Implementing Secure Applications
Thanks to an Integrated Secure Element
Sylvain Guilley
1,2,3 a
, Michel Le Rolland
4
and Damien Quenson
4
1
Secure-IC S.A.S., Tour Montparnasse, 27th floor, 75015 Paris, France
2
TELECOM-Paris, Institut Polytechnique de Paris, 19 Place Marguerite Perey, 91120 Palaiseau, France
3
École Normale Supérieure, 45 rue d’Ulm, Département d’informatique, CNRS, PSL University, 75005 Paris, France
4
Secure-IC S.A.S. Headquarters, 15 rue Claude Chappe, Bât. B, 35510 Cesson-Sévigné, France
Keywords:
Integrated Secure Element, Host Protection, Certification Schemes, Security Flexibility, Internet of Things.
Abstract:
More and more networked applications require security, with keys managed at the end-point. However, tradi-
tional Secure Elements have not been designed to be connected. There is thus a need to bridge the gap, and
novel kinds of Secure Elements have been introduced in this respect.
Connectivity has made it possible for a single chip to implement multiple usages. For instance, in a smart-
phone, security is about preventing the device from being rooted, but also about enabling user’s online privacy.
Therefore, Secure Elements shall be compatible with multiple requirements for various vertical markets (e.g.,
payment, contents protection, automotive, etc.). The solution to this versatility is the integration of the Secure
Element within the device main chip. Such approach, termed iSE (integrated Secure Element), consists in the
implementation of a subsystem, endowed to manage the chip security, within a host System-on-Chip.
The iSE offers flexibility in the security deployment. However, natural questions that arise are: how to pro-
gram security applications using an iSE? How to certify those applications, most likely according to several
different schemes?
This position paper addresses those questions, and comes up with some key concepts of on-chip security, in
terms of iSE secure usage. In particular, we will show in this paper that iSE nowadays shall be designed so that
the product it embeds is certifiable in a multiplicity of schemes, and so even before the product is launched on
the market.
1 INTRODUCTION
We are witnessing the rapid development of our dig-
ital society, which is characterized by the explosion
of the number of handheld and IoT devices. In this
context, the security of the end-points is essential, as
those objects represent an obvious entry point for at-
tackers into the whole platform. Actually, the end-
points are nowadays entrusted to secure the link start-
ing from the device up to the cloud. Such end-to-end
security is achieved by designing a solid root-of-trust
which stems from the device. It is the role of the “Se-
cure Element” to implement such fundamental secu-
rity foundation.
From a historical perspective, the Secure Element
has known several successive development stages:
It was initially discrete, and designed mainly to
a
https://orcid.org/0000-0002-5044-3534
defend themselves (i.e., it features anti-tampering
capabilities);
Then, it has become more aware of the security
of the application, and has thus been termed “em-
bedded”. It may consist in legacy discrete Secure
Element, but adapted to smaller form factors and
implemented closer to the circuit it protects;
Eventually, some Secure Elements have become
integrated into the main chip of the end-point de-
vice. Consequently, it got melted as a subsystem
within the host chip it protects.
The Secure Elements, under those three avatars,
enforce end-point security principles. Namely, they
implement:
Minimum Surface Exposition: in that the inter-
face of a Secure Element is reduced to as few pins
and functions as possible. For instance, one can
remember ISO/IEC 7816-2 standard (ISO/IEC
566
Guilley, S., Rolland, M. and Quenson, D.
Implementing Secure Applications Thanks to an Integrated Secure Element.
DOI: 10.5220/0010298205660571
In Proceedings of the 7th International Conference on Information Systems Security and Privacy (ICISSP 2021), pages 566-571
ISBN: 978-989-758-491-6
Copyright
c
2021 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
7816, 2014), published back in the late 80’s,
which required compliant smartcards to have only
one single in/out (IO) pin for the communication
with the external world, in half-duplex.
Defense in Depth: Secure Elements implement
several layers of protection. For instance, the Se-
cure Element is typically entrenched behind a de-
fensive layer, be it physical (e.g., SPI or Serial
line) or logical (e.g., protocol break through mail-
boxes).
Minimal Complexity Rationale: the control logic
in Secure Elements is crafted to be focused on its
security mission, and is thus not polluted by in-
terference with other businesses, which are left to
the “rich world”.
The proper implementation of those security princi-
ples is verified through a process known as “certi-
fication”, whereby a third party laboratory assesses
the product documentation and performs a counter-
expertise in terms of offensive security testing.
In this position paper, we account for the raise of
the integrated Secure Element (referred to as “iSE” in
the sequel), and explain the challenges linked with its
adoption. The argumentation is oriented towards ex-
plaining the stakes regarding security requirements,
the technological evolution (connected devices, life-
cycle management, etc.) and the moving normaliza-
tion landscape. The targeted audience is the security
solution architects aiming at gaining a pedagogical
overview about end-point security trends.
Contributions. We list the advantages of the iSE
over the traditional approach using a Secure Element
located outside of the main end-point chip. Then, we
discuss integration examples, for different representa-
tive applications. Most importantly, we provide a ra-
tionale for multiple simultaneous vertical market re-
quirements, and account for a strategy for the chip
hosting the iSE to get multiple certifications accord-
ing to various different schemes. The key concept of
the iSE “intellectual property” block pre-certification
is introduced and we demonstrate its advantages over
the state-of-the-art.
Outline. The rest of this article is structured as fol-
lows. The current trends in user’s requirements are
described and analyzed in Sec. 2. This section moti-
vates for the relevance of iSE paradigm. The instanti-
ation of iSE in different contexts is the topic of Sec. 3.
This section shows that iSE architecture can be pro-
tean within the host chip, depending whether the host
can be trusted or not. Finally, our conclusions and
perspectives are in Sec. 4.
2 CURRENT NEEDS
2.1 Embedded versus Integrated Secure
Elements
Comparison. The figure 1 illustrates the layout dif-
ference between embedded and integrated Secure El-
ements. The left side of this figure shows the secure
chip and its companion Secure Element, connected by
a SPI (Serial Peripheral Interface) link. In the right
side, one can see the iSE, as a security subsystem
buried into the applicative chip (termed the “host”).
The symbioticity of the iSE with its host provides
more natural and better performing security applica-
tion.
Since Secure Elements are off-the-shelf compo-
nents, an attacker can easily analyze a discrete Secure
Element in a “training mode” on some samples before
launching his sharpened attack on the targeted prod-
uct. At the opposite, the iSEs are more protected since
the attacker shall first do some reverse-engineering to
identify the iSE resources to target.
Embedded Secure Elements. are stereotyped cir-
cuits, such as Trusted Platform Modules (TPM).
Many such chips can be produced in large series, lead-
ing to economies of scale. But this gain comes with
the fact that the electronic board is more complex,
since there are two chips to integrate and solder. In
addition, embedded Secure Elements have finite re-
sources, and can be somehow restricted in their us-
age, e.g., lack memory. Moreover, although they can
be certified, they typically are so only for one appli-
cation.
integrated Secure Elements (iSEs). are cus-
tomized for each host chip. This does not preclude
some level of reuse, on the contrary, but the hardware
of the iSE shall be recompiled. One advantage is that
the iSE can be taylored to meet exactly the require-
ments of each application, in terms both of security
level and performances. Regarding board-level inte-
gration, the iSE solution is cheaper since it does not
add a dedicated chip (it is integrated into a System-
on-Chip).
In addition, the iSE has at its disposal large
amount of resources, as most of those can be shared
with the host (such as various kinds of memories
1
,
connectivity, debug capabilities, etc.). Moreover, the
1
For instance non-volatile FLASH, One-Time Pro-
grammable (OTP) or electric fuses, Random Access Mem-
ories (RAM), etc.
Implementing Secure Applications Thanks to an Integrated Secure Element
567
board (PCB)
printed circuit
SE
system bus
secure chip
iSE
host
secure chip
embed d ed Secure Element:
IO
(SPI)
integrated Secure Element:
Figure 1: Comparison between embedded versus integrated Secure Element.
iSE is more efficient since it can naturally interface
with various applications (e.g., it is on the same die
as the host applicative code it is responsible to check
upon “secure-boot”, the crypto services offered by the
iSE can access the data to encrypt/sign/authenticate
within one clock cycle, etc.). Reciprocally, the iSE
can also help with the verification of further exter-
nal resources, such as the configuration’s authentic-
ity verification of host clock/power module, memory
management unit (MMU), debug JTAG port, etc. In a
nutshell, the iSE is extraverted, in that it can directly
protect its host and other resources. This “mutual
win-win exchange” is illustrated in Fig. 2.
Regarding application finetuning, the iSE can
have its code debugged directly within the host, which
is an advantage for the ease of integrated chip pro-
totyping (including host along with iSE, simultane-
ously).
2.2 Modern Requirements
Beyond the form factor considerations, we now un-
derline that there are more fundamental driving fac-
tors for iSE to be considered as the best option. One
major compelling reason is that current chips are so
complex that they are not specialized for only one ap-
plication, but instead provide multiple services. For
instance, one single chip can enable payment, access
to digital contents, (private) social networks, etc. The
security management of heterogeneous applications
(so-called “apps”), each requiring a different security
scheme, can rapidly become a nightmare with discrete
Secure Elements, where it can be elegantly fulfilled
with the iSE paradigm.
Even more, iSE is future-proof, in that new apps
(with specific security requirements) can be installed
after deployment. This stays true even if applica-
tions are not known nor foreseen upon chip design.
This on-field flexibility is appreciable in the context
of highly competitive market segments where invest-
ments on the chip are carried out ahead of full knowl-
edge of market potentialities.
3 USE-CASES WITH iSE
This section presents some typical integration pat-
terns of a Secure Element in a security chip.
3.1 Programmer’s Model
The cornerstone of security management in a Secure
Element is the concept initially forged by RSA Labs
company in the standard PKCS #11 (OASIS, 2020).
It consists in segregating security objects from actions
which can be carried out on them. Namely, in PKCS
#11, keys are kept within the Secure Element, and the
user can only request actions on them through an Ap-
plication Programming Interface (API). When imple-
mented correctly (Delaune et al., 2008), it is proven
that whatever the actions from the user side, keys can-
not cross the border between the SE and the host. Cor-
rect implementation requires nonetheless some atten-
tion, such as carefully attributing roles to keys, other-
wise API-level attacks can apply (Clulow, 2003).
Now in practice, there can be several ways to in-
tegrate the iSE within the host. Those variants de-
pend on the perceived likelihood of host corruptibil-
ity. Namely, an illustration is provided in Fig. 3. The
security application is represented as
/
(for the
code) and (for the secret data). Three typi-
cal application partitioning schemes are presented in
this figure.
(a) When it is assumed that the host is likely to be
compromised (but not the iSE), then actions are
carried out on the Application Processor (AP),
and keys are kept concealed within the iSE;
(b) When the protocol cannot easily be split between
actions and objects, and that the host is assumed
untrusted, then the whole secure applications re-
sides within the iSE: actions are executed by the
iSE processor;
(c) When the host is assumed to be incorruptible,
then the iSE boundary can be dissolved: the keys
ICISSP 2021 - 7th International Conference on Information Systems Security and Privacy
568
iSE
Reset (released upon secure boot)
JTAG (for verification & debug)
controller
MMU
controller
Clock/power
secure (host) chip
iSE
FLASH
(for keys
OTP/eFuse
Commu ni cation ports (secure updates)
management)
storage)
(for life cycle
RAM (for complex crypto. or OS)
Host resources protected by iSE:
Host resources used by iSE:
Figure 2: (left) Resources the iSE requires from the host to extend its functional capabilities; (right) Host resources the iSE
can verify.
are managed by the iSE and the actions (even
complex, i.e., unproven in terms of harmlessness
should the code be manipulated, e.g. by a bug ex-
ploitation or a malware) are performed by the host
AP.
Notice that technologies such as Global Platform
Trusted Execution Environment (TEE (Global Plat-
form, 2020b)) can be seen as a virtualization of iSE.
Indeed, TEE consists in an API which allows for a
given host to behave two-fold:
in a regular manner, termed “Rich Execution En-
vironment”, where untrusted apps can be run, and
in a secure manner, termed “Trusted Execution
Environment”, with logical and/or physical seg-
regation of resources, including memory, timers,
random number generators, etc.
The security of TEE technology is mature, as there is
even a Protection Profile which formalizes it (Glob-
alPlatform Device Committee, 2016). Now, physical
security is explicitly out of the scope of TEE, hence a
TEE is not sufficient per se to play the role of a Se-
cure Element. However, many practical attacks have
been reported (Cerdeira et al., 2020; Novella, 2020).
Side-channel attacks, and in particular cache-timing
attacks, represent a large number of them.
3.2 Some Examples of iSE Usage
Some standards actually impose an iSE, such as
Global Platform (GP) Virtual Primary Platform
(VPP (Global Platform, 2018)) specification, as
used for instance in European telecommunications
standards institute (ETSI) Smart Secure Platform
(SSP (ETSI, 2019)) standard. Also, in the following
use-cases, we highlight in particular that different us-
ages require different certification schemes.
Application: Platform Security
The host runs the application, and asks the iSE for key
management and cryptographic services. Examples
of applications are PKCS #11 (OASIS, 2020) (as in
Hardware Security Modules) and TCG TPM (Trusted
Computing Group, 2019). This use-case is adapted to
an architecture corresponding to case (a) in Fig. 3.
Application: iSIM (Integrated SIM, i.e.,
Integrated Subscriber Identity Module)
The iSE is the enclave. It ensures the secure execution
of the application, such as GP VPP (Global Platform,
2018) or ETSI SSP (ETSI, 2019). The case (b) of
Fig. 3 is suitable for this use-case. Typical security re-
quirements are enunciated in Common Criteria (Con-
sortium, 2013) Protection Profile PP0084 (Bunde-
samt für Sicherheit in der Informationstechnik, 2014).
Since IoT devices cannot manage discrete SIM cards,
the market is shifting to this iSE paradigm.
Application: Secure Communication
In this application, the iSE is offloading the host for
cryptographic computations; therefore, the iSE can be
dissolved amidst the host, as in case (c) of Fig. 3.
Such architecture is suitable for protocols such as
EVITA (FP7 European Project, 2020), IEC 62443 (In-
ternational Society of Automation (ISA, https://www.
isa.org/), 2020), or Car2Car V2X PP (CAR 2 CAR
Communication Consortium, 2019). Specifically, this
latter reference clearly shows the paradigm shift from
embedded (Figure 1) to integrated (Figure 2) Secure
Element, called “HSM” for “Hardware Security Mod-
ule” in automotive market parlance.
Implementing Secure Applications Thanks to an Integrated Secure Element
569
/ / /
system bus
iSEAP
secure chip
system bus
iSEAP
secure chip
system bus
iSEAP
secure chip
Case (b): Case (c):Case (a):
Figure 3: Different instantiations of an iSE within an host chip. Cases (a) and (b) are suitable when the host can be subverted
(e.g., it can be “rooted”); Case (a) works for simple applications where the specification is designed to gracefully split control
and data, whereas (b) is used for complex applications, e.g., which require an Operating System (OS). Case (c) is the natural
solution for those chips which are shielded from the outside world, hence protected by design from remote infection.
Application: Root-of-Trust (e.g., Open Compute
Project (OCP Security, 2020), Open Titan (Open
Titan, 2020))
Here, the iSE is responsible for booting an external
host, which is typically a server in a Cloud infrastruc-
ture. As the root-of-trust is standalone in its chip, ar-
chitectural case (c) of Fig. 3 is also sufficient. There is
no security standard in place for now for this segment.
3.3 How to Ensure that the Device is
Pre-certifiable?
As one can see, there are multiple certification
schemes. It might happen that one of the certifica-
tion is required late in the product life cycle, e.g., after
that the product is manufactured. We define the term
“pre-certification” as the action to anticipate a future
certification. This concept is disruptive as, tradition-
ally, certification was a necessary milestone to enter a
market. However, nowadays, the context is different
for several reasons:
Some chips are so versatile that they can be de-
ployed in various market verticals, e.g., IoT, in-
dustry, telecommunication stations, etc. which are
each governed by different security standards.
Obtaining consensus on security regulation can
take a long time. For instance, Global Platform
SESIP (Global Platform, 2020a) certification has
been proposed by the industry to remedy for the
lack of relevant international normative require-
ments in the field of IoT.
Users sometimes adopt technologies before the
security model is completely understood; this is
for instance the case with IoT, massively deployed
as smart home assistants, smart assistants, wear-
able health trackers, etc. (see (ETSI, TC CYBER,
2019, §1)).
Such proactive attitude requires some precautions to
be taken, amongst which:
Preparation of a clear documentation, including
user manuals and test plans;
Sound design decisions, e.g., implementation of
roles, of error management, etc.
Functional features like keys zeroization (cf.
NIST FIPS 140 (NIST FIPS, 2009)) shall be sup-
ported, etc.
In addition, pre-silicon actions are necessary. For
instance, using a Physically Unclonable Func-
tion (Bruneau et al., 2018) to implement the mas-
ter key is a bullet-proof security-by-design presilicon
choice. Also, pre-silicon security evaluation (Souissi
et al., 2019) allows to attest of resistance against at-
tacks even before product production stage. Some
Common Criteria schemes (typically the NSCIB in
the Netherlands) have already started to evaluate vir-
tual products, i.e., their source code, before it is im-
plemented.
Clearly, adjustments in the design can be fine-
tuned at software level, and documentation might
need some elaboration, but those are minor changes
if the underlying precautions have been respected.
4 CONCLUSIONS AND
PERSPECTIVES
The integrated Secure Element (iSE) is emerging as
a novel paradigm for ensuring the end-point secu-
rity. The iSE brings flexibility in the security manage-
ment, in that iSE can support various security applica-
tions. Moreover, by being instantiated directly within
the host chip, it can protect several resources (e.g.,
clock, reset, MMU, debug, etc.), and can be recon-
figured, hence it is future-proof. This highlights the
ICISSP 2021 - 7th International Conference on Information Systems Security and Privacy
570
need for pre-certified iSE, as an asset for the chip to
comply with different certifications, even before be-
ing launched on different markets.
Leveraging the iSE concept, further security appli-
cations can be envisioned. For instance, the iSE can
behave actively as a “cybersecurity probe” within the
device, responsible for reporting security events. This
opens the door for end-point devices used as sentinels
for gathering security threats at system-level. With
the capability to garner intelligence this way, secu-
rity analytics can be carried out. This allows for the
security operator to understand very precisely the at-
tacks perpetrated on the chip. Such information en-
ables him to proactively anticipate them, thereby re-
versing the advantage in favor of the defense side.
ACKNOWLEDGEMENTS
This work has been partly funded by the ARCHI-SEC
project (number ANR-19-CE39-0008).
REFERENCES
Bruneau, N., Danger, J., Facon, A., Guilley, S., Hamaguchi,
S., Hori, Y., Kang, Y., and Schaub, A. (2018). De-
velopment of the Unified Security Requirements of
PUFs During the Standardization Process. In SecITC,
Bucharest, Romania, November 8-9, 2018, volume
11359, pages 314–330. Springer.
Bundesamt für Sicherheit in der Informationstechnik
(2014). SI-CC-PP-0084-2014: Security IC Platform
Protection Profile with Augmentation Packages. Ver-
sion 1.0. https://www.commoncriteriaportal.org/files/
ppfiles/pp0084a_pdf.pdf.
CAR 2 CAR Communication Consortium (2019). Protec-
tion Profile V2X Hardware Security Module. Version
1.4.0.
Cerdeira, D., Santos, N., Fonseca, P., and Pinto, S. (2020).
SoK: Understanding the Prevailing Security Vulnera-
bilities in TrustZone-assisted TEE Systems. In 2020
IEEE Symposium on Security and Privacy, SP 2020,
San Francisco, CA, USA, May 18-21, 2020, pages
1416–1432. IEEE.
Clulow, J. (2003). On the Security of PKCS#11. In Walter,
C. D., Çetin Kaya Koç, and Paar, C., editors, CHES,
volume 2779 of Lecture Notes in Computer Science,
pages 411–425. Springer.
Consortium, C. C. (2013). Common Criteria (aka
CC) for Information Technology Security Evalu-
ation (ISO/IEC 15408). Website: http://www.
commoncriteriaportal.org/.
Delaune, S., Kremer, S., and Steel, G. (2008). For-
mal Analysis of PKCS#11. In Proceedings of the
21st IEEE Computer Security Foundations Sympo-
sium, CSF 2008, Pittsburgh, Pennsylvania, USA, 23-
25 June 2008, pages 331–344. IEEE Computer Soci-
ety.
ETSI (2019). Smart Secure Platform (SSP) ETSI TS
103 666-1 V15.0.0. https://www.etsi.org/deliver/
etsi_ts/103600_103699/10366601/15.00.00_60/ts_
10366601v150000p.pdf.
ETSI, TC CYBER (2019). EN 303 645 V2.0.0, Cyber Se-
curity for Consumer Internet of Things.
FP7 European Project (2020). E-safety Vehicle Intru-
sion protecTed Applications (EVITA). https://www.
evita-project.org/, accessed on Oct 27, 2020.
Global Platform (2018). Virtual Primary Platform (VPP)
v1.0.1. https://tinyurl.com/yybvbge4.
Global Platform (2020a). Security Evaluation Standard for
IoT Platforms (SESIP). Version 1.0. Document Refer-
ence: GP_FST_070.
Global Platform (2020b). Trusted Execution
Environment (TEE) Committee. https:
//globalplatform.org/technical-committees/
trusted-execution-environment-tee-committee/.
GlobalPlatform Device Committee (2016). TEE Protection
Profile Version 1.2.1, GPD_SPE_021. https://www.
globalplatform.org/specificationform.asp?fid=7831.
International Society of Automation (ISA, https://www.isa.
org/) (2020). ISA-62443-1-5. Security for industrial
automation and control systems Industrial automation
and control system protection levels.
ISO/IEC 7816 (2014). (Joint technical committee (JTC) 1 /
Sub-Committee (SC) 17), Identification cards Inte-
grated circuit cards. (details).
NIST FIPS (2009). (Federal Information Processing Stan-
dards) publication 140-3, Security Requirements for
Cryptographic Modules (Draft, Revised). page 63.
http://csrc.nist.gov/groups/ST/FIPS140_3/.
Novella, E. (2020). TEE-reversing: A curated list of
public TEE resources for learning how to reverse-
engineer and achieve trusted code execution on ARM
devices. https://github.com/enovella/TEE-reversing
(accessed December 7, 2020).
OASIS (2020). PKCS #11 TC. https://www.oasis-open.org/
committees/tc_home.php?wg_abbrev=pkcs11.
OCP Security (2020). Specification of Root of Trust,
version 1.0. https://www.opencompute.org/projects/
security.
Open Titan (2020). Open source silicon Root of Trust (RoT)
project. https://opentitan.org/, accessed on Oct 28,
2020.
Souissi, Y., Facon, A., and Guilley, S. (2019). Virtual
Security Evaluation - An Operational Methodology
for Side-Channel Leakage Detection at Source-Code
Level. In Codes, Cryptology and Information Security
- Third International Conference, C2SI, Rabat, Mo-
rocco, April 22-24, 2019, pages 3–12.
Trusted Computing Group (2019). Trusted Platform Mod-
ule (TPM), Library Specification, Family “2.0”, Level
00, Revision 01.59. https://trustedcomputinggroup.
org/work-groups/trusted-platform-module/.
Implementing Secure Applications Thanks to an Integrated Secure Element
571