Presence Metadata in the Internet of Things: Challenges and
Opportunities
Robert Hegarty
1
and John Haggerty
2
1
School of Science, Engineering and Environment, University of Salford, The Crescent, Salford, U.K.
2
DC2D, Beeston, Nottingham, U.K.
Keywords: Internet of Things, Security, Privacy, Metadata.
Abstract: The Internet of Things is an emerging computing paradigm that promises to revolutionise society. The
widespread capture and aggregation of data from sensors and smart devices combined with processing using
machine learning in cloud computing platforms provides unrivalled insights into our environment. In addition
to the numerous benefits (smart healthcare, cities, transportation, etc.) such insights potentially jeopardise the
privacy of individuals, organisations, and society as whole. This is despite UK and EU regulations attempting
to mitigate the risk of individuals’ data exposure and the impact of it on their security. To demonstrate the
exploitation of metadata and its threat to privacy, this paper presents Meta-Blue, a Bluetooth Low Energy
metadata capture, analysis, and visualisation tool. The results of a case study are combined with an overview
of literature on IoT privacy to provide a holistic overview of the challenges and opportunities presented by
IoT metadata.
1 INTRODUCTION
The Internet of Things (IoT) has captured the
attention of academics and industry alike. Businesses
from start-ups to multinational enterprises are
embracing and developing IoT devices and services.
As these devices and services evolve, they generate,
gather and analyse ever increasing quantities of data
and metadata about their users. This revolution in data
capture, coupled with the processing power of cloud
computing and machine learning is enabling the
realisation of highly proactive and reactive
environments such as smart cities, autonomous
vehicles and personalised health care.
Debate continues around the definition of IoT.
Whether IoT is a new paradigm or the evolution of
embedded systems, wireless networks, artificial
intelligence, data mining and other existing
technologies is arguable. Regardless of such debate,
the market for IoT devices and services is exploding.
Gartner predict the market will grow from 4.9 Billion
devices to over 20 Billion devices by 2020 (Gartner,
2017). In addition, there is exponential growth in the
amount of data collected about users that is stored,
analysed and traded by a range of services that make
use of this technology.
As with most large-scale systems, there is conflict
between the requirements of the three main
stakeholders in IoT; end users, governments, and
service providers. All have very different
perspectives and goals. End users require their data
to be secure and easily accessible via apps and web
services. Governments also require security.
However, there is increasing pressure to provide
access to data by state agencies such as law
enforcement, intelligence services, etc. which
impacts the security of the wider public. Service
providers seek to aggregate and monetise data while
balancing user privacy with regulatory and legal
compliance. Traditionally, metadata has been a grey
area in these three perspectives whereby content, such
as the text of an email, is separated from
communications data, such as events in a network
recording the sending of that email.
However, the importance of metadata has
increasingly been recognized in Europe, which was
enshrined in law by the General Data Protection
Regulations (GDPR) (European Parliament, 2016).
These regulations impose responsibilities on
collectors and processors of data, including metadata,
to ensure transparency, proportionality, and
appropriateness of data collection. Moreover, GDPR
also ensures that data collectors and processors ensure
Hegarty, R. and Haggerty, J.
Presence Metadata in the Internet of Things: Challenges and Opportunities.
DOI: 10.5220/0009094106310638
In Proceedings of the 6th International Conference on Information Systems Security and Privacy (ICISSP 2020), pages 631-638
ISBN: 978-989-758-399-5; ISSN: 2184-4356
Copyright
c
2022 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
631
the security of any personal identifying data that they
store. In addition to the legal requirements, the
diversity of technology utilised by IoT complicates
matters further.
The importance of the IoT paradigm should not be
underestimated, nor should the potential for security
and privacy challenges and benefits in this emerging
domain. This article considers issues regarding
security of IoT in general and reviews the established
and emerging protocols, infrastructure and services
used in this domain. We also identify the security
challenges and potential benefits of existing real-
world IoT systems to support our position through a
novel case study.
This paper is organised as follows. Section 2
provides background on the IoT paradigm and
considers the broader issues of metadata and privacy.
Section 3 presents an overview of recent and
emerging threats, challenges and opportunities
affecting IoT devices. Section 4 presents Meta-Blue
our tool for visualising Bluetooth Low Energy
advertising packets; along with a case study to
determine the prevalence of IoT devices that
persistently broadcast uniquely identifiable metadata
(allowing users to be tracked) and analysis of some
popular fitness tracking devices. Finally, we make our
conclusions and discuss future work.
2 BACKGROUND
Despite a large amount of literature surrounding IoT
and the effort made to define it (Minerva et al, 2015),
there is no singular, universal definition. The National
Institute of Standards and Technology have published
a report on the Network of ‘Things’ (Voas, 2016)
which encompasses the foundations of IoT:
IoT involves sensing, computing, communication
and actuation;
Things can occur in physical space (e.g. people,
vehicles, switches, smart devices, etc.) or virtual
space (e.g., data streams, software, files, etc.);
There are five primitives that describe all aspects
of IoT systems; Sensors, Aggregators,
Communication Channels, External Utilities and
Decision triggers.
As we can see, these foundations are wide and
varied, making it problematic to define security and
privacy in this domain. Despite the legal changes in
Europe and the UK, there is still no ‘silver bullet’ to
address this problem.
To develop an understanding of the importance of
IoT, we must understand its characteristics. IoT
systems are typically distributed and heterogenous.
Much of the computational power used to aggregate,
process and provide data to external utilities, such as
other services or humans, resides in the core of the
network in the form of cloud platforms. Towards the
edge of the network, computational power drops off
as we shift from intermediate aggregators, such as
smart phones or local base stations, to low powered
sensors with restricted memory. This facilitates the
widespread distribution of sensors in an energy
efficient manner.
The three-tiered architecture proposed in IEEE
P2413 (IEEE, 2015) encompasses these foundations
and is illustrated in figure 1.
Figure 1: IoT three-tiered architecture, IEEE P2413.
As figure 1 demonstrates, there are three distinct
layers with the Application layer accessing data from
the Sensing layer via the Networking and Data
Communications layer. The authors of (Minerva et al,
2015) note that the definition of IoT varies amongst
its various proponents and stakeholders. For this
reason, we shall take an evidence-based approach to
our analysis of IoT and use this analysis to provide
context to the security challenges identified in IoT
environments.
IoT devices typically serve as data capture devices
which send data to be aggregated and processed by a
cloud service. The metadata associated with IoT
devices has some similarities with conventional
computing devices; for example, it may contain; IP
address, UUID, MAC address, etc. These artefacts are
used to enable connectivity between the IoT device,
intermediate devices and cloud services.
While not all of these characteristics are
considered PII (Personally Identifiable Information)
under GDPR legislation, the proposed European
ePrivacy legislation (European Commission, 2017)
directly targets communications metadata and non-
personal data. It is hoped that this ePrivacy legislation
will bring about a change in the IoT market and
enhance the privacy of IoT communications.
Applications
Networking and Data
Communications
Sensing
ICISSP 2020 - 6th International Conference on Information Systems Security and Privacy
632
Whilst much concern in the media and wider
literature has focused on user and data security,
metadata provides a greater challenge. For example,
users cannot access it without specialist equipment,
software or knowledge and may not even be aware
that it exists. If it is not open to the users, they may
not understand where their data is ultimately collected
and stored despite legal obligations to make this
transparent. If they do not understand this, then there
is very little awareness of how their data is being
processed or mined for events, locations or relational
information associated with their activities.
In addition to metadata associated with
applications, there is other metadata that is created
due to the wide variety of protocols required for IoT
communications over and above those used more
generally in TCP/IP networking. Moreover, each step
of data transfer from the device to its ultimate storage
may add metadata, for example, MAC (Media Access
Control) addresses, IP (Internet Protocol) addresses,
UUID (Universal Unique Identifier), etc.
Bluetooth Zigbee, Z-Ware, and WiFi facilitate
short range low power communication for IoT
devices and all make use of MAC addresses to
uniquely identify devices. With the exception of
Zigbee these addresses are 48 bits in length with the
first 6 bits representing the vendor ID. Zigbee an
802.11.4 standard protocol uses 64 bits for MAC
addressing. MAC address randomisation is supported
but not mandated by Bluetooth and WiFi, however it
has been demonstrated that fingerprinting techniques
that negate MAC address randomisation and facilitate
the tracking of devices via alternative attributes used
to create fingerprints (Vanhoef et al 2016).
Bluetooth has developed over the years to enable
a wide variety of features. As of 2017 Bluetooth Low
Energy (BLE) (Bluetooth.com, 2019) is the dominant
standard adopted by most Personal Fitness Tracking
Devices (PFTDs). This subset of IoT devices is of
interest as these devices are typically always on and
attached to a human subject. Moreover, these devices
are associated with highly sensitive data, such as
health, temporal geo-location and pattern of life.
Combined with communications protocols associated
with BLE, unique personal identifying information is
generated.
Before considering the details and specification
for BLE and the emerging Bluetooth Mesh standards,
we must understand how PFTDs utilise BLE. Figure
2 illustrates how PFTDs use BLE and are integrated
into the IoT architecture.
Figure 2: PFTD architecture.
PFTDs are typically constrained devices. Their
internal architecture consists of an embedded
processor responsible for retrieving data such as step
count, temperature and heart rate from sensors and
placing it in a small onboard storage area. This
memory is used to store data for approximately one
week, depending on capabilities and Internet
connectivity, to facilitate data collection whilst the
device is disconnected from the network. To
synchronise with the cloud service, the PFTD
connects to a staging device, typically a smart phone
or computer, using BLE to transfer data. This is in
turn transferred from the staging device to the cloud
service using cellular or TCP/IP networks for
aggregation and processing. The staging device then
retrieves the results from the aggregating and
processing service to display the data in graphical or
numerical format for the user to view.
To facilitate an ad-hoc connection between the
PFTD and a staging device, the BLE component(s) on
the PFTD periodically broadcasts advertising
packets. BLE advertising packets can contain various
types of metadata including; Advertisement Address
(often a MAC address), List of Services, Service
Class UUID, Shortened Local Name, Complete Local
Name and Custom Payload (e.g. Power Level). This
metadata is used to enable staging devices to
distinguish between devices in their vicinity and
determine which one to listen to (Argenox, 2019).
It is noted that many devices send advertising
packets containing a static MAC address which is
persistently used by the device (Hilts et al, 2016).
This is despite BLE privacy features being introduced
in version 4.0 and improved in version 4.2 of the
Presence Metadata in the Internet of Things: Challenges and Opportunities
633
Bluetooth protocol. Hilts et al used RaMBLE a
Bluetooth advertisement packet sniffing tool created
by Contextis (Contextis, 2019) to gather advertising
packets from Bluetooth devices. To validate their
findings we developed Meta-Blue, a Bluetooth
sniffing tool using open source libraries. Details of
the tool and case study can be found in section 4.
3 PRESENCE METADATA
CHALLENGES AND BENEFITS
IoT provides an ever-increasing list of benefits to
society. However, the rapid deployment of complex,
heterogeneous, distributed IoT platforms has also
brought with it a variety of challenges. In the race for
market share, products have been created using the
simple formula: take an existing appliance; add a
network stack and embedded operating system; and
connect it to the Internet for control by an application
and/or service. There are some obvious caveats and
flaws to this formula that have resulted in unforeseen
consequences to the security of the devices, data
privacy and the broader Internet. The authors of
(Kambourakis, 2017), highlight this describing how
the Mirai family of malware has infected vulnerable
IoT devices causing unprecedented large scale, high
intensity Denial of Services attacks against global
Internet infrastructure, impacting not only on the end
users of IoT devices but society as a whole.
3.1 Threats and Challenges
The data aggregated by fitness tracking platforms
such as Strava, FitBit, Garmin, etc. is used to create a
competitive online health social network. These
networks focus on comparison of fitness data
gathered from end user devices. Users of these social
networks compare steps taken, calories burned, geo-
locations and times, to compete. Like conventional
social networks such as Facebook and Twitter, the
interactions between the physical and digital worlds
have resulted in some unforeseen discoveries.
For example, while analysing Strava’s global
heatmap (Strava, 2018), security researcher Nathan
Ruser identified the locations of secret military
installations around the globe (Ruser, 2018). Ruser
could infer the locations and layouts of these
installations and the routes taken by front line combat
troops on active service. This breach of operational
security sent shockwaves throughout the military
establishment.
The Strava case study is a powerful reminder that
malicious actors may seek to infer information from
the data aggregated in the core of the network by the
providers of IoT services. Our work focuses on covert
data aggregation from devices at the edge of the
network and aims to highlight some challenges and
opportunities. Such data is routinely broadcast by
millions of wearables and other IoT devices around
the world every day. In addition to fitness trackers,
BLE emissions from smart watches, glasses,
headphones, vehicles, tags, children’s toys, etc. all
contribute to these readily observable broadcasts to
provide insights into our patterns of life (Craddock,
2016).
Metadata at the edge of the network potentially
poses a broader threat to privacy than centralised
aggregated datasets controlled by cloud services
providers. Users of IoT devices e.g. fitness trackers,
have consciously opted in to these services, the
metadata is subject to terms of service, and a single
entity is accountable for storage and access control.
Metadata collection at the edge of the network
indiscriminately and covertly captures data that can
be used surveil individuals, recording and/or
predicting their movements through pattern of life
analysis. More insidiously, such data can contribute
to mass surveillance of the public by unknown
entities.
It is important to consider that the data captured
during our case study was from a single capture
device at a fixed location. A much more sinister
scenario exists when we contemplate the impact of an
attack on a smart city in which many fixed and mobile
devices could be compromised and used to track the
movement of individuals across a large metropolitan
area.
Issoufaly and Tournoux, investigate the
exploitation of a Botnet of Bluetooth Low Energy
devices, for large scale individual tracking in BLEB
(Issoufaly, 2017).
Bluetooth Mesh (Bluetooth.com, 2017) is an
emerging standard that enables Bluetooth devices to
communicate with each other over a mesh network.
This presents the opportunity for a bad actor to
potentially create or infect an app and create a botnet
of tracking nodes in the form of Bluetooth enabled
mobile phones to geo-locate individuals across a
broad area via BLE advertisement packets.
As will be demonstrated in the case study section
of this paper, many of the threats and challenges
highlighted in this sub-section are feasible and the
tools required to implement such attacks are readily
available in the public domain.
ICISSP 2020 - 6th International Conference on Information Systems Security and Privacy
634
3.2 Benefits and Opportunities
The previous sub-section focused on the privacy
challenges arising from the use of IoT devices. In this
sub-section, we consider the safety and security
benefits that IoT devices can provide. It is outside the
scope of this paper to discuss the moral, ethical or
legal issues that the following raise should they be
implemented.
An obvious benefit of using and identifying
presence metadata is in the area of disaster recovery.
For example, this data could be used to identify and
recover individuals who are trapped by serious
environmental hazards, such as buildings destroyed
by earthquakes or floods. In addition, it could be used
to help co-ordinate the emergency services.
With a potential range of 100m, it is possible to
detect the presence of devices at a not insignificant
distance for asset recovery. When searching for the
proceeds of acquisitive crime, the presence of
multiple stolen devices in a single location may be
used to identify stolen assets. This approach could be
enhanced by either embedding battery powered BLE
chips into high value items or keeping a registry of
stolen device MAC addresses. This would facilitate a
targeted search for such devices by law enforcement.
Kao et al proposed an asset tracking system that
utilises Bluetooth Low Energy and WiFi (Kao, 2017)
to assist users tracking assets in large buildings.
Following the 2016 terror attack in Manchester,
many children were left unaccounted for as they were
evacuated from the scene of the attack. IoT devices
could be used for the identification of missing
children or vulnerable adults. By simply recording the
MAC address of the devices the children were
carrying (PFTDs, mobile phones, etc.) it would have
been possible to locate them by issuing simple
instructions to the many hotels that had provided
accommodation. This approach could be further
enhanced through the implementation of a Smart City
sensor grid or Bluetooth Mesh application that
responds to missing children reports in a similar way
to the US Amber Alert system. Identification
information could be obfuscated to preserve the
privacy of missing individuals once they have been
recovered.
Many serious offenders on licence are required by
law to wear a tag to enable the enforcement of
curfews and tracking of their location. IoT devices
could aid offender probation compliance checks.
Given the range of Bluetooth devices, it would be
possible to install BLE tracking devices at the
entrance of schools to enable staff to be alerted in real
time to the presence of a sex offender and take
appropriate action.
Roadside capture of advertising packets could be
used to ascertain the congestion levels of a route.
Such data could be aggregated and combined with
existing road sensors to better measure the traffic at
key junctions or intersections. This data could be sent
to the emergency services to improve emergency
response times. Transport for London carried out a
similar trial using Wi-Fi to measure foot traffic across
various routes on the London underground (TfL,
2017).
Lin et al proposed the use of Bluetooth low energy
tags to track dementia patients and keep them safe
(Lin, 2015).
It is noted that like many tools and techniques in
cyber security, what can be used for good can also be
subverted for malicious purposes. Therefore, careful
consideration would need to be made about
implementation of the examples above. The use of
ephemeral keys and end to end encryption in the
broadcast of hashed device identities and encrypted
device locations is used by some mobile device
tracking systems (Apple, 2020) to facilitate recovery
of devices while preventing unauthorised tracking.
Such an approach may be adapted to facilitate the
tracking of individuals in the scenarios previously
described.
4 META-BLUE CASE STUDY
This section demonstrates the extent to which PFTDs
advertising broadcasts may be employed to identify
individuals through presence metadata. The case
study captured and collated advertising packets from
BLE devices to determine if it is feasible to identify
an individual device over a sustained period.
To capture BLE advertising packets and validate
the results we developed Meta-Blue (Meta-Blue
2019), and licenced it freely under GNU 2.0 with the
hope that other academics and researchers will use it
help spread awareness on IoT security and privacy
issues. Meta-Blue uses the open source library
pyBluez (pyBluez, 2015) based on the Bluez
Bluetooth official Linux Bluetooth implementation
(Bluez, 2019). The tool captures and stores MAC
addresses from BLE advertising packets. The stored
MAC addresses are processed and visualised using
the MatPlotlib library (MatPlotLib, 2019) and
illustrated in figure 3.
Presence Metadata in the Internet of Things: Challenges and Opportunities
635
Figure 3: BLE device visualisation.
The plots in Figure 3 visualise the total number of
devices visible for each iteration of the scan, and the
device identification numbers associated with each
device where device IDs are derived from MAC
addresses. The scan was captured at the beginning of
a lecture session and more than 25 unique devices
where recorded over the duration of the capture. Even
without the context being supplied it is evident that
many devices arrived in the location in a short period
of time and could be individually identified.
In order to replicate this first experiment in a
public setting, scans were conducted at a railway
station. This is a more likely scenario whereby
individuals were not known to the researchers. The
plots in Figure 4 illustrate the arrival and departure of
trains at a major UK railway station in Manchester.
Over 400 unique devices where recorded over a 10-
minute time period.
Figure 4: Train arrival and departure visualised with Meta-
Blue.
From the capture alone, figure 4 demonstrates that
it is possible to infer the arrival and departure of
individuals by train, as indicated by the peaks and
toughs in upper graph (indicating the number of
devices visible at each time segment), and addition or
reduction of coloured markers in the lower graph that
represent individual devices. For example, between
100 and 200 seconds in the capture, train arrival is
marked by an increase in total devices, and the
emergence then disappearance of device ID’s in the
250 to 300 range, illustrating the transit of 50
devices/individuals past the capture device in the
busy railway station. Nearly 500 devices were
observed in the 10 minute (600 second) capture,
illustrating the popularity of BLE devices.
To better understand the circumstances under
which BLE devices persistently broadcast static
metadata a separate visualisation was created using
Meta-Blue and control devices. Figure 5 illustrates
the results of this experiment.
Figure 5: Control device visualisation with Meta-Blue.
Two control PFTDs were monitored for a fixed
period of time. Each of the devices was periodically
connected to its staging device, a mobile phone, to
synchronise data. The first control device with ID 1
persistently broadcast its MAC address regardless of
its connection with the staging device. The second
control device with ID’s 3, 6, 8 broadcast different
ID’s each time it was connected and disconnected
from its staging device, demonstrating adherence to
the Bluetooth LE Privacy standard. Devices ID’s 2, 4,
5, 7, 9, 10, 11 belong to unknown devices transiting
the vicinity of the experiment (demonstrating the
prevalence of BLE devices). It may be the case that
device ID’s 2 and 11 related to the same device,
however this could not be confirmed as these
broadcasts were not from the control devices. Both of
the control devices where periodically monitored
using Meta-Blue and RaMBLE (Contextis, 2019)
over a period of months. Despite power cycling and
battery depletion, device ID 1 never changed the
MAC address that was persistently used and
broadcast for months. It is clear from the data in
Figure 4, and through other evidence gathered using
Meta-Blue and RaMBLE (Contextis, 2019), that
many devices opt to broadcast persistent MAC
addresses.
As demonstrated by this case study, it is possible
to identify unique devices through presence metadata.
This could be further exploited to impact an
individual’s security and privacy by correlating this
metadata with other data.
ICISSP 2020 - 6th International Conference on Information Systems Security and Privacy
636
5 CONCLUSIONS AND FUTURE
WORK
BLE devices present a number of privacy and security
issues, both of which are at the heart of recently
proposed EU regulatory control. First and foremost,
devices are powered on by default and often cannot
be turned off. As such, most users are unlikely to be
aware of the advertising packets frequently broadcast
by their devices and their ability to identify individual
devices. This data can be used to identify personal
identifying information, particularly when correlated
with other data. In addition, users may not know
whether high standards in data encryption are being
employed by the device’s software or turned on at all.
Moreover, in many cases it is unclear as to how or
where the data is stored, or the level of encryption
and/or anonymization applied to discrete or
aggregated data stored in the cloud by the provider.
This goes against the transparency required by
regulatory bodies and may have serious implications
for controllers and processors of such data.
It is important that when considering any of the
scenarios discussed in this paper that MAC address
collisions can occur. While they are rare, they could
have a significant impact in some use cases. As such,
when employed in cases where identification of an
individual device has significant consequences, a
secondary check should be carried out to validate the
presence of a device or individual. This will be
addressed in future work.
The emergence of a Bluetooth Mesh standard will
enable existing BLE devices to communicate via a
mesh thereby significantly increasing the range at
which a device can be detected. This has implications
for both the security challenges and benefits
considered in this paper. Further research is also
required to determine the consequences of the
Bluetooth Mesh standard. As with the BLE standard
the way manufacturers implement the new standard
will play a key role in determining the privacy and
security of users. In addition to the privacy risks
outlined in this paper, the IoT poses a large threat to
societal privacy and trust. A much broader range of
threats to privacy are emerging as IoT matures; to
give a single example; private corporations are
constructing large scale unregulated surveillance
networks, marketed as a feature of smart connected
door bells. Aside from the recent attacks on these
devices and potential for them to disrupt the Internet
through Mirai type attacks. The threat to the
Universal Declaration of Human Rights Article 12
(United Nations, 1948) the right to privacy, by private
corporations with global reach demonstrates that
further regulation or enforcement of existing
legislation is required to balance the interests of the
market and the privacy of the individual.
REFERENCES
Hung, M., Gartner, 2017, Leading the IoT,
https://www.gartner.com/imagesrv/books/iot/iotEbook
_digital.pdf, accessed November 2019
European Parliament, 2016 , General Data Protection
Regulation, https://eur-lex.europa.eu/legal-content/EN/
TXT/PDF/?uri=CELEX:32016R0679 , accessed
November 2019
Minerva, R., Biru, A., Rotondi, D., 2015 Towards a
definition of the Internet of Things (IoT), IEEE Internet
of Things
ePrivacy Proposal https://eur-lex.europa.eu/legal-
content/EN/TXT/HTML/?uri=CELEX:52017PC0010
&from=EN accessed January 2020
Voas, J., 2016, Network of ‘Things’, NIST Special
Publication 900-183
www.bluetooth.com, accessed November 2019
https://www.argenox.com/a-ble-advertising-primer/ ,
accessed November 2019
Hilts, A., Parsons, C., Knockel, J., 2016, Every Step You
Fake: A Comparative Analysis of Fitness Tracker
Privacy and Security, Open Effect Report
https://www.contextis.com/en/resources/tools/ramble-ble-
app , accessed November 2019
Kambourakis, G., Kolias, A., 2017, The Mirai botnet of the
IoT Zombie Armies, IEEE Military Communications
Conference MILCOM.
https://www.strava.com/heatmap , accessed November
2019
Ruser, N., 2018, https://twitter.com/Nrg8000, accessed
November 2019
Craddock R., Watson D., Saunders W., 2016 Generic
Pattern of Life and behaviour analysis, IEEE
International Multi-Disciplinary Conference on
Cognitive Methods in Situation Awareness and
Decision Support (CogSIMA)
Issoufaly, T., Tournoux, P., 2017, BLEB : Bleutooth Lowe
Energy Botnet for large scale individual tracking, 1
st
International Conference on Next Generation
Computing Applications (NextComp)
Vanhoef, M., Matte, C., Cunche, M., Cardoso, L., Piessens,
F., Why MAC Address Randomization is not Enough:
An Analysis of Wi-Fi Network Discovery Mechanisms,
ASIA CCS ’16 Proceedings of the 11
th
ACM on Asia
Conference on Computer and Communications
Security, 2016.
https://www.bluetooth.com/blog/introducing-bluetooth-
mesh-networking/ , accessed November 2019
Kao C., Hsiano, R., Chen, P., Pan, M., 2017, A Hybrid
indoor positioning for asset tracking using Bluetooth
low energy and Wi-Fi, IEEE International Conference
of Consumer Electronics, Taiwan (ICCE-TW)
Presence Metadata in the Internet of Things: Challenges and Opportunities
637
Sager Weinstein, L., 2017, Review of the TfL WiFi Pilot,
http://content.tfl.gov.uk/review-tfl-wifi-pilot.pdf ,
accessed November 2019
Lin, Y., Chen, M., 2015, A cloud-based Bluetooth low
energy tracking system for dementia patients, 6
th
International Conference on Mobile Computing and
Ubiquitous Networking (ICMU)
Apple Find My, https://www.apple.com/icloud/find-my/,
accessed 2020
Hegarty, R., Haggerty J., https://github.com/arc2h/Meta-
Blue accessed January 2020
Karulis, P., 2015, https://github.com/karulis/pybluez ,
accessed November 2019
Bluez Official Linux Bluetooth protocol Stack,
http://www.bluez.org/ , accessed November 2019
Matplotlib, https://matplotlib.org/ , accessed November
2019
United Nations, Universal Declaration of Human Rights
1948, https://www.un.org/en/universal-declaration-
human-rights/ , accessed January 2020
ICISSP 2020 - 6th International Conference on Information Systems Security and Privacy
638