Embedded Systems Security Challenges
Konstantinos Fysarakis
1
, George Hatzivasilis
1
, Konstantinos Rantos
2
,
Alexandros Papanikolaou
1
and Charalampos Manifavas
3
1
Dept. of Electronic & Computer Engineering, Technical University of Crete, Chania, Greece
2
Dept. of Computer & Informatics Engineering, Eastern Macedonia and Thrace Institute of Technology, Kavala, Greece
3
Dept. of Informatics Engineering, Technological Educational Institute of Crete, Heraklion, Crete, Greece
Keywords:
Embedded Devices, Security, Denial of Service, Lightweight Cryptography, Location Privacy, Secure Routing.
Abstract:
In a world of pervasive computing, embedded systems can be found in a wide range of products and are em-
ployed in various heterogeneous domains. The abovementioned devices often need to access, store, manipulate
and/or communicate sensitive or even critical information, making the security of their resources and services
an important concern in their design process. These issues are further exacerbated by the resource-constrained
nature of the devices, in conjunction with the ever-present need for smaller size and lower production costs.
This paper aims to provide an overview of the challenges in designing secure embedded systems, covering
both node hardware and software issues, as well as relevant network protocols and cryptographic algorithms.
Moreover, recent advances in the field are identified, highlighting opportunities for future research.
1 INTRODUCTION
Embedded systems (ESs) permeate our lives in vari-
ous forms, ranging from avionics to e-textiles, auto-
mobiles, home automation and wireless sensor nodes.
In terms of their physical size, they range from minia-
ture wearable or sensor nodes (i.e. motes) to large
industrial deployments of programmable logic con-
trollers (PLCs).
The various intrinsic and application-specific
characteristics of ESs complicate the task of guaran-
teeing the security, namely handling the confidential-
ity, integrity and availability aspects of their applica-
tions and the data they handle. Their characteristics
habitually include resource constraints (namely com-
putational capabilities, storage capacity, memory and
power), dynamically formulated, remotely-managed
networking and even unattended operation in hostile
environment and time-critical applications. There-
fore, while securing networked computer systems is
not a novel concern, the techniques developed for per-
sonal and enterprise systems are often unsatisfactory
or even inapplicable to embedded devices.
In addition to the above, ES applications often fea-
ture direct interaction with the physical world, be-
ing responsible for vital, time-critical applications,
where a delay or a speed-up of even a fraction of a
second in system’s response or reaction could have
dire consequences. This further differentiates ES se-
curity, as a security incident in a critical application
may lead to asset damage or even personal injury
and death. Researchers recently demonstrated that
it is feasible to manipulate all critical sub-systems
in modern automobiles by using a wireless-enabled
MP3 player connected to the vehicle’s embedded con-
trol network (Koscher et al., 2010). The presented
attacks include accessing the brake controller, thus
disabling or forcibly activating the brakes and con-
sequently compromising the safety of the driver and
passengers, as well as injecting malicious code to
erase any evidence of tampering after a crash.
Moreover, next-generation ES services, like the
ones pertaining to the Internet of Things (IoT), may
require the integration of multiple administrative do-
mains (e.g. one domain may host the devices and en-
able access to devices and information,whereas an-
other domain may make use of the information for
designing innovative services). Each domain will typ-
ically have its own security requirements and con-
straints, therefore ensuring interoperability of security
is a challenging task (Alam et al., 2011).
The paper is organised as follows: Section 2
presents some physical security issues that are evi-
dent in embedded systems. The various access control
mechanisms used for controlling access to resources
are presented in Sec. 3. Indicative examples of crypto-
255
Fysarakis K., Hatzivasilis G., Rantos K., Papanikolaou A. and Manifavas C..
Embedded Systems Security Challenges.
DOI: 10.5220/0004901602550266
In Proceedings of the 4th International Conference on Pervasive and Embedded Computing and Communication Systems (MeSeCCS-2014), pages
255-266
ISBN: 978-989-758-000-0
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
graphic mechanisms specially-crafted for embedded
systems are presented in Sec. 4. Various protocol and
management issues are presented in Sec. 5, and the
paper concludes in Sec. 6.
2 PHYSICAL SECURITY ISSUES
Regarding the physical layer and given the often unat-
tended nature of deployed ESs, sometimes within
hostile environments, the risk of device tampering
should not be ignored. In the remainder of this sec-
tion, some aspects of physical security in ESs are pre-
sented.
2.1 Side Channel Attacks
A malicious entity’s physical access to a non-tamper-
resistant device, apart from providing physical ac-
cess to the system components, would also enable
the launch of various attacks like micro-probing and
reverse engineering or sophisticated side-channel at-
tacks (SCA), timing attacks, simple power analy-
sis (SPA), differential power analysis (DPA), as well
as their electro-magnetic counterparts, SEMA and
DEMA respectively, and also differential fault attacks
(DFA). The aforementioned methods can potentially
expose critical information concerning the operation
of the device (algorithms used, length of keys etc.)
which could prove critical to the security both of the
device itself and the network as a whole.
In the case of ESs utilising field-programmable
gate arrays (FPGAs), the concept of run-time re-
configuration (Dan
ˇ
ek et al., 2008) can be explored
to reduce component count and/or power consump-
tion, increase fault tolerance etc. as needed. Self-
reconfigurability can, for example, make a node more
secure against side-channel attacks through the mea-
surement of electromagnetic radiation and also im-
plement self-healing properties. Self-recovery mech-
anisms could reallocate functional blocks to mark
and replace faulty resources, through device repro-
gramming in the case of self-reconfigurable nodes or
through controlled degradation of service techniques
in less “intelligent” devices.
2.2 Trusted Platform Module
A Trusted Platform Module TPM (Trusted Plat-
form Module, 2009), is a microcontroller that can se-
curely store a relatively small amount of information
on it, which can then be used either for authenticat-
ing the platform (e.g. passwords, certificates, encryp-
tion keys) or for ensuring that the platform has not
been breached (e.g. platform measurements, configu-
ration data). It is worth clarifying that the microcon-
troller does not control the software that is running
on the platform itself. Instead, through the tamper-
resistant security functions it provides, the platform’s
operating system or any running applications access
the necessary information, so as to determine and im-
plement their security policies accordingly. The soft-
ware embedded on such a TPM component is directly
related to the component’s physical size (the higher
the memory requirements, the larger the module’s sur-
face) and consequently cost. Any optimisation of the
module’s software will have a direct impact on the
overall execution speed, as well as to the power con-
sumption (faster execution allows the module to re-
turn to a more power-sparring idle/sleep state).
2.3 Protection of Power Supply
Several types of embedded systems devices are
mostly battery-powered, something that creates issues
of energy constraints. Especially in cases where e.g.
small sensor nodes are meant to be used in unattended
environments, they are expected to operate for cer-
tain time intervals (sometimes spanning over a few
months) until they have their power source replaced
or recharged. An attacker could therefore launch a
Denial-of-Service (DoS) attack, aiming at draining
the battery power by forcing extensive use of the de-
vice’s wireless connection or CPU.
In order for the electronic parts of the embed-
ded device to function properly, continuous power is
required, with both voltage and current levels lying
within specific limits. The power source should also
be able to monitor its own state and react accordingly
in cases where an issue is detected that could affect
the normal operation of the system. Moreover, suit-
able fail-safe mechanisms should exist, implemented
in software and/or hardware, that would protect the
device and prevent any potential damage from spread-
ing across the rest of its components.
Most of the requirements mentioned above are sat-
isfied in modern uninterruptible power supply (UPS)
systems, nevertheless, they cannot easily be applied
to the operating environments of some embedded sys-
tems, due to strict size and cost constraints. Instead,
alternative solutions are implemented, such as energy
scavenging, super-capacitors, micro-solar cells and
remote/wireless power transferring schemes (Kurs
et al., 2007; Karalis et al., 2008), to name a few. Still,
such solutions must carefully be adopted to each spe-
cific scenario (e.g. a micro-solar cell is useless if the
device cannot be reached by enough direct sunlight),
also taking into consideration fail-safe options with
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
256
respect to the criticality of the possible failures and
the probability for them to occur, so as to protect the
device effectively.
3 ACCESS CONTROL
Access control mechanisms are essential to pre-
vent unauthorised/malicious entities to access the re-
sources, physical or otherwise, available to the ESs as
well as the hosting devices. The way access control is
implemented varies depending on the hardware capa-
bilities of the nodes, the type of network and the appli-
cation under consideration. Some common methods
include:
1. Profile authentication: If a node has some specific
characteristics (e.g. hardware specifications, oper-
ating system), it can join an existing network.
2. Access code: Demonstrating knowledge of the
code grants access to the network and its re-
sources. This code can either be programmable
or configurable. This category includes typical
password access, based on memory data, switch
configuration or any other procedure.
3. Predefined topology: Only pre-established nodes
can join the network (e.g. MAC filtering).
There is ongoing research on ES-specific access
control protocols since the commonly-used authen-
tication schemes, typically password-based, can be
impractical or even insecure when considering the
heterogeneous nature of ES networks can demon-
strate and the scalable remote manageability often re-
quired (Naedele, 2006). Moreover, even in wired em-
bedded networks and in industries such as automo-
tive and aviation, most control networks utilised (e.g.
Controller Area Network, Time-Triggered Protocol,
FlexRay) are designed with safety and reliability in
mind and do not feature any built-in security mech-
anisms like node authentication, data encryption or
prevention of DoS attacks (Szilagyi and Koopman,
2009; Szilagyi and Koopman, 2010), leading to crit-
ical vulnerabilities like the ones already mentioned
in (Koscher et al., 2010).
3.1 Policy-Based Access Control
ESs are often deployed in applications bound by strict
security requirements (e.g. e-Health applications are
a prime example), including secure transmission of
sensitive data to remote entities, instructions that need
to reach actuators in an unaltered form, robust entity
authentication and access control mechanisms (Al-
haqbani and Fidge, 2008).
Regarding the latter, among the proposed schemes
that have gained popularity are those where deci-
sions are made based on policy restrictions. Such a
scheme is the standardised by OASIS eXtensible Ac-
cess Control Markup Language (XACML), an XML-
based general-purpose policy decision language. Be-
sides being used for representing authorisation and
entitlement policies for managing access to resources,
it provides a processing model for evaluating requests
and making decisions based on the defined set of poli-
cies (Rantos et al., 2012). Policy-based access control
allows dynamic decision making on controlled nodes’
resources based on policy restrictions set by the sys-
tem owner.
An XACML architecture commonly consists of
the following components:
Policy Enforcement Point (PEP): The system en-
tity that performs access control, by making de-
cision requests and enforcing authorisation deci-
sions.
Policy Administration Point (PAP): The system
entity that creates a policy or policy set.
Policy Decision Point (PDP): The system entity
that evaluates applicable policy and renders an au-
thorisation decision.
Policy Information Point (PIP): The system entity
that acts as a source of attribute values.
Context Handler: It orchestrates the communica-
tions among the stakeholders, converts, if neces-
sary messages between their native forms and the
XACML canonical form, and collects all neces-
sary information for the PDP.
The aforementioned components are complemented
by a well-defined set of policies which define the rules
that should be taken into account when examining ac-
cess requests.
In a typical data flow model, authorisation re-
quests for accessing nodes’ resources, i.e. PEPs are
forwarded to PDPs together with the required at-
tributes taken from the PIP. The PDP evaluates the
request against the policy restrictions taken from PAP
and issues an authorisation decision which the PEP
has to enforce.
These components do not necessarily run on the
same node, mainly because resource-restricted nodes
do not have the capacity to accommodate them. In
this case more distributed approaches have to be
adopted to offload computationally expensive opera-
tions run by a PDP to more powerful devices.
Still, unprotected policy messages would expose
the system’s security, revealing private information to
attackers who might also try to identify policy restric-
tions and do a mapping of the security measures taken
EmbeddedSystemsSecurityChallenges
257
for the specific environments, hence exploiting poten-
tial vulnerabilities. Moreover, in a more active ap-
proach, an attacker might masquerade as a legitimate
entity or modify policy-related messages, such as au-
thorisation requests and/or decisions, obligations or
advices in an attempt to downgrade adopted measures
and bypass access controls. To avoid the aforemen-
tioned problems, appropriate security measures, like
the ones detailed in later sections of this paper, have
to be deployed to safeguard message confidentiality,
integrity and authentication.
3.2 Denial of Service (DoS)
The aim of a DoS or a Distributed DoS (DDoS) at-
tack is to harm the availability of a specific node or a
network of nodes, thus preventing or delaying legiti-
mate entities from accessing the services or resources
they wish to (Carl et al., 2006). DoS attacks on ES
nodes can take multiple forms, such as exploiting the
vulnerabilities in their software/firmware, or attack-
ing the network they belong to by jamming, misrout-
ing, flooding and so on. DoS attacks can be mounted
more or less on every layer, by exploiting the par-
ticular characteristics or mechanisms found in them.
Their aim is to cause the nodes to constantly process
or send dummy data, thus draining their power source
via unnecessary use of their wireless connection and
prolonged demand for memory and CPU cycles. The
effects of such attacks are particularly critical in the
case of nano and micro/personal nodes, where the
power reserve is usually rather limited. In addition,
flooding types of DoS attacks consume part of the net-
work’s bandwidth, which may also indirectly affect
the normal operation of the network or a significant
part of it, depending on the overall network’s capac-
ity.
Another type of DoS attack involves the case
where an attacker gains physical access to the device
and modifies or destroys it as a physical entity. De-
pending on the role of this particular node for the rest
of the network or cluster of nodes it belongs to, its un-
availability could have a significant impact on these
entities. For instance, it could lead to partial or total
loss of the data sink, selection of non-optimal routes,
or even the loss of a control node that is vital for the
normal operation of the system.
Given that large-scale networks are most proba-
bly heterogeneous in their nature, they contain differ-
ent capabilities and vulnerabilities, which need to be
addressed independently for achieving effective pro-
tection against DoS attacks. What is more, in cases
where there is provision for dynamic network size
variation, shielding against DoS attacks can become a
very challenging task (Raymond and Midkiff, 2008).
In addition, the problem becomes even more complex
and difficult to solve for cases where the available re-
sources and capabilities exhibit strict limitations.
In cases where the ES network is of an unat-
tended nature, the use of a remote management sys-
tem is vital. Nevertheless, such systems offer ad-
ditional attack surface and their compromise can al-
low an adversary to upload malicious firmware. In
this way, the attacker is able to corrupt memory and
data sent/received or even lead to permanent damage
of hardware sub-components by intended misconfig-
uration (Permanent DoS PDoS, or bricking). The
success of such attacks is based on the fact that the
various firmware upgrade mechanisms are usually in-
secure and do not employ complex security mecha-
nisms for verifying authenticity and integrity, such as
the ones used in digital certificates. This process of
rendering a device unbootable or non-reflashable is
also known as phlashing (Smith, 2008).
One of the main reasons for the DoS attacks be-
ing relatively easy to successfully launch is the use of
old protocols that suffer from lack of security require-
ments. For instance, the IP protocol takes for granted
various assumptions regarding the trust of network
nodes and consequently does not dispute the related
information found in the packet headers and/or pay-
load. Any integrity-checking mechanisms are rather
primitive and simple in nature (e.g. checksums), as
their aim is to detect accidental data corruption and
not deliberate modification of information. Therefore,
continuing to use such protocols as the base for build-
ing custom network communication protocols on it
makes it particularly hard to design (D)DoS-resilient
systems and services. An additional obstacle is the
fact that basic software methodologies do not take
into consideration security requirements, able to deal
with such kinds of attacks (Stefanidis and Serpanos,
2008).
The majority of the aforementioned attacks can
be avoided by employing suitable authentication, ac-
cess control and integrity-checking mechanisms. It is
also equally important to provide secure mechanisms
for node firmware deployment and software updates,
able to verify both the authenticity and integrity of
the firmware/software to be uploaded and reject it if it
does not pass the required checks. Furthermore, more
recent network protocols should be used, able to pro-
vide means and metrics for quantifying (D)DoS attack
resilience (Aad et al., 2008). The use of intrinsically
secure ES firmware offering various fail-safe mech-
anisms or even hardware redundancy could be em-
ployed in cases where dependability is highly critical
(e.g. avionics and the military), as they are expected
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
258
to increase the cost of the end-product.
4 CRYPTOGRAPHIC MECHANISMS
As has already been mentioned, embedded devices
often have inherent limitations in terms of process-
ing power, memory, storage and energy. Efficient
algorithm designs and implementations that adhere
to these constraints, while satisfying application de-
mands, can significantly impact battery lifetime and
allow the implementation of many applications.
Key management is an equally important issue,
both from a security and a management point of view.
The rather simple pre-shared key (PSK) scheme,
where every embedded device has the necessary cryp-
tographic keys pre-installed, is difficult to manage in
distributed and dynamic environments (physical ac-
cess to the device is required) or in cases where there
is a large number of such devices. Moreover, the dis-
closure of the master key leads to the instant compro-
mise of all the system/network. Having an appropri-
ate scheme that triggers periodic re-keying limits the
amount of ciphertext that has been encrypted with the
same key, thus increasing the system’s security level.
This section presents several lightweight crypto-
graphic and key management schemes, suitable for
resource-constrained devices.
4.1 Lightweight Cryptography
Lightweight Cryptography (LWC) refers to algorith-
mic designs and implementations best suited for de-
ployment in such devices (e.g. RFIDs, sensor nodes,
contactless smartcards, mobile devices). There has al-
ready been significant effort on the subject of crypto
optimisation, aiming to maintain the security level
that “traditional” algorithms and implementations of-
fer while narrowing what is often referred to as “bat-
tery gap” (Doomun and Soyjaudah, 2009), i.e. the
very high energy consumption overheads for support-
ing security on battery-constrained systems. A num-
ber of surveys (Preneel, 2009; Eisenbarth et al., 2007)
provide an overview of this subject.
ISO/IEC 29192 includes ciphers and crypto-
graphic mechanisms for LWC. The standards are
PRESENT (Bogdanov et al., 2007) and CLEFIA (Ak-
ishita and Hiwatari, 2012) for block ciphers, and
TRIVIUM (ECRYPT, 2008) and Enocoro (Watanabe
et al., 2008) for stream ciphers.
Compact implementations of “traditional” ci-
phers, like AES (Feldhofer et al., 2005), are also
applied to embedded devices. Newer lightweight
block ciphers are Humminbird-2 (Engels et al., 2011),
Piccolo (Shibutani et al., 2011), Simon (Beaulieu
et al., 2013), SPECK (Beaulieu et al., 2013),
ITUbee (Karakoc¸ et al., 2013). For stream ciphers,
the finalists of the eSTREAM project Grain, Salsa20,
Rabbit and HC128 (ECRYPT, 2008) are suitable for
LWC.
Hash functions design is another area where fur-
ther research is required. For LWC, compact im-
plementations of the standardised functions SHA-
2 (Bogdanov et al., 2008), SHA-3 (Gaj et al., 2012)
are examined. Newer functions that are suitable in
this domain are Blake (Gaj et al., 2012), Photon (Guo
et al., 2011), SPONGENT (Bogdanov et al., 2011)
and other hash functions based on lightweight block
ciphers, like DM-PRESENT (Bogdanov et al., 2008).
Asymmetric algorithms and protocols must also
be adapted to operate on devices with the afore-
mentioned resource limitations. This is an elaborate
task, since asymmetric ciphers are computationally
far more demanding than their symmetric counter-
parts and are usually executed on powerful hardware.
The performance gap is exacerbated on constrained
devices, such as 8-bit microcontrollers. Even an opti-
mised asymmetric algorithm (e.g. elliptic-curve cryp-
tography ECC), performs 100 to 1000 times more
slowly than a standard symmetric algorithm (e.g.
AES), which correlates to a proportionally higher
power consumption.
In terms of practical relevance, two families of es-
tablished public-key algorithms stand out: ECC and
NTRU (Kamal and Youssef, 2009). ECC in partic-
ular is considered the most attractive option in ESs,
due to its small operand length and its relatively low
processing requirements. NTRU is the most popular
lattice-based cryptosystem. Its security is based on
the shortest vector problem and it can efficiently be
deployed on embedded systems. Compared to ECC
in hardware, NTRU is 1.5 times faster with only the
1/7 of the memory footprint. Compared to RSA in
software, it is 200 times faster in key generation, al-
most 3 times faster in encryption and 30 times faster
in decryption.
The need for lightweight cryptography introduces
major multi-dimensional challenges in cryptographic
algorithms design, from the ES operating system (OS)
to the hardware and software cryptographic provi-
sions embedded on the device itself. Hardware and
software co-design seems to offer the best results in
terms of speed/size ratio for many ubiquitous comput-
ing applications (Eisenbarth et al., 2007). Regarding
primitives that cannot yet be effectively implemented
(e.g. hashes in the case of crypto and public key
crypto in the case of asymmetric), alternatives could
be investigated so that the protocols which are based
EmbeddedSystemsSecurityChallenges
259
upon them can be researched further and perhaps put
into practice. Special care should be taken during
the development of optimised implementations so that
they do not introduce new leakage channels which
could be exploited by Side-Channel Attacks (SCA).
A comparative analysis of lightweight ciphers on
embedded systems was performed recently, where the
authors evaluate the proposed schemes based on per-
formance metrics and classify them for various types
of embedded devices (Manifavas et al., 2013). Vari-
ous cryptographic libraries exist that offer key estab-
lishment mechanisms and communication protocols,
like for example ULCL (Hatzivasilis et al., 2014),
CyaSSL and OpenSSL.
4.2 Key Distribution Mechanisms
Key distribution, either for initialisation (Kuo et al.,
2007) or re-keying has been a challenging topic es-
pecially for dynamic, heterogeneous and resource-
limited environments. The majority of these schemes
is based on symmetric mechanisms, thus requiring
pre-distribution of the shared secret with all the dis-
advantages already discussed. Other schemes (Doyle
et al., 2006) are being proposed as well, some
of which feature location-aware and identity-based
mechanisms. Although some of the proposed
schemes are indeed energy efficient (Huang et al.,
2005a), key management based on shared secrets
has proven ineffective, especially in dynamically-
formulated infrastructures. There have been attempts
to correlate key establishment techniques to applica-
tions but these were based solely on the use of sym-
metric keys and on a framework level (Martin and Pa-
terson, 2008).
This has led part of the research community to fo-
cus its attention on public-key schemes (e.g. ECC).
This enables the distribution of authentic public keys
via insecure channels, as the verifying party does not
need to have a copy of the secret key. Therefore, a
mobile node’s key database may, for example, be up-
dated with all valid public keys once, according to a
pre-defined schedule or ad hoc, and from that point
onwards the device will be able to authenticate other
entities in off-line mode.
Identity-based cryptography IBC (Boneh and
Franklin, 2003) provides an alternative solution to
the very expensive, for many nodes, public-key cryp-
tography. IBC allows publicly-available information
that can uniquely identify participating nodes to be
utilised for the secure exchange of keys. Thus, nodes
do not have to depend on a public key infrastruc-
ture and digital certificates for the exchange of au-
thenticated public keys. The advent of identity-based
schemes and pairing-based cryptography has shown
that such schemes offer a promising solution for man-
aging keys in resource-constrained nodes (Oliveira
et al., 2011).
Even with a public-key scheme, we still need to
implement symmetric cryptography. Thus, alternative
schemes that are based only in symmetric cryptogra-
phy are also proposed (Chen and Chao, 2011; Simpli-
cio et al., 2010). Some of them are location-aware and
identity-based mechanisms and are energy-efficient.
Their disadvantage is the pre-distribution phase that
is required prior to first usage. The schemes are in-
efficient in dynamic environments but are proposed
in applications where public-key schemes cannot be
used.
5 NETWORK PROTOCOL AND
MANAGEMENT ISSUES
Certain applications of embedded systems, like Wire-
less Sensor Networks, rely on the integrity of the plat-
form for providing trustworthy services (e.g. mea-
surements taken by a sensor). It is therefore essen-
tial to have a method for validating this integrity and
assuring that system components have not been com-
promised. The integrity of the service-requester plat-
form, i.e. control node, must also be validated before
allowing it to allocate resources to the nodes it con-
trols or receive the data these nodes have collected.
In addition, it should be established that these secure
resource management mechanisms will not act as a
bottleneck in service performance. Examples of cur-
rent research on the subject are the WS-Attestation
mechanism (Yoshihama et al., 2007), which enables
Trusted Platform Module (TPM) remote platform at-
testation using web services.
5.1 Secure Resource Management
Inspecting the problem from a higher level, middle-
ware resources should be managed by monitoring
their availability, enforcing a policy based on which
of these resources are assigned, implementing a se-
cure model for the identification and authorisation of
requests, as well as an accounting system to track re-
source usage. Most of the above can be found in pro-
tocol Diameter (Calhoun et al., 2003), successor to
RADIUS, which offers strong authentication, autho-
risation, accounting and resource management mech-
anisms. Diameter is already adopted by many IP sys-
tems like in the 3rd Generation Partnership Project
(3GPP).
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
260
5.2 Reputation-based Schemes
Reputation-based schemes are a novel paradigm for
enhancing security in various applications, includ-
ing secure routing and intrusion detection systems for
Mobile Ad Hoc Networks MANETS (Hatzivasilis
and Manifavas, 2012). These systems are easy to im-
plement, lightweight and can protect a MANET from
a wide variety of attacks.
The basic concept is inspired from social be-
haviour and relies on the cooperation of the nodes.
Much like human interaction, each entity decides to
trust or ignore a new, unknown entity based on the
opinion of its peers about the individual in question.
Consequently, much like in social networks, trustwor-
thy behaviour is encouraged. The three main goals
identified (Resnick and Zeckhauser, 2002) for reputa-
tion systems are:
To provide the required information in order to
distinguish between a trustworthy principal and
an untrustworthy one.
To encourage principals to act in a trustworthy
manner.
To discourage untrustworthy principals from par-
ticipating in the service.
Reputation-based mechanisms are used in Intru-
sion Detection Systems (IDSs) and provide two main
functionalities: Secure routing and resource manage-
ment. Watchdog and Path-rater (Buchegger et al.,
2004) are the building blocks of such systems. Watch-
dog is a monitoring component and based on its
observations Path-rater ranks the available routing
paths. The main steps in the reasoning process of a
reputation-based scheme are as follows:
1. Gather information.
2. Score and rank.
3. Select entity.
4. Transaction.
5. Reward and punish.
Misbehaviour detection and intrusion detection can
either be distributed, where information about en-
tities’ reputation changes are immediately broad-
cast to the whole network (or local), in which case
each entity decides, based solely on its own data
about the reputation of other nodes. It should be
noted, however, that the latter is not as effective in
terms of speed in detecting and isolating of mali-
cious nodes. Known reputation-based systems for se-
cure routing are IRmIDS (Madhavi and Kim, 2011),
Reputated-ARAN (Abdalrazak and Sawant, 2012)
and CSRAN (Zhang et al., 2008).
While secure routing provides P2P protection,
reputation-based systems for secure resource man-
agement provide end-to-end protection. Such sys-
tems rank resources, providers and consumers. Based
on these values, network entities are able to recog-
nise legitimate providers and resources of good qual-
ity while keeping out selfish and malicious users.
Several reputation-based schemes have been pro-
posed for resource management in P2P and Grid net-
works (Damandeep and Jyotsna, 2012).
5.3 Anonymity and Location Privacy
Location-based applications constitute a rapidly ex-
panding market, owing to the widespread use and
advances in mobile devices and positioning systems
alike. Enhanced reality applications and other simi-
lar services are starting to emerge and are expected to
spread in the coming years. Other examples of such
smart services include location-aware emergency re-
sponse, enhanced entertainment and/or advertisement
services, or even location monitoring of personnel and
fleets of vehicles.
The location of individual users is necessary in or-
der to enable the abovementioned services but, even
though its disclosure may not pose a security risk for
the embedded device itself, said information consti-
tutes sensitive personal data of the user or users as-
sociated with each device and should be handled ac-
cordingly. Disclosure of such information can enable
a malicious user to harass, blackmail or even enter
the individual’s residence (e.g. when he/she is away).
There is on-going research on the subject, including
mechanisms for safeguarding location privacy (Gedik
and Liu, 2008; Zhong and Hengartner, 2008) as well
as reports on the weaknesses of current “sanitisation”
mechanisms (Golle and Partridge, 2009; Gruteser and
Grunwald, 2005).
Literature on this topic includes variations and en-
hancements of a few recurring methods. Anonymi-
sation methods aiming to remove identifying infor-
mation using generalisations and suppressions are the
most popular in the literature, with k-anonymity being
the basic mechanism used, as described in (Zhong and
Hengartner, 2008). The principle of k-anonymity in-
volves the use of a cloaking area, where there are at
least k users in it, and blurs their identities in order to
make each user’s identity indistinguishable from the
rest k 1 users. In general, there are two important
and mostly unavoidable trade-offs when choosing this
k-value: A trade-off between privacy and quality of
service and a trade-off between privacy and person-
alisation (Liu, 2007). K-anonymity-based schemes
have typically being deployed for privacy-preserving
EmbeddedSystemsSecurityChallenges
261
location monitoring and to allow mobile users to
take advantage of location-based personalised ser-
vices without compromising their privacy (Fysarakis
et al., 2013).
Elementary anonymity schemes (e.g. a pure k-
anonymity scheme) are inadequate if used indepen-
dently and hence they are often combined with aux-
iliary mechanisms, in order to achieve better privacy
safeguards (Golle and Partridge, 2009; Gruteser and
Hoh, 2005; Liu, 2009). Attempting to further en-
hance these basic schemes, the use of personalised k
values is proposed, for systems with context-sensitive
privacy requirements (Gedik and Liu, 2008). More-
over, if combined with l-diversity (Machanavajjhala
et al., 2007), another dimension can be added to k-
anonymity, where l is a set of distinct locations. Other
than the above suggested improvements, the use of
dummy locations (Gupta et al., 2011) and semantic
information (Byoungyoung et al., 2011) are also pro-
posed in the literature, in order to address issues that
are not solved by plain k-anonymity mechanisms.
Pseudonym-based methods are another com-
mon anonymisation theme, involving disposable
pseudonyms for each node in the location ser-
vice (Gruteser and Grunwald, 2005). As the
pseudonyms change over time, they are being used
as temporal identifiers, making it hard for potential
attackers to track the users. Silence periods can be in-
troduced to further enhance this concept, as detailed
in (Huang et al., 2005b). Other literature schemes rely
on path perturbation, i.e. trying to cross paths in ar-
eas where at least two users meet (Hoh and Gruteser,
2005).
5.4 Secure Service Discovery,
Composition and Delivery Protocols
Services in distributed networks must be discovered,
composed and delivered in a secure way. The Organi-
sation for the Advancement of Structured Information
Standards (OASIS) has released related standards in-
cluding WS-Security, WS-Policy, WS-Trust and WS-
Secure Conversation which have already been ap-
proved and the current trend is to bring web services
into ESs and it is thus imperative to adopt the afore-
mentioned specifications.
For this purpose, OASIS developed two standards:
Devices Profile for Web Services (DPWS) and Web
Services Dynamic Discovery (WS-Discovery), which
specify the use of web-services-based communica-
tions in resource-constrained and ad hoc environ-
ments. The profile’s architecture includes hosting and
hosted services. A single hosting service is associ-
ated with each device while the same device may ac-
commodate various hosted services. The latter rep-
resent the device’s various functional elements and
rely on the hosting service for discovery. Discov-
ery services are included as well, enabling devices to
“advertise” their presence on the network and search
for other devices. Metadata exchange services pro-
vide dynamic access to services hosted on a device
and their meta-data. Furthermore, publish/subscribe
eventing services allow other devices to subscribe to
messages provided by a certain service.
Additional research on this area has been con-
ducted by the Service-Oriented Architecture for De-
vices (SOA4D) open-source initiative which facili-
tates the development of service-oriented software
components adapted to the requirements of embed-
ded devices. Web Services for Devices (WS4D) is
another open source initiative, providing a number
of toolkits aimed at developing DPWS-compliant ap-
plications for resource-constrained devices in ad-hoc
networks, maintaining interoperability with regular
W3C-specified Web Services. A detailed overview
of the WS4D initiative can be found in (Zeeb et al.,
2010).
5.5 Communications Security
Embedded nodes have quite a few choices regard-
ing the protocols they adopt in their communica-
tion stack, depending on their computation capabili-
ties and needs. One of the predominant solutions is
the 6LoWPAN (Hui and Thubert, 2011) stack, i.e.
IPv6 over 802.15.4 (IEEE Standard for Local and
metropolitan area networks, 2011). Such an approach
benefits from the adoption of well-known and stan-
dardised solutions for providing security at the net-
work layer. This is IPsec which, however, has to
utilise compressed header format (Raza et al., 2011;
Rantos et al., 2013a; Rantos et al., 2013b) to fit into
the limited message space provided by IEEE802.15.4,
i.e. a for low-power, low data rate wireless communi-
cation standard for small devices.
The network layer, however, is not the only layer
in the TCP/IP stack where messages can be pro-
tected. The others are application and data link layer.
The corresponding security mechanisms for these two
layers are the Datagram Transport Layer Security
DTLS protocol (Kothmayr et al., 2012), which is
based on the well-known TLS (Transport Layer Se-
curity) but utilises datagram protocols and the in-
herent security mechanism of IEEE802.15.4 which
is defined in the same standard. All these mech-
anisms were designed to provide at least confiden-
tiality, integrity and message authentication, yet they
have slightly different properties.
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
262
Security mechanisms found at the bottom layers
of the communication stack relieve applications from
deploying their own distinct security mechanisms.
However, this comes at a cost. With regards to the
IEEE 802.15.4 security protocol which protects mes-
sages at the data link layer, protection takes place
on a node-by-node basis. This introduces signifi-
cant computational overhead to the nodes which have
to decrypt incoming messages, verify their integrity,
and re-encrypt them, typically using a different set of
keys, prior to forwarding them to the next node. This
process consumes valuable node resources on routing
nodes.
As opposed to 802.15.4, IPsec offers end to end
protection of messages. Therefore, intermediate rout-
ing nodes’ resources are only used for routing packets
and not for message protection. In this sense, security
provided at the network layer can be considered as
a valuable mechanism for low power and lossy net-
works.
IPsec can be either used within such a network
to secure communications among participating nodes
in cases where these are deployed in a hostile envi-
ronment to secure communications among participat-
ing nodes, or between a node and remote party. This
second choice requires utilising a gateway, e.g. sink
node, which can simply forward messages or set up
a tunnel to further protect messages and also provide
communicating node details and traffic flow confiden-
tiality. As an example, consider the secure remote
access to an aircraft’s device to control it in case of
an emergency. The aircraft’s gateway can be used
to further protect the message and hide the addresses
of communicating entities, while padding can conceal
communication patterns and characteristics.
Compared to IPsec, DTLS demonstrates similar
characteristics, in terms of end-to-end message pro-
tection, but it suffers from an expensive handshake
mechanism and the inability to cope with applications
that utilise the TCP protocol.
One of the problems one should consider when
deploying one of these three solutions, is the use
of robust key management mechanisms designed for
resource-constrained devices. Traditional public key
cryptography solutions are considered inappropri-
ate in some environments and alternatives, includ-
ing those that utilise lightweight cryptography, ellip-
tic curves and identity based cryptography should be
considered.
6 CONCLUSIONS
This paper provided an overview of the various se-
curity issues in ESs design and implementation. The
particular characteristics of such resource-constrained
devices and the varied requirements of their applica-
tions not only introduce new vulnerabilities but also
intensify existing ones. Moreover, mechanisms and
techniques (e.g. for access control, cryptography, net-
work routing etc.) that would typically be deployed
to secure other types of computing devices are not al-
ways applicable or have limited efficacy in the context
of embedded systems.
Further research is required for establishing se-
cure mechanisms tailored for ESs, in order to address
potential threats to their secure operation, including
those exacerbated by the intrinsic characteristics of
the devices and their application fields, especially in
the case of critical systems’ applications. What is
more, given the widespread adoption of smart de-
vices in our everyday lives (e.g. smart vehicles, smart
houses, smart clothing), it is important to deal effec-
tively with the inherent security concerns. The chal-
lenge lies with the researchers to put more effort in
the above matters and come up with appropriate so-
lutions, thus helping realise the promise of pervasive
computing and the Internet of Things (IoT).
ACKNOWLEDGEMENTS
This work has been supported by the Greek
General Secretariat for Research and Technology
(GSRT), under the ARTEMIS JU research program
nSHIELD (new embedded Systems arcHItecturE for
multi-Layer Dependable solutions) project. Call:
ARTEMIS-2010-1, Grant Agreement No.: 269317.
REFERENCES
Aad, I., Hubaux, J., and Knightly, E. W. (2008). Im-
pact of denial of service attacks on ad hoc net-
works. IEEE/ACM Transactions on Networking
(TON), 16(4):791–802.
Abdalrazak, T. R. and Sawant, H. K. (2012). Collaborative
trust-based secure routing based ad-hoc routing pro-
tocol. International Journal of Modern Engineering
Research (IJMER), 2(2):95–101.
Akishita, T. and Hiwatari, H. (2012). Very compact hard-
ware implementations of the blockcipher CLEFIA. In
Miri, A. and Vaudenay, S., editors, 18th international
conference on Selected Areas in Cryptography (SAC
’11), volume 7118 of LNCS, pages 278–292, Toronto,
ON, Canada.
EmbeddedSystemsSecurityChallenges
263
Alam, S., Chowdhury, M. M. R., and Noll, J. (2011). In-
teroperability of security-enabled Internet of Things.
Wireless Personal Communications, 61(3):567–586.
Alhaqbani, B. and Fidge, C. (2008). Access control re-
quirements for processing electronic health records.
In Hofstede, A., Benatallah, B., and Paik, H.-Y., ed-
itors, Business Process Management Workshops, vol-
ume 4928 of Lecture Notes in Computer Science,
pages 371–382. Springer Berlin Heidelberg.
Beaulieu, R., Shors, D., Smith, J., Treatman-Clark, S.,
Weeks, B., and Wingers, L. (2013). The SIMON and
SPECK families of lightweight block ciphers. Cryp-
tology ePrint Archive, Report 2013/404.
Bogdanov, A., Knezevic, M., Leander, G., Toz, D., Varici,
K., and Verbauwhede, I. (2011). SPONGENT: A
lightweight hash function. In Preneel, B. and Tak-
agi, T., editors, 13th International Workshop on Cryp-
tographic Hardware and Embedded Systems (CHES
’11), volume 6917 of LNCS, pages 312–325, Nara,
Japan.
Bogdanov, A., Knudsen, L. R., Leander, G., Paar, C.,
Poschmann, A., Robshaw, M. J. B., Seurin, Y.,
and Vikkelsoe, C. (2007). PRESENT: An ultra-
lightweight block cipher. In Paillier, P. and Ver-
bauwhede, I., editors, 9th International Workshop
on Cryptographic Hardware and Embedded Systems
(CHES 2007), volume 4727 of LNCS, pages 450–466,
Vienna, Austria.
Bogdanov, A., Leander, G., Paar, C., Poschmann, A., Rob-
shaw, M. J. B., and Seurin, Y. (2008). Hash func-
tions and RFID tags: Mind the gap. In Oswald, E.
and Rohatgi, P., editors, 10th International Workshop
on Cryptographic Hardware and Embedded Systems
(CHES 2008), volume 5154 of LNCS, pages 283–299,
Washington, D.C., USA.
Boneh, D. and Franklin, M. (2003). Identity-based encryp-
tion from the Weil pairing. SIAM Journal on Comput-
ing, 32(3):586–615. Also appeared in CRYPTO ’01.
Buchegger, S., Tissieres, C., and Le Boudec, J.-Y. (2004).
A test-bed for misbehavior detection in mobile ad-hoc
networks, how much can watchdogs really do? In
6th IEEE Workshop on Mobile Computing Systems
and Applications (WMCSA ’04), pages 102–111, Low
Wood, Lake Windermere, UK.
Byoungyoung, L., Jinoh, O., Hwanjo, Y., and Jong, K.
(2011). Protecting location privacy using location se-
mantics. In 17th ACM SIGKDD International Con-
ference on Knowledge Discovery and Data Mining
(KDD ’11), pages 1289–1297, San Diego, California,
USA.
Calhoun, P., Loughney, J., Guttman, E., Zorn, G., and
Arkko, J. (2003). Diameter base protocol. RFC 3588,
IETF.
Carl, G., Kesidis, G., Brooks, R. R., and Rai, S. (2006).
Denial-of-service attack detection techniques. IEEE
Internet Computing, 10(1):82–89.
Chen, C.-Y. and Chao, H.-C. (2011). A survey of key dis-
tribution in wireless sensor networks. Security and
Communication Networks.
Damandeep, K. and Jyotsna, S. (2012). Proposed p2p trust
and reputation based model to secure grid. IJCA Pro-
ceedings on International Conference on Recent Ad-
vances and Future Trends in Information Technology
(iRAFIT 2012), iRAFIT(2):19–24.
Dan
ˇ
ek, M., Kadlec, J., Bartosinski, R., and Kohout, L.
(2008). Increasing the level of abstraction in FPGA-
based designs. In Udo, K., editor, International Con-
ference on Field Programmable Logic and Applica-
tions (FPL 2008), pages 5–10, Heidelberg, Germany.
Doomun, M. R. and Soyjaudah, K. M. S. (2009). Analytical
comparison of cryptographic techniques for resource-
constrained wireless security. International Journal of
Network Security, 9(1):82–94.
Doyle, B., Bell, S., Smeaton, A. F., McCusker, K., and
O’Connor, N. (2006). Security considerations and key
negotiation techniques for power constrained sensor
networks. The Computer Journal, 49(4):443–453.
ECRYPT (2008). The eSTREAM project. Available online
at: http://www.ecrypt.eu.org/stream/.
Eisenbarth, T., Kumar, S., Paar, C., Poschmann, A.,
and Uhsadel, L. (2007). A survey of lightweight-
cryptography implementations. IEEE Design & Test,
24(6).
Engels, D., Saarinen, M.-J. O., Schweitzer, P., and Smith,
E. M. (2011). The hummingbird-2 lightweight authen-
ticated encryption algorithm. In Juels, A. and Paar, C.,
editors, 7th Workshop of RFID Security and Privacy
(RFIDSec ’11), volume 7055 of LNCS, pages 19–31,
Amherst, Massachusetts, USA.
Feldhofer, M., Wolkerstorfer, J., and Rijmen, V. (2005).
AES implementation on a grain of sand. IEE Pro-
ceedings on Information Security, 152(1):13–20.
Fysarakis, K., Manifavas, C., Papaefstathiou, I., and
Adamopoulos, A. (2013). A lightweight anonymity
& location privacy service. In IEEE Int. Symposium
on Signal Processing and Information Technology (IS-
SPIT 2013), Athens, Greece. To appear.
Gaj, K., Homsirikamol, E., Rogawski, M., Shahid, R.,
and Sharif, M. U. (2012). Comprehensive evaluation
of high-speed and medium-speed implementations of
five SHA-3 finalists using Xilinx and Altera FPGAs.
Cryptology ePrint Archive, Report 2012/368.
Gedik, B. and Liu, L. (2008). Protecting location privacy
with personalized k-anonymity: Architecture and al-
gorithms. IEEE Transactions on Mobile Computing,
7(1):1–18.
Golle, P. and Partridge, K. (2009). On the anonymity of
home/work location pairs. In Tokuda, H., Beigl, M.,
Friday, A., Brush, A. J. B., and Tobe, Y., editors,
7th International Conference on Pervasive Computing
(Pervasive 2009), volume 5538 of LNCS, pages 390–
397, Nara, Japan.
Gruteser, M. and Grunwald, D. (2005). Enhancing location
privacy in wireless LAN through disposable interface
identifiers: A quantitative analysis. Mobile Networks
and Applications, 10(3):315–325.
Gruteser, M. and Hoh, B. (2005). On the anonymity of pe-
riodic location samples. In Hutter, D. and Ullmann,
M., editors, 2nd International Conference on Security
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
264
in Pervasive Computing (SPC 2005), volume 3450 of
LNCS, pages 179–192, Boppard, Germany.
Guo, J., Peyrin, T., and Poschmann, A. (2011). The PHO-
TON family of lightweight hash functions. In Rog-
away, P., editor, 31st annual conference on Advances
in cryptology (CRYPTO ’11), volume 6841 of LNCS,
Santa Barbara, CA, USA.
Gupta, K., Yadav, A. S., and Yadav, S. (2011). Location pri-
vacy using user anonymity and dummy locations. In-
ternational Journal of Innovative Technology & Cre-
ative Engineering, 1(10):5–8.
Hatzivasilis, G. and Manifavas, C. (2012). Building trust
in ad hoc distributed resource-sharing networks using
reputation-based systems. In 16th Panhellenic Confer-
ence on Informatics, pages 416–421, Piraeus, Greece.
Hatzivasilis, G., Theodoridis, A., Gasparis, H., Manifavas,
C., and Papaefstathiou, I. (2014). ULCL: An ultra-
lightweight cryptographic library for embedded sys-
tems. In Measurable security for Embedded Comput-
ing and Communication Systems (MeSeCCS 2014),
Lisbon, Portugal. To appear.
Hoh, B. and Gruteser, M. (2005). Protecting location pri-
vacy through path confusion. In 1st International
Conference on Security and Privacy for Emerging Ar-
eas in Communications Networks (SECURECOMM
’05), pages 194–205, Athens, Greece.
Huang, J., Buckingham, J., and Han, R. (2005a). A
level key infrastructure for secure and efficient group
communication in wireless sensor networks. In 1st
IEEE/CreateNet Conference on Security and Privacy
for Emerging Areas in Communication Networks (Se-
cureComm 2005), pages 249–260, Athens, Greece.
Huang, L., Matsuura, K., Yamane, H., and Sezaki, K.
(2005b). Enhancing wireless location privacy using
silent period. In IEEE Wireless Communications and
Networking Conference (WCNC ’05), volume 2, New
Orleans, LA, USA.
Hui, J. and Thubert, P. (2011). Compression format for IPv6
datagrams over IEEE 802.15.4-based networks. RFC
6282, IETF.
IEEE Standard for Local and metropolitan area networks
(2011). IEEE standard for local and metropoli-
tan area networks–Part 15.4: Low-Rate Wireless
Personal Area Networks (LR-WPANs). Avail-
able online at: http://standards.ieee.org/getieee802/
download/802.15.4-2011.pdf.
Kamal, A. A. and Youssef, A. M. (2009). An FPGA im-
plementation of the NTRUEncrypt cryptosystem. In
International Conference on Microelectronics (ICM),
pages 209–212.
Karakoc¸, F., Demirci, H., and Harmanci, A. E. (2013).
ITUbee: A software oriented lightweight block cipher.
In 2nd International workshop on lightweight cryp-
tography for security & privacy (LightSEC 2013).
Karalis, A., Joannopoulos, J. D., and Solja
ˇ
ci
´
c, M. (2008).
Efficient wireless non-radiative mid-range energy
transfer. Annals of Physics, 323(1):34–48.
Koscher, K., Czeskis, A., Roesner, F., Patel, S., Kohno, T.,
Checkoway, S., McCoy, D., Kantor, B., Anderson, D.,
Shacham, H., and Savage, S. (2010). Experimental
security analysis of a modern automobile. In IEEE
Symposium on Security and Privacy (SP), pages 447–
462, Oakland, CA, USA.
Kothmayr, T., Schmitt, C., Hu, W., Bruenig, M., and Carle,
G. (2012). A DTLS based end-to-end security archi-
tecture for the Internet of Things with two-way au-
thentication. In IEEE 37th Conference on Local Com-
puter Networks Workshops (LCN Workshops), pages
956–963, Clearwater, FL, USA.
Kuo, C., Luk, M., Negi, R., and Perrig, A. (2007). Message-
in-a-bottle: User-friendly and secure key deployment
for sensor nodes. In 5th International Conference on
Embedded Networked Sensor Systems (SenSys ’07),
pages 233–246, Sydney, Australia.
Kurs, A., Karalis, A., Moffatt, R., Joannopoulos, J. D.,
Fisher, P., and Solja
ˇ
ci
´
c, M. (2007). Wireless power
transfer via strongly coupled magnetic resonances.
Science, 317(5834):83–86.
Liu, L. (2007). From data privacy to location privacy: Mod-
els and algorithms. In 33rd international conference
on Very large data bases (VLDB ’07), pages 1429–
1430, Vienna, Austria. VLDB Endowment.
Liu, L. (2009). Privacy and location anonymization
in location-based services. SIGSPATIAL Special,
1(2):15–22.
Machanavajjhala, A., Kifer, D., Gehrke, J., and Venkitasub-
ramaniam, M. (2007). L-diversity: Privacy beyond
k-anonymity. ACM Transactions on Knowledge Dis-
covery from Data, 1(1). Article No. 3.
Madhavi, S. and Kim, T. H. (2011). An intelligent dis-
tributed reputation based mobile intrusion detection
system. International Journal of Computer Science
and Telecommunications, 2(7):53–58.
Manifavas, C., Hatzivasilis, G., Fysarakis, K., and Rantos,
K. (2013). Lightweight cryptography for embedded
systems A comparative analysis. In 6th Interna-
tions Workshop on Autonomous and Spontaneous Se-
curity (SETOP 2013), volume 8247 of Lecture Notes
in Computer Science, pages 371–382. Springer-Verlag
Berlin Heidelberg, RHUL, Egham, U.K.
Martin, K. M. and Paterson, M. B. (2008). An application-
oriented framework for wireless sensor network key
establishment. Electronic Notes in Theoretical Com-
puter Science, 192(2):31–41.
Naedele, M. (2006). An access control protocol for embed-
ded devices. In 4th International IEEE Conference on
Industrial Informatics (INDIN ’06), pages 565–569,
Singapore.
Oliveira, L. B., Aranha, D. F., Gouv
ˆ
ea, C. P. L., Scott,
M., C
ˆ
amara, D. F., L
´
opez, J., and Dahab, R. (2011).
TinyPBC: Pairings for authenticated identity-based
non-interactive key distribution in sensor networks.
Computer Communications, 34(3):485–493.
Preneel, B. (2009). Research challenges in lightweight
cryptography. Key Note Talk at the 2nd ACM con-
ference on Wireless network security (WiSec ’09).
Rantos, K., Papanikolaou, A., Fysarakis, K., and Manifavas,
C. (2012). Secure policy-based management solu-
tions in heterogeneous embedded systems networks.
EmbeddedSystemsSecurityChallenges
265
In IEEE International conference on Telecomunica-
tions and Multimedia (TEMU 2012), pages 227–232,
Heraklion, Crete, Grece.
Rantos, K., Papanikolaou, A., and Manifavas, C. (2013a).
IPsec over IEEE 802.15.4 for low power and lossy
networks. In 11th ACM International Symposium on
Mobility Management and Wireless Access (MobiWac
2013), pages 59–64, Barcelona, Spain.
Rantos, K., Papanikolaou, A., Manifavas, C., and Papaefs-
tathiou, I. (2013b). IPv6 security for low power and
lossy networks. In Wireless Days 2013 (WD 2013),
Valencia, Spain. To appear.
Raymond, D. R. and Midkiff, S. F. (2008). Denial-of-
service in wireless sensor networks: Attacks and de-
fenses. IEEE Pervasive Computing, 7(1):74–81.
Raza, S., Duquennoy, S., Chung, T., Yazar, D., Voigt, T.,
and Roedig, U. (2011). Securing communication in
6LoWPAN with compressed IPsec. In 7th IEEE In-
ternational Conference on Distributed Computing in
Sensor Systems (DCOSS ’11), pages 1–8, Barcelona,
Spain.
Resnick, P. and Zeckhauser, R. (2002). Trust among
strangers in internet transactions: Empirical analysis
of eBay’s reputation system. In Baye, M. R., edi-
tor, The Economics of the Internet and E-commerce,
volume 11 of Advances in Applied Microeconomics,
pages 127–157. Emerald Group Publishing Limited.
Shibutani, K., Isobe, T., Hiwarati, H., Mitsuda, A., Ak-
ishita, T., and Shirai, T. (2011). Piccolo: An ultra-
lightweight blockcipher. In Preneel, B. and Takagi,
T., editors, 13th International Workshop on Cryp-
tographic Hardware and Embedded Systems (CHES
’11), volume 6917 of LNCS, pages 342–357, Nara,
Japan.
Simplicio, Jr., M. A., Barreto, P. S. L. M., Margi, C. B., and
Carvalho, T. C. M. B. (2010). A survey on key man-
agement mechanisms for distributed wireless sensor
networks. Computer Networks, 54(15):2591–2612.
Smith, R. (2008). Phlashdance, discovering permanent de-
nial of service attacks against embedded systems. Talk
at the EUSecWest Applied Security Conference. Lon-
don.
Stefanidis, K. and Serpanos, D. N. (2008). Implementing
filtering and traceback mechanism for packet-marking
IP-based traceback schemes against DDoS attacks. In
4th International IEEE Conference on Intelligent Sys-
tems (IS ’08), volume 2, pages 14–28, Varna, Bul-
garia.
Szilagyi, C. and Koopman, P. (2009). Flexible multicast au-
thentication for time-triggered embedded control net-
work applications. In IEEE/IFIP International Con-
ference on Dependable Systems & Networks (DSN
’09), pages 165–174, Lisbon, Portugal.
Szilagyi, C. and Koopman, P. (2010). Low cost multicast
authentication via validity voting in time-triggered
embedded control networks. In 5th Workshop on Em-
bedded Systems Security (WESS ’10), pages 10:1–
10:10, Scottsdale, Arizona.
Trusted Platform Module (2009). Information tech-
nology Trusted Platform Module Part 1:
Overview. ISO/IEC 11889-1:2009. Available online
at: http://www.trustedcomputinggroup.org/resources/
tpm main specification.
Watanabe, D., Ideguchi, K., Kitahara, J., Muto, K., and Fu-
ruichi, H. (2008). Enocoro-80: A hardware oriented
stream cipher. In 3rd International Conference on
Availability, Reliability and Security (ARES 08), pages
1294–1300.
Yoshihama, S., Ebringer, T., Nakamura, M., Munetoh,
S., Mishina, T., and Maruyama, H. (2007). WS-
attestation: Enabling trusted computing on web ser-
vices. In Baresi, L. and Di Nitto, E., editors, Test and
Analysis of Web Services, pages 441–469. Springer
Berlin Heidelberg.
Zeeb, E., Moritz, G., Timmermann, D., and Golatowski, F.
(2010). WS4D: Toolkits for networked embedded sys-
tems based on the devices profile for web services. In
39th International Conference on Parallel Processing
Workshops (ICPPW ’10), pages 1–8, San Diego, CA,
USA.
Zhang, Y., Xu, L., and Wang, X. (2008). A cooperative
secure routing protocol based on reputation system
for ad hoc networks. Journal of Communications,
3(6):43–50.
Zhong, G. and Hengartner, U. (2008). Toward a distributed
k-anonymity protocol for location privacy. In 7th
ACM workshop on Privacy in the electronic society
(WPES ’08), pages 33–38, Alexandria, VA, USA.
PECCS2014-InternationalConferenceonPervasiveandEmbeddedComputingandCommunicationSystems
266