in the unlicensed spectrum, to avoid any license fee or
[25]. Developers cannot easily develop in it or debug certain associated administrative steps. They thus tap into one of the few
problems in their simulations using it. In [26], Pahlevan and license-exempt frequency bands used by Wi-Fi (most often, in
Obermaisser do not have their TSN implementation, which is the 2.4 GHz band). This means many devices share the same
based on OPNET, publicly available. The networking frequencies, and no protection against interference can be
community cannot add to, use or extend it [19]. Moreover, if guaranteed. Since many devices are able to communicate over
work is not open-source it is not easy to compare it against this frequency band, Khorov [27] inferred that the channel could
similar applications and tools for benchmarking or other easily be occupied by other devices’ transmission activities.
purposes. Falk et al. [19] stated that they could not compare the
OPNET TSN implementation in [26] to their own TSN Furthermore, Wi-Fi is not even the sole technology that
simulator (NeSTiNg) with regards to its performance and how shares this spectrum: highly popular Bluetooth and Zigbee
accurate it is. technologies, both preferred for their low hardware costs and
low energy consumption at the end-devices [28], are among
Additionally, OMNeT++ is fully public- or open-source and many that sit at the 2.4 GHz band. Each day, more and more
widely extensible. It encourages researchers and developers to devices incorporate these IoT technologies, which only adds to
add to its frameworks and leverage its underlying power and the congestion and interference in the airwaves. Cellular
capabilities. Varga and Hornig [25] credit this as being the unlicensed spectrum technology is another technology that
reason for its plentiful simulation models and frameworks. shares spectrum with Wi-Fi. In addition to the crowded
Some studies have OMNeT++ as faring better than OPNET frequency range, mechanisms inherent in Wi-Fi, such as the
and NS-2; according to [23], OMNeT++ is the more powerful randomised medium access algorithm and channel contention
simulator as attested to by results from a WSN simulation procedure it employs to deal with collisions, further add to its
experiment they conducted with these three, based on criteria of non-determinism. Other technologies such as cellular and
run time, delivery rate, and memory requirement. That, along Ethernet do not suffer from these issues. Ethernet, in contrast,
with the greater number of supported functionalities, such as the through the employ of switches, can, for example, create duplex
powerful simulation library and debugging capabilities, made it links between these switches and other devices, thereby
come out on top. But because this work isn’t a comparative eliminating any chance of collisions with other communicating
study of simulation frameworks, it will not go more in-depth devices.
than this into what makes OMNeT++ better than other tools. As regards TSN, implementing its standards over Wi-Fi can
This work is rather one more concerned with the tool best used be difficult to achieve because of the above-discussed spectrum
for the simulation of TSN. sharing leading to congestion and associated errors and thus an
Finally, aside from OMNeT++, all the underlying tools indeterministic channel. While the channel is perfectly fitting for
chosen for this particular work can be freely obtained over the many consumer applications, it would typically be unreliable in
Internet. an industrial scenario with inflexible timing demands to the
order of milliseconds.
 V. TSN EMULATION MODEL
After assessing the requirements and estimating what would
be needed for the components, the design of the model followed.
On the conceiving of said model, the review of the literature
brought forth some concerns. The findings on wireless TSN,
particularly those in [6], [7], [29], [30], point towards the need
to include several significant enhancements in the present
wireless systems including Wi-Fi, to effect a time-sensitive
communication that reflects the TSN standards. These
modifications advocate for the definition of new standards and
modes of operation, some of which will require nothing short of
redesigning the physical layout of wireless devices to produce a
new hardware solution as in [27].
Our work followed a different, simpler approach. We instead
decided to leverage the current state-of-the-art hardware
components as they are and tried to integrate them with the
present wireless capabilities as suggested by [2].
Inspired by the roadmap envisaged by Bush and Mantelet Fig. 1. TSN Simulation Environment
[2], the writer conceptualises a different phased approach
towards going about this method. It proceeded in four key NeSTiNg was used primarily to build the base of the
phases: environment. Then the underlying INET framework models
were called on to extend the wired capabilities of NeSTiNg TSN
1) Introduce wireless hardware into the wireless domain. After the network devices were built
2) Configure the switch and access point interfaces and it was verified that the devices could communicate with one
3) Design the interconnection between wired TSN part and another, OMNeT++ was used to stage the simulation of the
wireless part network. The proposed system modelled the strict-priority
4) Establish communication between wired and wireless feature of NeSTiNg to prioritise critical frames over best-effort
components traffic.
In the model, to briefly outline the main points on the time-
Because these design considerations suggest some specific sensitive aspects of TSN, TSN packets are tagged with a special
degree of customisation, an open-source simulation program VLAN ID, which identifies the way they will be handled by
was seen as a good environment to build the model in. A switches (in terms of priority and urgency of the traffic flow).
simulator would offer the opportunity to trial-run experiments Just as important, queues are defined in the switches that help to
of different setups upon each modification of the code. Two directly shape the network traffic they receive. The earlier-
tools were chosen that look to fit this purpose: OMNeT++ and mentioned VLAN ID attribute and others (destination MAC
NeSTiNg. address and Class of Service) in the frames tell a TSN switch
OMNeT++ comprises a library built on extensible, modular, how to process frames and in which queues.
and open-source principles – it is a discrete event simulator [21]. Various tests of different topologies in several different
The Network Simulator for Time-Sensitive Networking scenarios were run [35]. The case detailed in this paper was set
(stylised as NeSTiNg) [31] also handles discrete event up with the topology shown in Fig. 2. The premise is that of a
simulation as it is based on OMNeT++. It started as an internal network with time-sensitive traffic and non-critical traffic
tool for TSN evaluation [32], and it can be used to analyse the accommodated in the same infrastructure. There are two distinct
behaviour of converged IEEE 802.1 TSN networks, as network traffic flows:
evidenced in the work of Falk et al. [19]. In this way, it is more
specialised in its function. (Instructions supporting its use can be 1) Flow 1: A traffic generator originating within the
found on their GitLab page [31].) nonTSNtalker devices tries to occupy the network with
background traffic whose destination is the nonTSNlistener
Since our setup involves a wireless environment different
from the traditional case of wired Ethernet, it calls for the need device. 1500-byte frames were sent out at an interval of 12 μs.
to leverage other capabilities of the existing networking tools. The packets travel via accessPoint1 to TSN switches A and B
The INET framework is an OMNeT++ library for wired, and then via accessPoint2 to the destination.
wireless, and mobile networks; it contains models for the 2) Flow 2: Amidst these ongoing transmissions, the
Internet stack, wired, and wireless link layer protocols [33]. It TSNtalker (talker with critical/time-sensitive traffic) transmits
has extensible support for the wireless use case intended to be to the TSNlistener to control its movements. In the talker, the
employed as a building block for the setup. cycle time was defined as 400 μs for the 354-byte TSN frames.
Fig. 1 overviews our multi-layer environment, built similarly The packets travel via the same path as for Flow 1.
to [34, Fig. 1].
 From these experiments, it became evident where a problem
lay: the majority of the packets in the network were lost due to
the access point. Although it had been expected that Ethernet
would outperform Wi-Fi, given the high rate of collisions in
Flow 1
freely accessible wireless mediums, it presented an interesting
case in seeing this large a magnitude of packets lost and the
access point inefficiently utilising the channel to such a degree.
Through further experimentation, after modifying the
parameters of these setups, better results were obtained. This
modification included reducing the contention window and
setting a low back-off interval so that frames waited for shorter
durations in the queues (this was applied uniformly for both
critical and non-critical frames) for faster recovery times, and
varying the send intervals for both types of frames. In some
Fig. 2. Simulation Setup cases where the bandwidth was constrained (for example, by
using 10 Mbps Ethernet) and the non-critical traffic was limited,
Some of the parameters used for the setup in Fig. 2 are shown the critical traffic that could be transmitted rose to above 99%
in Table I. [35]. Yet still, in a large number of cases there [35], the overall
efficiency of the channel was seen to be nonoptimal when the
capacity used was above a certain minimum (circa 10%).
TABLE I. PARAMETERS FOR THE MODEL Generally, the critical traffic was processed in a timely fashion
at the switches, however, it got bottlenecked at the Wi-Fi access
Parameter Value
points.
1 Centre frequency (bandwidth) 5 GHz (20 MHz)
VI. ANALYSIS AND CONCLUSION
2 Radio transmitter power 100 mW
The access point was reporting such low numbers seemingly
3 Radio receiver sensitivity -85 dBm because it did not have the shaping capabilities needed to
4 Radio receiver energy detection -85 dBm efficiently deliver TSN frames, especially in situations of high
throughput.
5 Radio receiver SNIR Threshold 4 dB
QoS parameters can be defined for an access point to
6 Ethernet LAN datarate 1 Gbps (highest)
determine how it deals with TCP/UDP traffic. Using the
7 Wireless LAN bitrate 693.3 Mbps (highest) Enhanced Distributed Channel Access (EDCA) mechanism,
which is a part of the MAC’s HCF (Hybrid Coordination
Function) submodule, the MAC layer applies QoS for packets
Apart from the setup in Fig. 2, two other setups were tested in belonging to one of four access categories, namely, voice, video,
the following order. best effort and background, in order of decreasing priority.
1. The first was a fully wired setup (excluding both However, in the case of the novel model in [19], no higher layer
wireless access points). protocols are used; it is purely Layer 2 Ethernet frames. So, this
QoS feature was not applicable.
2. The second setup was one with the left-hand side
without wireless, that is, accessPoint1 was removed, Operating in Wi-Fi frequencies, where there is already a high
and the talkers were connected directly to switchA. chance of collisions, the situation was worsened since access
points generally do not forward frames as they are received; they
3. Then finally the third setup with two access points have to first check and acknowledge the availability of the
(APs) shown in Fig. 2. channel. The access points’ queues quickly got full, and so
critical frames were dropped in the process, on pair with any
It was noted that in the first two setups, in the span of 1
other frames.
second, 83,128 packets successfully arrived at switchB, which
then transmitted this same number. Of the about 86,000 packets With the first access point, even with the fore-mentioned
sent out by the talkers, this corresponds to a delivery rate of modified parameters, the AP’s innate abilities do not allow it to
96.7%. distinguish between TSN vs non-TSN frames, and so they were
treated as the same kind of traffic. Secondly, even with an
In the third setup, only a few arrived (4,844), a fraction
optimum combination of parameters such as those in Table I, the
(5.6%) of the ~86,000 sent out by the talkers.
access point still could not process frames fast enough, which
Similarly, on the outgoing interface, eth[1], there were 4844 filled up the queues in the wireless devices and led to many
packets sent out (packets in = packets out). frames being dropped.
The only difference between this third and the two prior Based on our results, Wi-Fi’s ability to offer determinism
experiments was the inclusion of a wireless access point before could be improved by including some features into the standard.
the first switch. One solution could be for its access categories to redefine their
priority scheme to include support for a TSN traffic class.
Further, TSN could borrow from the DetNet specification [8], Stream Reservation Protocol (SRP) Enhancements and Performance
[36] and introduce a bounded or maximum bandwidth existing Improvements, vol. 2018. 2018.
for different kinds of traffic in the network. That way, the best- [17] IEEE Standards Association, IEEE Standard for Local and metropolitan
area networks--Frame Replication and Elimination for Reliability. 2017.
effort traffic will not be allowed to overwhelm the network and
[18] D. Cavalcanti, G. Venkatesan, and C. Cordeiro, “PAW-IETF 104-Prague
choke out the critical/time-sensitive traffic as the best-effort Time-Sensitive applications support in 802.11ax and 802.11be (EHT),”
traffic’s throughput will be capped at a certain limit. 2019.
Expounding on this, say, no more best-effort traffic should be [19] J. Falk et al., “NeSTiNg: Simulating IEEE Time-sensitive Networking
allowed to flow through the network past a certain threshold (TSN) in OMNeT++,” in 2019 International Conference on Networked
volume, maybe unless the critical traffic throughput and delivery Systems (NetSys), 2019, pp. 1–8, doi: 10.1109/NetSys.2019.8854500.
rate are kept above a certain level or above a certain percentage [20] S. Brooks and E. Uludag, “Time-Sensitive Networking: From Theory to
of the whole network’s traffic at any time. Implementation in Industrial Automation,” TTTech Intel White Pap., Oct.
2018.
REFERENCES [21] OMNeT++, “OMNeT++.” [Online]. Available: https://omnetpp.org/.
[Accessed: 26-Nov-2019].
[1] L. Columbus, “2018 Roundup Of Internet Of Things Forecasts And
Market Estimates,” Forbes, Dec-2018. [Online]. Available: [22] R. Sharma, V. Vashisht, and U. Singh, “Modelling and simulation
https://www.forbes.com/sites/louiscolumbus/2018/12/13/2018-roundup- frameworks for wireless sensor networks: a comparative study,” IET
of-internet-of-things-forecasts-and-market-estimates/#386fb5e37d83. Wirel. Sens. Syst., vol. 10, no. 5, pp. 181–197, Oct. 2020, doi:
[Accessed: 25-Nov-2019]. 10.1049/iet-wss.2020.0046.
[2] S. F. Bush and G. Mantelet, “Industrial Wireless Time-Sensitive [23] Xiaodong Xian, Weiren Shi, and He Huang, “Comparison of OMNET++
Networking: RFC on the Path Forward,” in Avnu Alliance® White Paper, and other simulator for WSN simulation,” in 2008 3rd IEEE Conference
2018, pp. 1–12. on Industrial Electronics and Applications, 2008, pp. 1439–1443, doi:
10.1109/ICIEA.2008.4582757.
[3] J. Hibbard, “5 Real-Time, Ethernet-Based Fieldbuses Compared,” AIA
(Automated Imaging Assoc. Vis. Online White Pap., Mar. 2016. [24] S. Mittal, “OPNET: An Integrated Design Paradigm for Simulations,”
Softw. Eng. An Int. J., vol. Vol. 2, no. No. 2, pp. 68–84, Sep. 2012.
[4] A. Nasrallah et al., “Ultra-Low Latency (ULL) Networks: The IEEE TSN
and IETF DetNet Standards and Related 5G ULL Research,” IEEE [25] A. Varga and R. Hornig, “An Overview of the OMNeT++ Simulation
Commun. Surv. Tutorials, vol. 21, no. 1, pp. 88–145, 2019, doi: Environment,” in Proceedings of the 1st International Conference on
10.1109/COMST.2018.2869350. Simulation Tools and Techniques for Communications, Networks and
Systems & Workshops (Simutools ’08), 2008, pp. 1–10, doi:
[5] A. Mildner, “Time Sensitive Networking for Wireless Networks-A State 10.5555/1416222.1416290.
of the Art Analysis,” Semin. IITM WS 18/19, 2019, doi: 10.2313/NET-
2019-06-1_07. [26] M. Pahlevan and R. Obermaisser, “Evaluation of Time-Triggered Traffic
in Time-Sensitive Networks Using the OPNET Simulation Framework,”
[6] D. Bankov, E. Khorov, A. Lyakhov, and M. Sandal, “Enabling real-time in 2018 26th Euromicro International Conference on Parallel, Distributed
applications in Wi-Fi networks,” Int. J. Distrib. Sens. Networks, vol. 15, and Network-based Processing (PDP), 2018, pp. 283–287, doi:
no. 5, p. 155014771984531, May 2019, doi: 10.1109/PDP2018.2018.00048.
10.1177/1550147719845312.
[27] E. Khorov, “Wi-Fi Time Sensitive Networking,” IEEE Mentor, Nov-
[7] T. Adame, M. Carrascosa, and B. Bellalta, “Time-Sensitive Networking 2017. [Online]. Available: https://mentor.ieee.org/802.11/dcn/17/11-17-
in IEEE 802.11be: On the Way to Low-latency WiFi 7,” Dep. Inf. 1734-00-0wng-wtsn.pptx. [Accessed: 30-Sep-2019].
Commun. Technol. Univ. Pompeu Fabra, Dec. 2019.
[28] L. Zhang, Y.-C. Liang, and M. Xiao, “Spectrum Sharing for Internet of
[8] X. Yang, D. Scholz, and M. Helm, “Deterministic Networking (DetNet) Things: A Survey,” IEEE Wirel. Commun., vol. 26, no. 3, pp. 132–139,
vs Time Sensitive Networking (TSN),” in Proceedings of the Seminar Oct. 2018, doi: 10.1109/MWC.2018.1800259.
Innovative Internet Technologies and Mobile Communications (IITM),
2019, pp. 79–84, doi: 10.2313/NET-2019-10-1_15. [29] C. Mannweiler et al., “Reliable and Deterministic Mobile
Communications for Industry 4.0: Key Challenges and Solutions for the
[9] A. B. D. Kinabo, J. B. Mwangama, and A. A. Lysko, “An Overview of Integration of the 3GPP 5G System with IEEE - VDE Conference
Time-Sensitive Communications for the Factory Floor,” in IST-Africa Publication,” in Mobile Communication - Technologies and
2021 Conference Proceedings, 2021, pp. 1–9. Applications; 24. ITG-Symposium, 2019, pp. 1–6.
[10] J. Farkas, “Introduction to IEEE 802.1 - Focus on the Time-Sensitive [30] S. Sultana, “LS on 3GPP 5G System support for integration with IEEE
Networking Task Group,” Ericsson, Mar. 2018. TSN networks,” in Standards Association (SA) Working Group #2
[11] IEEE Standards Association, 802.1AS-2011 - IEEE Standard for Local (WG2), 2019, pp. 1–2.
and Metropolitan Area Networks - Timing and Synchronization for Time- [31] D. Hellmanns and J. Falk, “ipvs / nesting project · GitLab.” Distributed
Sensitive Applications in Bridged Local Area Networks, no. March 30. Systems group of IPVS, University of Stuttgart.
IEEE, 2011.
[32] D. Hellmanns, “NeSTiNg - A Network Simulator for Time-sensitive
[12] M. D. Johas Teener and G. M. Garner, “Overview and timing Networking,” in September 2018 Interim Meeting in Oslo, Norway, 2018.
performance of IEEE 802.1AS,” in 2008 IEEE International Symposium
on Precision Clock Synchronization for Measurement, Control and [33] INET, “INET Framework.” [Online]. Available:
Communication, 2008, pp. 49–53, doi: 10.1109/ISPCS.2008.4659212. https://inet.omnetpp.org/. [Accessed: 26-Nov-2019].
[13] IEEE Standards Association, IEEE Standard for Local and metropolitan [34] J. Jiang, Y. Li, S. H. Hong, A. Xu, and K. Wang, “A Time-sensitive
area networks -- Bridges and Bridged Networks - Amendment 25: Networking (TSN) Simulation Model Based on OMNET++,” in 2018
Enhancements for Scheduled Traffic. 2016. IEEE International Conference on Mechatronics and Automation
(ICMA), 2018, pp. 643–648, doi: 10.1109/ICMA.2018.8484302.
[14] TTTech, “IEEE TSN (Time-Sensitive Networking): A Deterministic
Ethernet Standard,” TTTech, 2015. [35] A. B. D. Kinabo, “Time Sensitive Networking for Wi-Fi Based Wireless
Industrial Environments,” University of Cape Town, 2021.
[15] IEEE Standards Association, IEEE Standard for Local and metropolitan
area networks -- Bridges and Bridged Networks -- Amendment 26: Frame [36] IETF, “Deterministic Networking (detnet) - Documents,” Datatracker -
Preemption. 2016. Internet Engineering Task Force. [Online]. Available:
https://datatracker.ietf.org/wg/detnet/documents/. [Accessed: 18-Dec-
[16] IEEE Standards Association, IEEE Standard for Local and Metropolitan 2020].
Area Networks--Bridges and Bridged Networks -- Amendment 31:

View publication stats