Privacy Aura for Transparent Authentication on Multiple Smart Devices
Takoua Guiga
1,2
, Jean-Jacques Schwartzmann
1,2
and Christophe Rosenberger
2 a
1
Orange Labs, France
2
Normandie Univ., UNICAEN, ENSICAEN, CNRS, GREYC, 14000 Caen, France
Keywords:
Transparent Authentication, Aura, Privacy, Multi-devices.
Abstract:
Nowadays, users carry multiple connected devices such as a smartphone, laptop, connected watch. . . . Security
constraints limit user’s usability especially when using all of them intensively during the day (social media,
work). In this paper, we propose the privacy Aura concept corresponding to the circle of trust in the
neighborhood of each smart device to facilitate user authentication. Many data (phone calls habits, biometrics,
localization) can be collected to realize a transparent and privacy compliant authentication on each device. The
confidence on user authentication on each device can be transferred to another one if it is located in the same
Aura. This is the main contribution of the paper. We show through illustrations the benefit of the proposed
solution.
1 INTRODUCTION
Research on innovative solutions for user
authentication is very active. However, while
people are daily surrounded by different and multiple
devices, they are still using one single device
to access or protect their electronic equipment.
Despite the significant rise of connected objects
and Internet of Things (IoT), few are interested in
the user experience taken in the totality of his/her
digital interactions. If a device is compromised
by attackers, security is no longer guaranteed and
attackers can easily gain access to user personal data.
Moreover, the authentication task requires repetitive
interventions by the user as he/she has to act with
different devices in its neighborhood in order to prove
his/her identity to each one. With the multiplicity
of authentication factors and the diversity of owned
terminals, this action becomes painful and disruptive,
creates stress, wastes time and clutters our daily lives
with unnecessary tasks. With the intention of getting
rid of the superfluous in everyday life and to ensure
better security and privacy, we want to consider the
multiple devices of the user in the authentication
process and to delegate the authentication task to
all of them. So, in case of compromise, all the
devices create together a strong circle of trust, so
that even when a device is stolen, it should be able
to protect user’s personal data. Subsequently, this
a
https://orcid.org/0000-0002-2042-9029
multi-devices circle is called Authentication Aura.
This concept is not new as it has been proposed
in (Hocking et al., 2011), we propose in this paper
to extend some notions especially with a great
focus on privacy. In this respect, in a ubiquitous
digital environment, we propose to provide a privacy
transparent authentication to the user throughout
his/her day while ensuring a good privacy protection.
The paper is organized as follows. Section 2 is
dedicated to the related works in the area on user
authentication on multiple devices. In section 3, we
detail the proposed method, describing the concept,
the process service and the privacy protection aspect.
Section 4 provides some illustrations showing the
benefit of the proposed approach. We conclude in
section 5 and give some perspectives of this work.
2 RELATED WORKS
Over the recent years, it has been proven that Internet
of Things (IoT) has the potential to make a society-
wide impact by changing diverse sectors but also
our daily lives. Despite the impressive growth
of IoT-connected devices number, as mentioned
in the latest Juniper Research (Juniper, 2020)
claiming that the number of IoT-connected devices
will reach 83 billion by 2024, rising from 35
billion connections in 2020, few research works are
focusing on involving multi-devices authentication
Guiga, T., Schwartzmann, J. and Rosenberger, C.
Privacy Aura for Transparent Authentication on Multiple Smart Devices.
DOI: 10.5220/0010576407410746
In Proceedings of the 18th International Conference on Security and Cryptography (SECRYPT 2021), pages 741-746
ISBN: 978-989-758-524-1
Copyright
c
 2021 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
741

solutions. Multi-devices authentication, particularly
the authentication Aura concept was firstly introduced
by Hocking et al.(Hocking et al., 2011) as a new
approach to identity authentication on mobile devices
based upon a framework that can transparently
improve user security confidence. Information
pertaining to user authentication is shared amongst
the user’ devices, collectively enabling a near field
adaptive security envelope to be established and
maintained around the user. We can note that
privacy protection was not considered in this work
which is an important drawback once the General
Data Protection Regulation (GDPR) is in place now
for European citizens. Riva et al.(Riva et al.,
2012) presented progressive authentication based on
associating multiple sources of authentication data. A
two-device authentication model is proposed by Cha
et al. (Cha et al., 2015) for micro-payment systems
using a mobile and wearable devices. Xu (Xu, 2015)
focuses on biometric authentication using wearable,
namely on face recognition using smart-glass and
gait recognition using a smart watch. Furthermore,
in order to control and secure the access to the
storage services based on the cloud, Gonzalez et al.
(Gonzalez-Manzano et al., 2015) propose a multi-
devices solution with a symmetric cryptographic
scheme. Hajny et al. (Hajny et al., 2016)
presented also a cryptographic scheme providing a
multi-devices authentication using wearable and IoT.
Move2auth, a proximity-based mechanism for IoT
device authentication was introduced by Zhang et
al. (Zhang et al., 2017) based on performing hand
gestures (moving towards and away, and rotating)
to detect proximity and authenticate IoT devices.
Nevertheless, shown results are limited on a single
device which is the smartphone and there is no
further analysis on other IoT devices. All these
approaches lack more details about data privacy
protection. On the basis of this brief overview of the
literature, we can therefore agree that multi-devices
authentication solutions are both limited in number
and in consideration of privacy protection. Existing
solutions do not focus at the same time on usability,
security and privacy. In the next section, we propose a
new solution extending the Aura concept by (Hocking
et al., 2011). The proposed solution defines a new
trust party service to Internet users respecting GDPR
requirements with a great focus on usability. We
describe the concept and all the steps of its usage in
the following section.
3 PROPOSED METHOD
Authentication is definitely an essential and important
task to secure access to our devices. At the same time,
it is repetitive and painful for the user and in some
cases vulnerable to attacks. Our goal in this work is
to make the authentication task as simple as possible
for the user while respecting his/her privacy and data
security. In this section, we first introduce the concept
of the authentication Aura we use in this work.
3.1 Concept Description
Surrounded by digital technology, human beings
are actively interacting with digital devices during
their daily life. We call this interaction digital
aura translating the communication between the
user and his/her digital devices through a mutual
authentication. In a typical authentication scenario,
the user has to realize an explicit action each
time he/she wants to authenticate to all owned
devices, by presenting a different code for each
device, which makes ten or even twenty codes
to be learned and typed hundred times a day, in
order to authorize access. Our approach aims to
create a multi-devices authentication system based
on mutual communication between devices thanks
to a trust party service (see Figure 1). We assume
in this work that each device realizes a transparent
authentication of the user. The idea is to define a
confidence architecture between the different smart
objects of the user, capable of continuously collecting
behavioural or morphological data. This data must
enable continuous authentication of the user. In
other words, we wish to ensure sufficient interactions
between a user and its devices to guarantee a high
level of trust that could be transferred between them.
Let’s consider the following definitions:
• A: the authentication Aura of user U,
• O
i
: a device ∈ A, with i ∈ [1, n
d
], n
d
is the number
of devices of the user U,
• C(O
i
): Confidence in a device O
i
. It is computed
at any time with a transparent authentication
solution based on many factors (passwords,
biometrics, geolocation. . . ).
We consider that each device has its own
Aura, that we call Aura device AO
i
. Two
types of information are sent to the trust party
service: some data are transmitted to compute the
confidence level associated to a device (transparent
authentication scheme) and geolocation information
are also transmitted in order to update the confidence
level of a device O
i
if it is in the Aura of a device O
j
SECRYPT 2021 - 18th International Conference on Security and Cryptography
742

Figure 1: Principle of the proposed method: A trust party
realizes the monitoring of the confidence level of each
device at any time.
(i 6= j). The amount of confidence transferred to the
device O
j
is a ratio of the confidence level associated
to O
i
and also depends on the proximity of both
devices. Figure 2 illustrates this process. Sending this
information could be a security problem in case of
interception by an attacker or if the service is honest
but curious. We propose a privacy protection scheme
that enables user’s personal information protection
and let the service compute the confidence level
without knowing which type of information has been
used for transparent authentication.
Figure 2: Illustration: The user’s smartphone has an Aura
where the confidence is monitored by the trust party. If
the computer is located in the smartphone Aura (not far in
some sense), a part of the confidence level associated to the
smartphone can be transferred to the computer one.
3.2 Privacy Protection
Privacy protection is a main and important issue
to consider in our approach. As the verification
process could be done by a trust party considered
as honest but curious, a privacy protection of data
coming from all devices is required. The concept of
privacy protection of biometric data has been defined
in 2001 in a seminal paper (Ratha et al., 2001).
Since then, many methods have been proposed among
random projections approaches (Pillai et al., 2010),
BioHashing methods (Teoh et al., 2004), Bloom
filters (Rathgeb et al., 2014), to cite just a few.
The BioHashing algorithm is applied on biometric
templates that are represented by real-valued vectors
of fixed length (so the metric used to evaluate the
similarity between two biometric features is the
Euclidean distance). It generates binary templates of
length lower than or equal to the original length (here,
the metric D
T
used to evaluate the similarity between
two transformed templates is the Hamming distance).
This algorithm has been originally proposed for face
and fingerprints by Teoh et al. in (Jin et al.,
2004). Then, the BioHashing algorithm transforms
the biometric template T = (T
1
, .. . T
n
) into a binary
template B = (B
1
, . . . B
m
), with m ≤ n in Algorithm 1.
A complete review of cancelable biometric systems
can be found in (Patel et al., 2015).
More generally, a security analysis of the
biometric system protecting the biometric template
based on transformations (Rosenberger, 2018) are
considered. The specificity of the BioHashing
algorithm is that it uses a one way function and a
random seed of m bits. It is important to note that
every behavioral feature uses a different seed in order
to create a specific BioCode. The performance of this
algorithm is ensured by the scalar products with the
orthonormal vectors. The quantization process of the
last step ensures the non-invertibility of the data (even
if n = m, because each coordinate of the input T is a
real value, whereas the coordinates of the output B is
a single bit). Finally, the random seed guarantees both
the diversity and revocability properties.
Algorithm 1: BioHashing.
1: Inputs
2: T = (T
1
, . . . , T
n
): biometric template,
3: K
z
: secret seed
4: Output B = (B
1
, . . . , B
m
): BioCode
5: Generation with the seed K
z
of m pseudorandom
vectors V
1
, . . . ,V
m
of length n,
6: Orthogonalize vectors with the Gram-Schmidt
algorithm,
7: for i = 1, . . . , m do compute x
i
=< T,V
i
>.
8: end for
9: Compute BioCode:
B
i
=

0 if x
i
< τ
1 if x
i
≥ τ,
where τ is a given threshold, generally equal to 0.
3.3 Process Service Description
Let’s consider a user U, having a number of devices n
d
and a number of hotspots nh he/she defined a priori.
A hotspot is a trusted area such as home or work
place. This user has the possibility to register his/her
devices and hotspots via an application provided by
a trust party, by entering the IP address of each of
possessed devices and the GPS coordinates of his/her
hotspots. Once registered on this application, they
Privacy Aura for Transparent Authentication on Multiple Smart Devices
743

are considered as trusted devices and hotspots. The
confidence on a hotspot (denoted by NAH) is variable
and is set by the user. For example, in the case of a
fully trusted hotspot (such as my home), NAH can
be set to 100%. We suppose that the user has on
each of its device a transparent authentication solution
that allows a trust party to compute a confidence level
guarantying that at any time, the device is used by
the right user. A solution proposed by authors of
(Guiga et al., 2020) could be used. This confidence
level evolves with time i.e. the device sends at a set
interval of time (as for example, each 3 minutes) some
information (behavior, face biometric data. . . ) to the
trust service for the confidence computation. This
confidence value decreases automatically by the trust
party to ensure that if the device is not used by the
legitimate user, it cannot be used by an impostor. In
this work, we want at any time, to determine the trust
level of each device by taking into account its Aura.
This level can be increased based on belonging the
trusted hotspots where the user and device are located.
For this purpose, we illustrate the process
considering the following scenario. Let Alice be
the user with 3 devices: a smartphone (S), a
laptop (L) and an office computer (PC). In order to
correctly authenticate herself to her devices, Alice
uses passwords, either the same password for all 3
devices or different passwords for each device. In
both cases, this authentication method is weak and
vulnerable to different attacks. We wish to establish
a connection between Alice’s devices in order to
allow the transfer a part of the confidence level on
authentication from one to another device without
Alice’s intervention. Alice creates with her devices
an authentication aura A. Let AO
1
be the smartphone
aura, AO
2
the laptop aura and AO
3
the computer
aura. Let A
T
be Alice’s Aura in trusted hotspots, in
order to compute the confidence of Alice’s device (the
smartphone as for example), we first want to check
if it belongs to A
T
(i.e. if this device is in a trusted
hotspot). We consider the confidence C(O
i
) of the
device O
i
calculated individually (not in the hotspot).
We also define the confidence C
H
(O
i
) of the device O
i
belonging to a hotspot by the minimum of the sum of
the confidence products of a device O
j
(for any i6= j)
belonging to the same hotspot (having the trust NAH)
and 100 (the maximal value of a confidence), given
by the following equation:
C
H
(O
i
)= min (
n
H
∑
i6= j
NAH ×C(O
j
)+ C(O
i
), 100) (1)
With n
H
the number of Alice’s devices belonging
to the same hotspot. This value is updated at each
interval of time set by the trust service or user. We
assume that the initial confidence of a new device to
which Alice wishes to authenticate is zero. In order
to determine the possible transfer of confidence level
among devices, we need to verify if they are located
on the same hotspot. To achieve this goal, we can
measure their geolocation using many data such as
GPS coordinates, IP address or via the WIFI list.
Let(g
1
, g
2
,...,g
n
) be the geolocation data of Alice’s
smartphone. Then, to ensure the security and privacy
of the geolocation data, we apply the BioHashing
algorithm to generate a geolocation Biocode with
Alice’s secret key (here, a random seed value). In
order to decide if the device is located in a known
trusted hotspot, we compute the Hamming distance
dist
H
between the Biocodes of the device geolocation
data and the hotspot. The level of proximity to a
hotspot is defined by the value of the Hamming
distance. Among a decision threshold set by the trust
party (risk management), the trust service can decide
if the device is located in one of Alice’s trusted
hotspots. Note that the trust service is not able to
know where is located this hotspot as it only knows
its geolocation BioCode and do not know Alice’s
secret key. Figure 3, illustrates the adopted scenario.
Figure 3: Illustration of the adopted scenario to determine
the presence or not in a trusted hotspot.
For a number of devices n
H
= 2, let’s imagine a
typical day in Alice’s life described by the following
scenario: Alice uses her smartphone every morning
to read the news and consult her social networks,
with the time spent logging on, her smartphone
gains more confidence on authentication (transparent
authentication). On her way to work, she continues
to use her smartphone to call her mum. Now, when
she arrives at her office, that has been previously
declared as a high trusted hotspot, she wants to
authenticate to her laptop. Having a sufficient level of
confidence on her smartphone, she can use her laptop
without the need for a re-authentication (transparent
authentication gained with the confidence transferred
from the smartphone). The confidence level C
H
(L)
of her laptop can be calculated by the equation given
SECRYPT 2021 - 18th International Conference on Security and Cryptography
744

Figure 4: Alice’s Aura confidence evolution curve Vs Alice’s smartphone, Laptop and computer confidence evolution curve
along a day.
below 2:
C
H
(L)= min (NAH× C(S) + C(L), 100) (2)
In case Alice uses her smartphone, but not
enough to establish a confidence to be shared with
other devices, another proof of authentication can be
requested (a PIN code for example). We can see that,
when trying to authenticate with a new device, we
do not loose the confidence level already acquired on
other devices in the Aura. So, initially, the new device
gains the confidence of the whole Aura, whatever the
type of hotspot. Note that the confidence associated
to a device is updated at each interval of time, with
the transfer from other devices (it could be at the
previous time not equal to zero). In the next section,
we illustrate the proposed method on simulated data.
4 ILLUSTRATIONS
In order to study the confidence evolution of
Alice’s devices, We consider a specific scenario
describing a typical day of Alice, mainly while
using her smartphone, laptop and computer,
where Alice has declared 4 trusted hotspots as
following:Home(NAH=100),Office(NAH=90),
Parents home(NAH=75),Cafeteria(NAH=50). We
set in this illustration the natural decrease in the
confidence level over time with an exponential decay
(interval of time when data are sent to the trust
party). We can consider figure 4 showing the impact
of the Privacy Authentication Aura solution on the
confidence level of Alice’s devices, respectively
on her smartphone, office computer and laptop vs
the confidence level on Alice’s devices, along the
day, in a transparent mono-device authentication
context. We can clearly see that the Aura confidence
curve is at least equal to the confidence curve of one
device without applying the Aura solution. This is
normal as defined in equation 1. In this work, we
only considered the positive impact of the Aura on
the confidence level on authentication on a device.
Having other devices in the proximity or belonging
to the same hotspot, leads to a higher confidence
level, thanks to the transfer of confidence between
devices. With this solution, we do not require another
authentication request. Devices take benefit of the
belonging to the same hotspot and of the transparent
authentication acquired on one device. For example,
we can see on figure 4, when Alice is in her office
at t=170 (i.e. 8:30 am as we have 20 intervals
per hour), the confidence level without the Aura
solution is of 23% which is not considered enough
to be authenticated, in case Alice wants to use her
smartphone, another proof is requested. However,
when using the Aura solution, we can see that the
confidence level at t=170 is equal to 80%, which
allows Alice to use her phone without any additional
proof, thanks to the proximity with her computer
and being in a trusted hotspot. Another example,
when Alice goes to visit her parents, she has her
laptop in her bag but she rarely use it, so at t=360, the
Privacy Aura for Transparent Authentication on Multiple Smart Devices
745

confidence level without the Aura concept is almost
equal to zero. Yet, if she decides to use it, she has
already gained on confidence level which increases to
69% when using the Aura solution by just being at her
parents home which is declared as a trusted hotspot
and by relying on the confidence on her smartphone
over the time spent using it. In terms of privacy,
the trust party, in order to authorize authentication,
collects information to know whether a device is in a
specific hotspot or not, but it is not allowed to know
the content of data. In our case, the trust party has
no right to have access to the geolocation data. It
receives only the Biocodes BG because all collected
data are protected by the Biohashing algorithm as
mentioned in the section 3.2. So, the trust party can
be informed of Alice’s presence in a trusted hotspot
without knowing exactly where she is. Noting that
the mono-device transparent authentication privacy
is respected as well, and we can refer to this work
(Guiga et al., 2020) for more details.
5 CONCLUSION
We proposed in this paper a multidevices
transparent authentication solution, called Privacy
Authentication Aura, that improves the confidence
level authentication comparing to a mono-device
solution and ensures data privacy protection. A
higher confidence is provided by the Aura when
devices are located at the same trusted hotspot and
it can be transferred from a device to another. It is
true that in our process, the confidence decreases
over time, but to keep transparency, it cannot be
decreased abruptly, and, in fact, this can lead to
intrusion attacks. Therefore, we aim to improve our
process so the user can be alerted when one of her
devices is not detected in the Aura but still have a
high confidence. The user can decide to decrease the
confidence of a device if it is not located in the same
Aura. This process is classical to detect payment
frauds (as for example, detecting a withdrawal of
money in a foreign country), we plan to improve the
proposed solution with a negative impact of the Aura
on the authentication confidence level.
REFERENCES
Cha, B.-R., Lee, S.-H., Park, S.-B., Ji, G.-K. L. Y.-K.,
et al. (2015). Design of micro-payment to strengthen
security by 2 factor authentication with mobile &
wearable devices. Advanced Science and Technology
Letters, 109(7):28–32.
Gonzalez-Manzano, L., de Fuentes, J. M., and Orfila,
A. (2015). Access control for the cloud based
on multi-device authentication. In 2015 IEEE
Trustcom/BigDataSE/ISPA, volume 1, pages 856–
863. IEEE.
Guiga, T., Rosenberger, C., and Schwartzmann, J.-J. (2020).
When my behavior enhances my smartphone security.
In 2020 International Conference on Cyberworlds
(CW), pages 280–284. IEEE.
Hajny, J., Dzurenda, P., and Malina, L. (2016). Multi-device
authentication using wearables and iot. In SECRYPT,
pages 483–488.
Hocking, C. G., Furnell, S. M., Clarke, N. L., and
Reynolds, P. L. (2011). Authentication aura-a
distributed approach to user authentication. Journal of
Information Assurance and Security, 6(2):149–156.
Jin, A. T. B., Ling, D. N. C., and Goh, A. (2004).
Biohashing: two factor authentication featuring
fingerprint data and tokenised random number.
Pattern recognition, 37(11):2245–2255.
Juniper (2020). Juniper research press releases,
”iot connections to reach 83 billion by 2024,
driven by maturing industrial use cases”. https:
//www.juniperresearch.com/press/press-releases/
iot-connections-to-reach-83-billion-by-2024-driven.
[Online; accessed 27-January-2021].
Patel, V. M., Ratha, N. K., and Chellappa, R. (2015).
Cancelable biometrics: A review. IEEE Signal
Processing Magazine, 32(5):54–65.
Pillai, J. K., Patel, V. M., Chellappa, R., and Ratha,
N. K. (2010). Sectored random projections
for cancelable iris biometrics. In 2010 IEEE
International Conference on Acoustics, Speech and
Signal Processing, pages 1838–1841. IEEE.
Ratha, N. K., Connell, J. H., and Bolle, R. M. (2001).
Enhancing security and privacy in biometrics-based
authentication systems. IBM systems Journal,
40(3):614–634.
Rathgeb, C., Breitinger, F., Busch, C., and Baier, H. (2014).
On application of bloom filters to iris biometrics. IET
Biometrics, 3(4):207–218.
Riva, O., Qin, C., Strauss, K., and Lymberopoulos, D.
(2012). Progressive authentication: deciding when to
authenticate on mobile phones. In 21st {USENIX}
Security Symposium ({USENIX} Security 12), pages
301–316.
Rosenberger, C. (2018). Evaluation of biometric template
protection schemes based on a transformation. In
ICISSP, pages 216–224.
Teoh, A., Ngo, D., and Goh, A. (2004). Biohashing: two
factor authentication featuring fingerprint data and
tokenised random number. Pattern recognition, 40.
Xu, W. (2015). Mobile applications based on smart
wearable devices. In Proceedings of the 13th ACM
Conference on Embedded Networked Sensor Systems,
pages 505–506.
Zhang, J., Wang, Z., Yang, Z., and Zhang, Q. (2017).
Proximity based iot device authentication. In IEEE
INFOCOM 2017-IEEE Conference on Computer
Communications, pages 1–9. IEEE.
SECRYPT 2021 - 18th International Conference on Security and Cryptography
746