FOCUS ON SMART CITIES

A Framework for the Design and

Deployment of Large-Scale LPWAN
Networks for Smart Cities Applications
Bassel Al Homssi, Akram Al-Hourani, Kagiso Magowe, James Delaney, Neil Tom, Joan Ying, Hans Wolf, Simon Maselli,
Sithamparanathan Kandeepan, Ke Wang, and Karina Mabell Gomez

Abstract
The Internet of Things (IoT) enables the connectivity of massive numbers of sensors to facilitate the delivery of immense sets of data. This
valuable data promotes higher productivity, effectiveness, and sustainability. With the accelerating global urbanization, many cities and sub-
urbs are expected to face increasing challenges related to urban planning, resources and livability. These challenges stimulate new oppor-
tunities for communities and governments to collaboratively incubate new planning and development strategies relying on the insightful
IoT data. This concept is commonly referred to as smart cities. Accordingly, a wave of smart cities implementations is currently emerging.
However, the underlying IoT network deployments are developed in silos, and thus require more harmonization. In this paper, we present
a design and deployment framework for large-scale low power wireless access networks (LPWAN) in order to harmonize and guide the
end-to-end process. This framework is derived from the acquired expertise and learned lessons from the largest open-access LPWAN deploy-
ment in Australia. The framework includes various aspects related to the state-of-the-art design techniques, smart cities use-case scenarios,
project development and implementation, network management, and finally a robust performance testing mechanism for quality of service
assurance. All these components were put into rigorous verification and testing under this realistic large-scale IoT network deployment.

Introduction ed battery life of several years. A widely deployed LPWAN

technology is LoRaWAN, supported by the LoRa Alliance
More than 68 percent of the world’s population is projected which is an association of more than 500 member companies.
to live in urban cities by 2050 [1] almost doubling the number LoRaWAN utilizes a proprietary modulation and coding scheme
of their residents. Growth of this magnitude would certainly based on a chirp spread spectrum (CSS) technique which
provide many economic opportunities, but will equally present spreads over a lager radio spectrum for better interference
several challenges including higher traffic congestion, more pres- immunity. Signals with different spreading factors can coexist in
sure on natural resources, and increased levels of CO2 emissions the channel. Moreover, increasing the spreading factor enhanc-
and other pollutants. This projection has led many governments es the communication link budget, making the signal more
around the world to adopt a transformative smart cities vision to resilient to remote deployments. However, a higher spreading
address these arising challenges. The renewed focus on smarter factor also lowers the data rate, leading to longer transmission
cities will not only lead to a better utilization of infrastructure time and increasing energy drainage [5].
and improvements in energy efficiency and services, but will Over the past few decades, the deployment of large-scale
also ensure that cities remain globally competitive and attractive. networks was almost always associated with large telco oper-
Smart cities include many application scenarios which have ators. However, with the advent of simpler LPWAN technol-
particularly been of a great interest to local governments [2] ogies, emergent network deployments can be carried out by
ranging from smart parking, pollution monitoring, occupancy smaller sized new comers to the market or by the end-users
estimation, smart lighting, smart metering, asset tracking, waste themselves. Accordingly, the harmonization and integration of
management to water-level monitoring. Among the many smart such deployments becomes of utmost importance to allow the
cities technology drivers is the Internet of Things (IoT) that plays LPWAN to provide a seamless continuum and avoid unnec-
a major role in shaping smart cities and their future road-map essary capital investment. The harmonization needs to be
[3]. At the heart of the smart city model, a massive number of accounted for not only in the technology, but also in the way
IoT sensors collect data from their surroundings to enable back- the network is designed, tested, and in the quality assurance.
end analytics [4]. These analytics draw insightful understand- Ongoing standardization efforts for IoT data network infrastruc-
ing of the current state of a city and project different solutions tures exist [6]; nevertheless, standardization for the network
based on the evolving Artificial Intelligence (AI). design and implementation phases is still underexplored.
To cater to battery-powered IoT wireless devices, several In this article, we present an end-to-end framework based
technologies have been proposed, collectively referred to as on the lessons learned from the deployment of a large-scale
low power wide area networks (LPWAN) aiming for an extend- LPWAN for smart cities application in Melbourne, Australia. The
framework provides a blueprint to guide the process of network
Bassel Al Homssi, Akram Al-Hourani, Kagiso Magowe, James Delaney, Neil Tom, development all the way from the use-case scenario creation,
Sithamparanathan Kandeepan, Karina Mabell Gomez, and Ke Wang are with the to the radio design methodology, network architecture, net-
School of Engineering, RMIT University. work management, implementation and finally the testing and
Hans Wolf and Joan Ying are with the City of Whittlesea Council and Banyule City quality assurance. The framework is developed and put into
Council, respectively. action through the development of the Northern Melbourne
Simon Maselli is with Minnovation Australia. Smart Cities Network (NMSCN) project, which is the largest
Corresponding author A. Al-Hourani: e-mail: [email protected] single deployment of an open-access LPWAN network in Vic-
Digital Object Identifier: 10.1109/IOTM.001.2000179 toria, Australia. The NMSCN network has already integrated

IEEE Internet of Things Magazine • March 2021 2576-3180/21/$25.00 © 2021 IEEE 53

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 53 3/22/21 5:27 AM

 Figure 1. IoT Smart city concept in different urban environments where radio coverage largely depends on the interference levels and
the propagation conditions.

48 LoRaWAN gateways and 294 sensors across the five con- • Offer new services such as advising the community about air
tributing local council areas. The framework underscores the quality and water levels in the local creeks.
network design, radio frequency testing, connectivity testing, • Develop trigger-based alert systems for both public safety and
and quality assurance of the network. Ultimately, the frame- internal staff use.
work is aimed to help network planning engineers, researchers • Understand community demands and better allocate assets
and decision makers alike to gain insightful understanding of and services.
the challenges and opportunities of large scale deployments, • Invest in future services based on data trends and community
as well as facilitate better harmonization in future IoT networks behavior prediction.
and expansions. • Motivate innovation by allowing open-access to the network
by the public.
Northern Melbourne Smart Cities Network The technical objective of the NMSCN project is to establish
The Australian Government launched the smart cities plan in an open-access LoRaWAN network consisting of an initial roll
2016 as a long-term framework to support the growth of the out of 48 gateways and 294 sensors. Nevertheless, the ultimate
cities and suburbs. The aim of this framework is to achieve high- goal of the project is for the network to be scalable and engulf
er sustainability, better productivity, and more liveable cities. more sensors managed by the councils or the community in the
One of the main pillars of this plan is the use of Smart Technol- short-term future. The choice of the sensors was based on the
ogy to drive investment and policies [7]. Smart Technology is overwhelming interest shown by the councils in five use-cases
a broad term, but the plan has further emphasized the impor- during the initial consultation phases of the project:
tance of open and real time data for informing better policies Waste management sensors to capture the bin usage and
and planning. To kick start the plan, AUD $50 million (around determine trends of usage over time across the community.
USD $35 million) were awarded in the form of competitive Water level sensors to monitor the levels of water in creeks
grants to the bidding councils. and alert councils in the case of flooding and to monitor levels
The bid to build the NMSCN network was successfully select- in regional area fire fighting water towers.
ed and the project was awarded AUD $1.5 million to build an Occupancy estimation sensors to anonymously capture
open IoT network covering spots across the five contributing data of pedestrian traffic in shopping strips, public parks and
councils [8]. The network is purposed to gather sensor data other public areas of interest.
to support further analytics required to monitor and improve Environmental sensors as a new service to understand local-
the efficiency of existing services and to ultimately support the ized air quality, humidity, temperature and other environmental
potential of new service delivery. The NMSCN is overarching a parameters.
diverse combination of inner city, interface, and rural councils Asset tracking sensors to better track mobile council ser-
as illustrated in Fig. 1. To offer services to their respective com- vices such as street sweepers, mowers and council vehicles to
munities, the contributing local councils maintain many assets better allocate these resources.
such as public buildings, parks, footpaths, roads, playgrounds, In total, eight organizations were working together to
public toilets, and public BBQs, among many others. However, deliver this largest open-access IoT network in Victoria. The
prior to the Australian Government initiative, councils invested project is a manifestation of the collaboration work between
a considerable amount of time and money in the construction councils, universities and industry. Please refer to the official
and maintenance of public assets with little data showcasing web page of the project for more details about the contrib-
their actual usage. Hence, one of the main objectives of the uting partners. Five Councils came together to be part of the
NMSCN project is to enable the collection of data on the usage NMSCN project: City of Whittlesea Council (lead council),
of public assets by the community. This enables councils to pro- Banyule City Council, Moreland City Council, Mitchell Shire
mote the following objectives: Council and Nillumbik Shire Council. Royal Melbourne Insti-

54 IEEE Internet of Things Magazine • March 2021

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 54 3/22/21 5:27 AM

 Figure 4. Cellular IoT model network architecture adopted in NMSCN.

tute of Technology (RMIT) University is a partner responsible the management in large-scale IoT networks. In addition to con-
for grant preparation, network design, testing and quality trol plane management, the developed NMS is also responsible
assurance. La Trobe University is another partner respon- for the aggregation of the payload from both the TTN and LTE
sible for the data analytics and visualization. Minnovation networks, where the aggregated payload data is forwarded to
Australia is the chosen partner responsible for the supply of the data analytic/visualization platform.
the sensors and gateways along with all associated installa- Cloud Data Platform: utilized for data storage, analytics,
tion services as well as hardware maintenance and support. and visualization. The cybersecurity of the LoRa network is
This project attracted two prestigious industry awards from managed by TTN, whereas that of the 4G network is managed
Municipal Association of Victoria (MAV) Technology Awards by the LTE provider. The authentic security of the gateways and
for Excellence 2020, and IoT Alliance Australia (IoTAA) devices is managed by Minnovation instead.
Smart Cities Award for 2020.
Design Methodology
Design Framework The main objective of the radio access network design phase
is to determine suitable locations for the gateways. Figure 3
Network Architecture illustrates the network design process that was developed in
In principle, the adopted layout in NMSCN follows the typical NMSCN. Ideally, each gateway is expected to serve multiple
cellular network topology, where IoT devices communicate sensors to optimize the cost-effectiveness of the deployment.
directly with their serving LoRaWAN gateways. The network Hence, optimal gateway locations are those nearest to clusters
primarily utilizes LTE backhaul over commercial telecom oper- of sensors. Geographical areas that include clusters of sensors
ators. In some cases, where the LoRaWAN coverage becomes are referred to as sites. To determine the optimal location of
inadequate, specifically when the IoT devices are quite sparse, the gateway for a given site we propose the following scheme:
4G LTE radio modules are utilized such that the IoT traffic is Site survey: Performing site visits is a crucial step to identify
handled directly by the commercial telecom operators. Figure 2 candidate locations for installing the gateways and sensors. For
illustrates the end-to-end connection and highlights the adopted instance, it is preferred to install the gateway in an open area
network architecture in NMSCN which is constituted of the that maximizes the line-of-sight probability with the sensors [9].
following: RF Ray Tracing: After identifying the candidate locations, we
IoT Devices: having different types of sensors as indicated in further optimize the selection based on iterative RF ray-tracing
the previous section, equipped with the radio network access simulations. In addition to the candidate locations, the simulator
modules. also incorporates two important geographic data: (i) the terrain
LoRaWAN Gateways: responsible for the reception, demod- data, often called digital elevation model (DEM), and (ii) the
ulation and decoding of the LoRaWAN transmissions. LoRaWAN building geometry and heights, often referred as the buildings
Gateways relay the aggregated traffic via LTE backhaul. footprint which is a 2.5D data-set.
LoRaWAN Management Network: used for extracting the Coverage Mapping: The coverage of radio service is
payload data sent from the gateways. For this project, an open determined by the ability of the receiver to decode the
platform, The Things Network (TTN), is utilized to enable public transmitted message correctly. This is dependent on both
access to community users. the received signal power and the surrounding noise and
Network Management System (NMS): developed by the interference. Thus, in addition to the ray-tracing simulation
industry partner, the platform is called Alpha-X1. The platform it is important to get the interference statistics (or at least
monitors and manages the health of the network and its ele- the average interference power) around the selected sites
ments through the control plane. The NMS reduces the man- because the network is operating in the license-free ISM
agement workload and classifies and predicts battery usage band. See the references cited herewith for further details
based on consumption trends. It also provides tools that allow about interference modeling [10]–[12]. The power level at
the user to set up templates of device types and network types which an acceptable success rate is maintained is called the
to efficiently assign those settings to multiple devices optimizing sensitivity of the gateway and is dependent on the utilized

IEEE Internet of Things Magazine • March 2021 55

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 55 3/22/21 5:27 AM

 Figure 3. Network design methodology utilized in NMSCN to achieve the the design objectives.

spreading factors in LoRa modulation [5]. Higher spreading LoRaWAN device referred to as a packet tester and further
factors provide better coverage, but require longer time-on- explained hereafter.
air, which shortens the device’s battery lifetime. Hence, it is Packet tester: A device we developed at RMIT University in
desirable to utilize the lowest spreading factor to lower the order to facilitate the radio coverage validation. This device sends
operational costs arising from battery replacement. It is sim- LoRaWAN frames to the deployed network and logs the proper-
pler to communicate the outcome of this step as a coverage ties of these frames including geographic coordinates, spreading
contour, similar to the one in Fig. 6. Radio coverage design factor, transmit power, and frame sequence number. The devel-
is typically an iterative process since the obtained coverage oped device is shown in Fig. 5 alongside the basic end-to-end
by simulation is still an estimate and cannot incorporate pre- architecture of the network testing system. The testing mech-
cisely all real-wold factors, including surface materials and anism utilizes similar infrastructure to the gateways deployed
foliage, among many more. Thus, some sites may require along with an additional cloud-based server to extract and store
optimization post-installation using radio frequency (RF) test- the data received from the testing device via the LoRaWAN
ing, which is further discussed in a later section. network. The packet tester is primarily built using available off-
the-shelf modules to achieve better interoperability with common
Implementation LoRaWAN network server and/or application server. The param-
The industry partner developed the majority of the IoT eters of testing scenarios are highly configurable, providing the
devices by integrating commercially available modules and flexibility to configure the LoRaWAN serial keys as well as the
in-house developed firmware. The integration also includes spreading factor, transmit power, frequency sub-band, and finally
the mechanical housing of IoT devices and the battery man- the total number of frames to send for a specific test.
age.ment system. The gateways and some sensors required Pin-point test: The first test is to perform pin-point measure-
larger rechargeable batteries with solar power system inte- ments. The packet tester is configured to send a fixed number
gration. The rooftop installations of the gateways typically of frames once it is positioned at the potential locations of the
require 60 minutes to complete onsite since the gateways IoT sensors. The ratio of the number of successful frames to
themselves are self-contained packages that are specifically the total number of frames is used to estimate the probability
designed to minimize site installation. Similarly, each sensor of frame success rate. If the probability is higher than a certain
requires around 45 minutes to install since the sensors are threshold, the sensor location is approved for installation. If that
also self-contained units, whereas larger sensors that require is not the case, another location is proposed instead.
an external solar panel each needed approximately 1.5 hours Coverage test: The second test is conducted via a drive
to complete. around the gateway location. The packet tester is configured
One of the main challenges of the project is the stringent to continuously transmit LoRaWAN frames while in-motion on
installation timeline in addition to the COVID-19 pandemic that a selected route. The drive routes are selected to account for
restricted the typical access to sites and imposed certain work various distances around the gateway as depicted in Fig. 5.
restrictions. Some installations also included hurdles such as sun The geographic coordinates are also logged by the packet
shading which limited the flexibility for solar panel installations tester in order to identify the locations from where the frames
and did not always coincide with the optimal location for cover- where transmitted. The transmitter log is extracted from the
age. Figure 4 illustrates the implementation phase and highlights packet tester, whereas the receiver log is obtained from the
examples of the different installation types completed in the cloud server. Both logs are compared in order to identify the
network. lost and successful frames.
The implementation process started with the installation of To expand the measured coverage to a larger grid, we fur-
the gateways first in order for the coverage testing to take place ther use these frames to tune the simulation-based coverage
and validate the sensor locations. Accordingly, the results from model. This is done by tuning the simulated coverage such that
the testing are utilized to feedback the design phase in order to the discrepancy between the simulation and measurements are
sustain the gateway connectivity. minimized, in particular by minimizing the root mean square
error (RMSE) [13]. The resulting coverage of the gateways in
Coverage Validation and Quality Assurance the NMSCN project is depicted in Fig. 6 for data rates DR0 and
DR5 (spreading factors SF12 and SF7 respectively). The overall
RF Testing and Optimization coverage guides the councils to direct future candidate locations
The validation of the radio network performance is an import- of new sensors as well as identifying black spots where the cov-
ant part of the overall verification and quality assurance since erage requires enhancement using future gateway installations.
only field measurements can reveal the real performance of
the network in comparison to the expected performance in Installation Quality Assurance and Troubleshooting
the design phase. The proposed methodology utilized for Quality assurance is performed to assess the implementation
the testing is illustrated in Fig. 5 where we utilize a hand-held and ensure the connectivity of the different IoT radio network

56 IEEE Internet of Things Magazine • March 2021

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 56 3/22/21 5:27 AM

 Figure 4. Installation photos for all the different network elements included in the NMSCN project.

elements. The quality assurance employed in the NMSCN proj- and awareness, the technical project partners delivered multi-
ect is based on two key criteria: ple informative workshops followed by targeted meetings to
Connectivity status of network elements: In addition to engage the councils. During the final stages of the project, sev-
the radio coverage verification, we further test the end-to-end eral training sessions have been conducted to provide the coun-
connectivity of gateways and sensors with the NMS. A trouble- cil staff with expertise to manage the network prior to handover.
shooting plan is developed to tackle the connectivity issues that
may arise due to coverage limitations or intermittent connec- Lessons Learned
tion. The plan also includes troubleshooting steps which include This project has provided a great opportunity to get detailed
tuning the spreading factor, transmit power, frequency sub- insights both on the technical level and the administrative level
band as well as identifying hardware and software deficiencies. when deploying large-scale IoT networks.
If the issue persists, a 4G module is proposed to substitute the
LoRaWAN module. Technical Lessons
Quality of installation: A procedure conducted to ensure Lesson 1–Conduct a thorough radio planning: When
that the physical installation is following the project specifica- deploying IoT networks in urban environments it is import-
tions and the different international and national organizations’ ant to validate the marketing claims of many IoT technologies
requirements. This includes the verticality of the antennas, RF in regards to the distance and performance, as most of the
and electrical cabling ratings such as UV protection, as well as marketing figures do not report the success rate or the level
battery performance, PV panels’ performance, cable conduits, of interference at which the measurements are conducted.
enclosure ratings among many other standards and regulations. Thus, radio planning is a crucial step for successful planning
and dimensioning of the networks, especially for large-scale
Project Management of Network Deployment multi-gateway networks. This was one of the key success factors
The overall timeline of the project spanned over 18 months on the project.
to achieve the following key milestones: (i) smart cities net- Lesson 2–Get the right sensor: With the hype associated
work design, (ii) procurement of gateways, sensors and installa- with IoT, it is important to frame the targeted outcome from
tion services, (iii) implementation of the network elements, (iv) the collected data, because the choice of the type of sensors
implementation of data visualizations, and (v) broader engage- is crucial in achieving the desired outcomes. For instance, an
ment with the community. The project was managed using IoT motion detection sensor still cannot provide occupancy
PRINCE2 project management methodology and processes. data whereas a sensing camera can. This project leveraged the
Product-based planning is utilized to identify and document a strength of the industry partner in developing custom-made sen-
list of activities to ensure clear communication among the team. sors and was another key success factor.
Due to the collaboration of eight different project partners, Lesson 3–Develop a clear quality assurance process: A
monthly team meetings to communicate the project progress, quality assurance process is a paramount framework aiming
discuss issues, plan for activities and make decisions were con- to ensure complete implementation, quality installation and
ducted. Furthermore, the goal of this project is to change the reliable connectivity of IoT devices according to the planned
councils’ conventional methodology from reactive to proac- use-cases and network design. This process has both technical
tive decision making to provide services which are now driven and financial implications and thus needs to be clearly devel-
through data and trends. On the other hand, to raise interest oped based on industry best practices. In this project, a clear

IEEE Internet of Things Magazine • March 2021 57

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 57 3/22/21 5:27 AM

 Figure 5. An illustration of the packet tester and its architecture and the network performance testing methodology.

Figure 6. Overall NMSCN outdoor coverage map. The light green represents the coverage for DR0 while the dark green represents
the coverage for DR5.

QA process was continuously developed and visited that has

Advancement to 5G Technologies and Deployment
ensured high quality standards. On one hand 5G, formally known as the IMT-2020 stan-
dard, includes three classes of use cases: (i) mobile broad-
Administrative Lessons band, (ii) ultra reliable low latency (URLL), and (iii) massive
Lesson 4–End-users’ engagement: The end-user engage- machine-type communications (mMTC) with a focus on tel-
ment is imperative to enable a more directed and timely project co-deployed networks which sustain a controlled level of
delivery. In this project, the councils were fully engaged, facil- interference and quality of service. On the other hand, unli-
itating a seamless transition between different project phases censed LPWAN such as LoRaWAN particularly complement
which has led to a timely project delivery. mMTC use case in 5G by empowering end-users to deploy
Lesson 5–Inclusive project management: The importance their own networks (similar to the amalgam of LoRaWAN
of having periodical meetings that involve different stakeholders and 4G deployed in this project). This hybrid approach in
(contractor, network designers, data scientists, and clients) leads having complementary systems, i.e. 5G-mMTC and LPWAN,
to adaptive and robust decision making. This project followed is significantly accelerating IoT innovation and addresses a
inclusive decision making that has led to its agile roll-out and wider range of techno-economic needs. This complementary
timely completion. nature is akin to the very successful legacy co-existing of
cellular and wireless local area network (WLAN) technolo-

58 IEEE Internet of Things Magazine • March 2021

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 58 3/22/21 5:27 AM

 gies that is continuing to play a major roll in last-mile access Akram Al-Hourani is the lead investigator on this project at RMIT University. He
is a senior lecturer and the program manager for the Master of Engineering (Tele-
through 5G and Wi-Fi. communication and Networks) at the School of Engineering, RMIT University,
Melbourne, Australia. He completed the Ph.D. degree in 2016 from RMIT Uni-
Conclusion versity. He has extensive industry/government engagement as a chief investigator
in multiple research projects related to the Internet-of-Things, smart cities, and
The deployment of a large-scale network has almost always
satellite/wireless communications.
been associated with large telecom operators. However, with
the introduction of simplified IoT access technologies, the Kagiso Magowe holds a doctoral degree in electronics and telecommunications
deployment of large-scale networks is now the responsibility engineering and is currently a research fellow with the School of Engineering,
of broader entities using free-access radio bands. Therefore, RMIT University. His research interests include communication theory, informa-
tion theory, statistical signal processing, application of linear algebra, cognitive
it is crucial to develop a harmonized framework that can radios, localization, Internet of Things, and the application of machine learning.
facilitate the different aspects of network deployments. In
this article, we detailed some of the lessons learned from a James Delaney is currently working toward his master of engineering at the School
large-scale deployment of an open IoT radio access network, of Manufacturing, Materials and Mechatronics Engineering, RMIT University,
Melbourne, Australia. He received a bachelor of engineering in 2018 also from
where we presented the developed framework for design- RMIT University, specializing in electrical and electronic engineering. His current
ing, implementing, testing and managing this deployment. research interests include Internet of Things, smart cities, and machine learning for
We highlighted the different challenges faced in the project wireless communications.
on both technical and management levels. Moreover, the
Neil Tom completed his master of engineering specialized in telecommunications
NMSCN project attracted two prestigious industry awards and networks from RMIT University in 2019. He graduated with a bachelor of
from Municipal Association of Victoria (MAV) Technolo- technology in 2017 specialized in electronics and communication engineering.
gy Awards for Excellence 2020, and IoT Alliance Australia His current work and interests focus on the field of IoT.
(IoTAA) Smart Cities Award for 2020.
Joan Ying is the IT Infrastructure Project Manager at the Banyule City Council
References and a project manager of the Northern Melbourne Smart Cities Network project.
With a master’s degree in information technology and over 10 industrial certifi-
[1] A. Zanella et al., “Internet of Things for Smart Cities,” IEEE J. Internet of Things,
cations, she has successfully delivered many projects, including cloud migrations,
vol. 1, no. 1, Feb. 2014, pp. 22–32.
network deployments, security services and unified communication solutions.
[2] A. Solanas et al., “Smart Health: A Context-aware Health Paradigm within
Smart Cities,” IEEE Commun. Mag., vol. 52, no. 8, 2014, pp. 74–81.
Hans Wolf holds a masters of business (information technology). He is the Chief
[3] W. Ejaz et al., “Efficient Energy Management for the Internet of Things in
Information Officer at the City of Whittlesea and sponsor of the Northern Mel-
Smart Cities,” IEEE Commun. Mag., vol. 55, no. 1, 2017, pp. 84–91.
bourne Smart Cities project. He has led technology teams in various public and
[4] Y. Mehmood et al., “Internet-of-Things-based Smart Cities: Recent Advances
private sector organizations ensuring stakeholder value by leveraging technology.
and Challenges,” IEEE Commun. Mag., vol. 55, no. 9, 2017, pp. 16–24.
[5] B. Al Homssi et al., “On the Bound of Energy Consumption in Cellular IoT Net-
Simon Maselli is the founder and CEO of Minnovation Australia (the industry
works,” IEEE Trans. Green Commun. and Networking, pp. 1–1, 2019.
partner in this project). His Australian based company manufactures IoT and
[6] “IEEE Standard for an Architectural Framework for the Internet of Things
telemetry products that have been deployed across Australia and New Zealand,
(IoT),” IEEE Std 2413-2019, 2020, pp. 1–269.
South East Asia and Europe for dozens of global brands.
[7] Australian Government, Department of the Prime Minister and Cabinet,
“Smart Cities Plan,” Australian Government, May 2016, [online]. Available:
Sithamparanathan Kandeepan is a full professor and the Discipline Leader for
https://www.infrastructure.gov.au/cities/smart-cities/plan/ files/Smart Cities
Telecommunications and Photonics, School of Engineering, RMIT University. His
Plan.pdf [Accessed: May 2, 2020].
areas of expertise include communication engineering, signal processing and
[8] Australian Government, Department of Infrastructure, Re.gional Development
networks. He has extensive experience in working with the industry for bringing
and Cities, “Northern Melbourne Smart Cities Network, enabling data to drive
research to practice as well as conducting fundamental research.
change,” [online]. Available: https://www.infrastructure.gov.au/cities/smart-cit-
ies/collaboration-platform/northern-melbourne-smart-cities-network.aspx
K e W ang is a senior lecturer and Australian Research Council (ARC) DECRA
[9] A. Al-Hourani, “On the Probability of Line-of-Sight in Urban Environments,”
Fellow at the School of Engineering, RMIT University. He has published more than
IEEE Wireless Commun. Lett., 2020, pp. 1–1.
120 papers, including more than 20 invited papers. His research interests mainly
[10] B. Al Homssi et al., “Free Spectrum for IoT: How Much Can It Take?” in 2018
include optical wireless communications, underwater wireless communications,
IEEE International Conference on Communications Workshops (ICC Work-
optical interconnects, and silicon photonic integration.
shops), May 2018, pp. 1–6.
[11] B. A. Homssi et al., “Machine Learning Framework for Sensing and Modeling
Karina Mabell Gomez is a network project manager at Milano Teleport, Italy. In
Interference in IoT Frequency Bands,” in IEEE J. Internet of Things.
2015, she served as a senior lecturer at the School of Engineering at RMIT Univer-
[12] A. Al-Hourani, “Interference Modeling in Low-altitude Unmanned Aerial Vehi-
sity, where her role was to coordinate several networking courses and supervise
cles,” IEEE Wireless Commun. Lett., 2020.
several Ph.D. and master students. She obtained her Ph.D. degree in telecommu-
[13] R. J. Rossi, Mathematical Statistics : An Introduction to Likelihood Based Infer-
nications from the University of Trento, Italy.
ence. Newark: John Wiley & Sons, 2018.

Biographies FOOTNOTES
Bassel Al Homssi is currently working toward his Ph.D. degree in the School of This work was partially funded by the Smart Cities and Suburbs Program, Department of Infrastructure, Transport,
Engineering, RMIT University, Melbourne, Australia. His research interests are Regional Development and Communications, Australian Government. Under the project, Northern Melbourne
focused on wireless sensor networks, the Internet of Things, and the practical Smart Cities Network, Enabling Data to Drive Change, grant number SCS69259.
deployment of large-scale IoT networks. 1 http://alphax.com.au/

IEEE Internet of Things Magazine • March 2021 59

Authorized licensed use limited to: University of Exeter. Downloaded on June 01,2021 at 21:42:36 UTC from IEEE Xplore. Restrictions apply.

HOMSSI_LAYOUT.indd 59 3/22/21 5:27 AM