Link reliable and Trust aware RPL routing protocol for Internet of

Things

Asma LAHBIB∗ , Khalifa Toumi∗ , Sameh Elleuch† , Anis Laouiti∗ ,and Steven Martin‡
∗ SAMOVAR, Télécom SudParis, CNRS, Université Paris-Saclay, 9 rue Charles Fourier 91011 Evry, France.
Email: {asma.lahbib, khalifa.toumi, anis.laouiti}@telecom-sudparis.eu
† École nationale des sciences de l’informatique, Campus Universitaire de la Manouba, 2010 Manouba, Tunisie
Email: sameh.elleuch@ensi-uma.tn
‡ LRI, Université Paris-Sud, 15 Rue Georges Clemenceau, 91400 Orsay, France.
Email: steven.martin@lri.fr

Abstract— Internet of Things (IoT) is characterized by
heterogeneous devices that interact with each other on
a collaborative basis to fulfill a common goal. In this
scenario, some of the deployed devices are expected to be
constrained in terms of memory usage, power consumption
and processing resources. To address the specific proper-
ties and constraints of such networks, a complete stack of
standardized protocols has been developed, among them
the Routing Protocol for Low-Power and lossy networks
(RPL). However, this protocol is exposed to a large variety
of attacks from the inside of the network itself. To fill this
gap, this paper focuses on the design and the integration of
a novel Link reliable and Trust aware model into the RPL
protocol. Our approach aims to ensure Trust among enti-
ties and to provide QoS guarantees during the construction
and the maintenance of the network routing topology. Our
model targets both node and link Trust and follows a
multidimensional approach to enable an accurate Trust
value computation for IoT entities. To prove the efficiency
of our proposal, this last has been implemented and tested
successfully within an IoT environment. Therefore, a set
of experiments has been made to show the high accuracy
level of our system.
Index Terms— Trust model, RPL, IoT, Energy efficiency,
QoS, Security.
I. INTRODUCTION
As an emerging technology, the Internet of Things (IoT) is
expected to offer promising solutions to potentially transform
and revolutionize the services provided by various application
domains. This paradigm could be defined as the integration of
the physical world with the digital, cyber, and virtual worlds
on a global scale. The requirement of remote sensing makes
sensor networks and low power personal area networks one
of the key enabling technologies of IoT [1], [2], [3]. These
networks present several challenges, especially in communi-
cation and networking due to their inherent constraints (i.e.
scare resources, limited capabilities, etc.). For this reason,
many IoT standards were proposed, among them the ipv6
over LOW power wireless Personal Area Networks (6lowPAN)
adaptation layer introduced by the 6LoWPAN IETF group [4]
to make IPv6 addressing appropriate for low power wireless
networks [4], [5]. Taking into consideration the specification
of 6LoWPAN, an increasing need for an efficient routing
protocol for 6LoWPAN-compliant IoT networks has arisen.
To address this gap, the IETF ROLL working group [6],
[7] has proposed a routing protocol, referred to as Routing
Protocol for Low power and lossy networks (RPL).
This protocol was specially designed to overcome the different
978-1-5386-1465-5/17/$31.00 2017
c IEEE
challenges posed by these networks (i.e. lossy links, high
Packet Error Rate (PER), link outages, constrained nodes,
and scare energy resources, etc.). However, their inherent
constraints and their exposure to varied risks and security
attacks have turned the design of RPL challenging especially
since the proposed security specifications in [6] do not address
all possible attacks that may compromise the RPL network
security. In this context, several solutions [8], [9] have been
proposed in an attempt to bring some enhancements to
the RPL standard specification and to meet the Quality of
Service (QoS) requirements of these networks. However, little
attention have been paid to the security concept and just some
works [16], [17], [18] have incorporated the Trust management
aspect within the RPL routing procedures.
In this direction, this work focuses mainly on the Trust
aspect of RPL. We propose some enhancements in order to
ensure Trust among entities and to provide QoS guarantees
during the construction and the maintenance of the routing
topology. To deal with the different characteristics of IoT
networks, this paper proposes LT-RPL, a novel Link reliable
and Trust aware model for RPL protocol. Such model follows
a multidimensional approach to enable an accurate Trust
value computation for IoT entities.
In particular, unlike previous models that usually only con-
sider experience and reputation parameters to evaluate Trust,
this model uses security aspects, QoS factors and energy
considerations in addition to the reputation parameters which
are used by network entities to assess the Trustworthiness
of their neighboring ones. Thereafter, the proposed model
has been implemented and integrated into the RPL topology
construction and maintenance phases. Thus a set of evalua-
tion results are analyzed and discussed to demonstrate the
feasibility of LT-RPL.
The remainder of this paper is structured as follows. Section
2 discusses the related proposals regarding Trust models for
routing protocols. Thereafter, Section 3 provides a brief de-
scription of the network model. Section 4 provides an overview
of the proposed scheme and the main blocks it relies on, it
delves then into the integration of the proposed scheme within
RPL. In Section 5, a set of experimental results validating our
approach are shown, and finally in Section 6, the paper ends
up with some conclusions and an outlook of the envisaged
future work.
 II. Review of existing works
In the context of IoT, we found RPL as the main candidate
for acting as the standard routing protocol for IoT based
networks and more specifically for LLN networks [10], [11].
Current works regarding RPL have focused on performance
evaluation [12], [13] and just few ones have addressed its
security aspects [14], [15]. In fact, this protocol originally
provides support for integrity, confidentiality and authentic-
ity of messages [10] and it has basically three basic secu-
rity modes namely unsecured, preinstalled and authenticated
modes. However, even RPL is expected to support message
confidentiality, integrity and authenticity, its security may be
compromised and it remains exposed to various threats and
attacks [15].
To secure the communication in an RPL-based network,
authors in [16] made use of the Trusted Platform Module
(TPM) based approach to establish the nodes Trust before
exchanging keying material. However, the Trustworthiness
assessment was considered only for the keys exchange mech-
anism to verify the identity of the keys suppliers and not for
the routing path selection and establishment.
In [17], authors have proposed Trusted-RPL in order to
strengthen the RPL protocol by adding a new Trust metric
based on the evaluation of the neighboring nodes’ behav-
iors during the construction and the maintenance of RPL
instances using selfishness, energy, and honesty components.
However, they had not considered the Trust value along the
path nor proven the defense of the proposed scheme against
the routing attacks that could be launched.
In [18], authors have presented SecTrust, a lightweight SE-
Cure Trust-based routing framework for IoT. The Trust
evaluation is based specifically on the successful interactions
between IoT nodes and its calculation is based on some
metrics such as the prospect of the positive interactions, the
nodes’ satisfaction and their energy level as well. However,
although the proposed framework was designed to isolate
common routing attacks, its effectiveness under these threats
has not been proven nor evaluated.
Current literature still lacks a comprehensive study on Trust
based secure routing in IoT. In this context several issues need
to be seriously considered and more investigated.
First, the Trust integration within the routing functions could
affect the performance of the routing protocol as longer paths
could be selected to avoid malicious nodes and thus it could
cause a more important delay and energy. Hence, we need
further research to minimize these effects and to optimize the
security considerations to be implemented.
Second, more research is required in terms of energy consump-
tion, quality of service consideration, and attacks resiliency
issues within the routing protocols. In fact proposed schemes
are generally designed to defend a specific class of attacks
(those targeting the RPL topology) while Trust management
could deal with various kind of attacks while meeting both
the energy and the QoS requirements that could effectively
improve the security of such a protocol.
III. System model and problem statement
Before elaborating the detailed design of the proposed
scheme, we will provide first a broad overview of the RPL
routing protocol as well as its main features. Then we will
present a brief description of the network model, its archi-
tecture and the nature of the nodes to be involved as well.
Afterwards, a representation of the important properties to
be considered within the proposed scheme will be given within
the problem statement paragraph.
A. Overview on the RPL routing protocol
RPL is an IPv6 routing protocol specifically designed for
LLNs with very limited resources in terms of energy, computa-
tion and bandwidth. This protocol mainly targets collection-
based networks made up of nodes interconnected according
to a specific topology called Destination Oriented Directed
Acyclic Graphs (DODAG), where most popular destination
nodes (i.e. sinks) or those providing a default route to the
Internet (i.e. gateways) act as the roots of the Directed
Acyclic Graphs (DAGs). Within each DODAG each node is
assigned a rank representing its position in the graph. Its
computation depends on a set of specific routing metrics (e.g.
delay, link quality, throughput, etc.). The translation of these
metrics into ranks is based on an Objective Function (OF)
responsible for rank computation and parent selection. The
DODAG construction and maintenance phases are based on
a set of control messages namely DODAG Information Object
(DIO), DODAG Information Solicitation (DIS), Destination
Advertisement Object (DAO) and Destination Advertisement
Object Acknowledgement (DAO-ACK) messages.
B. System model
1) Network model
We consider an IoT network consisting of several groups
made up by a set of smart objects randomly scattered in a
circular network. Each object has a Trust Manager which is
an entity in charge of assessing the Trustworthiness degree of
another object or link according to the proposed model in Sec.
IV. For the constrained objects (i.e., wireless sensors, RFID
readers, etc.), the Trust manager is assumed to be deployed
in another powerful network entity. Furthermore, each group
is monitored by a monitoring and analysis tool that captures
the 6LowPAN traffic, renders the network state and identifies
the abnormal behaviors related to the RPL routing protocol.
2) Adversary model
We consider a behavior as malicious if it is characterized
by a node aiming to modify the topology of the network
and to degrade the performance of the routing protocol. This
node after managing to attract the traffic by advertising
falsified information data, can drop either all, randomly or
selectively packets received and supposed to be forwarded.
Respectively a selfish behavior is conducted by a node that
does not forward packets in order to save its residual energy
for example. Consequently, a malicious node can simulate the
following two types of attacks:
(i) Black-Hole Attack: In this attack, the malicious node
dumps all the packets supposed to be forwarded. However, it
participates devotedly in the route establishment process [15].
(ii) Gray-Hole Attack: This attack is similar to the black-
hole attack. However, the malicious node here only discard a
specific subpart of the network traffic and forward the other
part at random interval.
C. Problem statement
In what follows, we will identify the properties that should
be respected in order to establish Trusted routes in IoT
networks.
(i) Trust: a route is Trusted if only Trusted nodes can
participate in its establishment, design and maintenance. On
the other side malicious and selfish nodes will be isolated
and excluded from participating in the routing procedures.
This property can be assessed accurately on the basis of the
entities historical interactions and their observation of each
other forwarding behavior.
(ii) Delay awareness: a delay aware route should be
able to provide low end-to-end delays from one source to
a destination through a particular route, thus it can be
measured through the offered end-to-end delay.
(iii) Energy Efficiency: a route is energy efficient if it is
made up of nodes that have more energy than any other
node. Obviously nodes with low battery levels should be
avoided in the routing process to extend the network lifetime.
The nodes remaining energy could be so considered within
the topology construction process to maximize the network
lifetime.
(iv) Reliability: a route is reliable if it continuously provide
available and high quality of the communication links along
the path. This property can be evaluated through the link
quality estimators, such as the Packet Reception Ratio
(PRR), the Packet Error Rate (PER) and the Expected
Transmission count (ETX).
Taking into account these four properties, our model shall
respect several constraints such as the maximization of the
network lifetime, the optimization of the quality of service
and the improvement of the security performance.
IV. Proposed scheme
In this section, we will describe the overall architecture of
LT-RPL. To efficiently compute the Trust values of IoT nodes,
we first need to clearly understand the main meaning of Trust
as it is considered in LT-RPL. Afterwards, a brief description
of the different blocks and dimensions will be provided.
A. Trust presentation and definition
Trust here could be defined as a relationship between
two entities, a Trustor and a Trustee. The Trustor is the
evaluating entity willing to join the DAG routing structure
or to update its preferred DAG parents to send its data
packets. On the other side, the Trustee is the evaluated entity
which represents the candidate entity that will be chosen as
the preferred parent. This relationship is derived from direct
observations and interactions referred as the direct Trust and
the recommendations exchanged between neighboring nodes
termed as the indirect Trust. We use hereafter the following
notation to specify the Trust relationship between ei and ej at
a given time t as Trust is time dependent: T rust(ei −→ ej )t .
B. Overview of the proposed model
The design of LT-RPL consists of several phases namely:
Information gathering, Trust composition, Trust database
and Trust application as presented in Fig. 1. It includes as
well four main dimensions specifically QoS dimension, Energy
awareness dimension, Reputation dimension and Security di-
mension.
1) Information gathering
The Trust related information are collected regarding the
entities’ behavior as well as the links’ quality indicators
and taking into consideration two different sources, (a) the
Fig. 1. Architecture of the proposed model
direct observations and (b) the recommendations given by the
neighboring entities.
The Trust structure to be sent to the Trust manager is made
up of the following information: node ID, neighbor ID, RE
percentage, PFR value, ETX value, PRR value, PER value,
the transmission delay as well as the entity time.
2) Trust composition
The Trust Manager compute the Trust level for each entity
specifically the Node related Trust (NT) and the Link related
Trust (LT) as follow:

0 if (alert generated)
T (ei −→ ej )t = else
w1 ∗ N T (ei −→ ej )t + w2 ∗ LT (ei −→ ej )t
Where T (ei −→ ej )t denotes the value of Trust the entity ei
has for its candidate parent ej at time t and it is limited to
a continuous range from 0 to 1, where 0 denotes complete
distrust whereas 1 represents absolute Trust. The weight
factors w1 and w2 are assigned to N T (ei −→ ej )t and
LT (ei −→ ej )t respectively where w1 + w2 = 1; 0 ≤ w1 ≤ 1
and 0 ≤ w2 ≤ 1.
Each computation is based on a set of properties where
N T (ei −→ ej )t represents the NT level calculated based
on the Trustor’s direct observation of its one hop neighbors’
behavior referred as the direct Trust N T (ei −→ ej )dt and on
the other hand, on the third parties’ attributed recommenda-
tions called the indirect or the relative Trust N T (ei −→ ej )rt
as follow: N T (ei −→ ej )t = wd ∗ N T (ei −→ ej )dt + wr ∗
N T (ei −→ ej )rt , wd and wr are the weights assigned to the
direct and the indirect Trust respectively.
The direct Trust is calculated by considering both node
cooperativeness (coop) and node competence (comp). At time
t, it is defined as:
N T (ei −→ ej )dt = N T (ei −→ ej )coop
t ∗ N T (ei −→ ej )comp
t ,
where N T (ei −→ ej )coop
t reflects the cooperativeness level
evaluated during the time interval [0..t] and calculated using
the Packet Forwarding Ratio (PFR), while N T (ei −→ ej )comp t
provides the degree of the entity’s ability to perform its
intended tasks within the routing process and it is assessed
based on the Remaining Energy (RE) percentage hence the
energy dimension of our model.
On the other side, the indirect Trust N T (ei −→ ej )rt is
set up upon recommendations of other entities within the
neighborhood which reflects here the reputation dimension. In
order to obtain Trust recommendations, we first need to select
Trust recommenders with a Trusted communication link, and
thus get rid of the impact of malicious recommendations. To
do so each recommendation RTkj about ej given by ek is
subject to a credibility factor CRk in the interval [0..1], where
 Simulation parameter Value
1 represents the highest credibility and 0 the lowest one.
Simulation tool Contiki/Cooja 2.7
When it comes to the link Trust LT (ei −→ ej )t , its evaluation Mote type Tmote Sky
aims mainly to reinforce the routing DAG establishment and Simulation run time 3600 seconds
maintenance processes by considering both the quality (qual) Simulation coverage area 300m x 300m
and the performance (perf) of the different links connecting Interference range 100m
Wireless communication 50m
the participating entities in order to successfully meet the QoS range
requirements, its value is defined as follow: Total number of nodes 10..50
LT (ei −→ ej )t = LT (ei −→ ej )qual t ∗ LT (ei −→ ej )perf
t Number of malicious nodes 3..15
qual
where LT (ei −→ ej )t reflects the belief that the connecting Radio environment DGRM (Directed
Graph Radio Medium)
link is efficient enough to respect the QoS guarantees. It is
measured by ETX and PRR as indicators of the link quality TABLE I
between the entity and its neighbor. While LT (ei −→ ej )perf
t Network related parameters used in simulation analysis
characterize the performance of the link based on the PER
and the transmission delay L.
The combination of these properties will produce an overall
Trust value that can be used efficiently and effectively to more specifically with RPL based on the Minimum Rank with
ensure security improvement for the RPL routing process. Hysteresis Objective Function (MRHOF) which uses hystere-
3) Trust storage sis while selecting the path with the smallest metric value. We
Once the Trust composition process has been completed, have varied the number of network nodes, the percentage of
Trust related evidence will be stored by the Trust manager in malicious ones and we have analyzed the corresponding effect
a Trust record table that contains apart the Trust information respectively on the packet loss ratio, the remaining energy
that each entity has gathered for its candidate neighbors, the percentage and the resiliency to black-hole attacks.
Trust value computed according to the different properties as 1) Remaining energy
it has been already explained. To enforce the security aspect We measure here the RE while varying the number of
of the proposed scheme, a hash algorithm has been employed nodes within a network composed of a single DODAG. Fig.
to encrypt the Trust values when stored and retrieved from 2 illustrates the distribution of the RE between the network
the Trust database or when sent to the evaluating entities. nodes.
4) Trust application: LT-RPL: Link reliable and Trust As seen, the RE distribution in LT-RPL based network is
aware RPL routing protocol
The Trust model previously described has been integrated
into the DODAG construction and maintenance phases of
RPL. For the implementation of LT-RPL, each node sends to
its neighbors the value of its rank which is included by default
in the DIO message, once received the evaluating node ejoin
will check its record table for the most recent Trust values
of its p ≥ 1 candidate parents ecand1 ..ecandp , already sent by
the Trust manager. Afterwards, the rank R(ejoin −→ ecandq )
will be computed according to the Trust based OF and with
respect to each candidate parent ecandq , according to the
formula below: Fig. 2. Remaining energy distribution

R(ejoin −→ ecandq ) = R(ecandq ) − T (ejoin −→ ecandq )t
where R(ecandq ) is the rank value of the candidate parent.
Afterwards, the node with the minimum rank R(ejoin −→
ecandq ) will be chosen as the preferred parent to reach the
root. Otherwise, it will be rejected and identified as malicious
parent, thus punished and removed from the list of candidate
entities.
V. Simulation and Results
A. Simulation setup
Our experiments are performed using the InstantContiki
2.7 platform while integrating the proposed Trust model
(presented in Section IV) into the RPL routing protocol.
The various simulation parameters are listed in Table I. In
the simulation study, we have assumed that the attacking
nodes behave as good nodes from inception and begin their
malicious activities during time (when activated).
B. Performance evaluation results
In order to prove the performance of LT-RPL, we have
performed several simulations in comparison with RPL and
more balanced than the normal RPL insofar as relay nodes are
chosen function of their RE level during the DAG construction
phase which avoids having the same relay node within the
routing topology and consequently depleting the power of this
last in a faster way as well. The simulation results here reveal
that within LT-RPL, 44% of nodes have a RE above 80%, 14%
above 90% and 64% between 70% and 80%. However, we note
that within RPL-MRHOF, just 5% of nodes have a RE above
90% and 16% under 70% which could adversely affect the
network lifetime over time as 16% of nodes are likely becoming
to exhaust their batteries in a faster way in comparison with
LT-RPL.
2) Packet loss ratio
We measure here the Packet Loss Ratio (PLR) while vary-
ing the number of legitimate and malicious nodes.
Fig. 3 shows the variation of the PLR in the presence of a
number of malicious nodes.
The simulation result in Fig. 3 shows that LT-RPL experi-
ences a more reduced packet loss than RPL under black-hole
attacks scenarios where malicious nodes advertise themselves
to their neighbor nodes as better parents and routes in a guise

Fig. 3. Packet loss ratio comparison between RPL (MRHOF) and
LT-RPL under black-hole attacks
to attract these unsuspecting nodes while eventually dropping
their packets supposed to be forwarded. In fact, while LT-RPL
related PLR stayed below 0.4, the standard RPL (MRHOF)
recorded a staggering one between 0.6 and 0.95.
To demonstrate the reliability of LT-RPL based network in
the presence of malicious nodes (launching black-hole at-
tacks), we have presented the impact of their presence on
the packet loss. The simulation result in Fig. 4 shows well
Fig. 4. Packet loss ratio evolution of LT-RPL in a 50 nodes network
size
how LT-RPL detects the malicious behavior of nodes and
how it reacts to its presence. The network here is composed
of 50 nodes where 15 among them start to act maliciously
over a certain time after the network initialization. As shown
in Fig. 4, once the malicious nodes begin their malicious
activities consisting of dropping their neighbors’ packets, the
PLR considerably increase in response to the occurring event.
However the integration of the proposed scheme within RPL
was shown and proven to be effective in reducing the impact
of such behaviors insofar as the attacking nodes will not be
chosen anymore as a next hop nodes during the network
routing topology construction. As a result thereof, the PLR
is reduced and maintained considerably stable over time.
VI. Conclusion
In this paper, we have proposed a Trust management
scheme for securing the network routing topology construc-
tion and maintenance within the RPL routing protocol. Our
Trust model was applied to the RPL protocol, evaluated and
compared with ETX based MRHOF objective function. As a
general remark, LT-RPL proves to be efficient and effective
in detecting and preventing malicious nodes launching black-
hole and gray-hole attacks while suitably providing QoS guar-
antees required for energy-efficient routing in constrained IoT
networks. As a future work, It would be interesting to improve
the performance of our model in case of other types of attack
scenarios, we may consider the attacks targeting the RPL
network traffic as well as the network nodes resources. We also
aim at deploying LT-RPL in different application scenarios in
order to prove its adaptability and applicability to different
platforms and environments.
References
[1] Miorandi, D., Sicari, S., De Pellegrini, F., & Chlamtac, I.
(2012). Internet of things: Vision, applications and research
challenges. Ad Hoc Networks, 10(7), 1497-1516.
[2] Gubbi, J., Buyya, R., Marusic, S., & Palaniswami, M. (2013).
Internet of Things (IoT): A vision, architectural elements, and
future directions. Future generation computer systems, 29(7),
1645-1660.
[3] Al-Fuqaha, A., Guizani, M., Mohammadi, M., Aledhari, M.,
& Ayyash, M. (2015). Internet of things: A survey on enabling
technologies, protocols, and applications. IEEE Communica-
tions Surveys & Tutorials, 17(4), 2347-2376.
[4] Montenegro, G., Kushalnagar, N., Hui, J., & Culler. D. (2007).
Transmission of IPv6 packets over IEEE 802.15. 4 networks.
Internet Eng. Task Force (IETF), Fremont, CA, USA, Internet
Proposed Std. RFC 4944.
[5] Kushalnagar, N., Montenegro, G., & Schumacher, C. (2007).
IPv6 over Low-Power Wireless Personal Area Networks
(6LoWPANs): Overview, assumptions, problem statement,
and goals. Internet Eng. Task Force (IETF), Fremont, CA,
USA, RFC4919, vol. 10.
[6] Winter, T. (2012). RPL: IPv6 routing protocol for low-power
and lossy networks.
[7] Routing Over Low Power and Lossy Networks (ROLL), 2004.
Available: https://datatracker.ietf.org/wg/roll/charter/.
[8] Di Marco, P., Fischione, C., Athanasiou, G., Mekikis, P.V.
(2013). MAC-aware routing metrics for low power and lossy
networks. INFOCOM, 2013 Proceedings IEEE, 13-14.
[9] Iova, O., Theoleyre, F., Noel, T. (2015). Using multiparent
routing in RPL to increase the stability and the lifetime of the
network. Ad Hoc Networks 29(0), 45-62.
[10] Winter, T. (2012). RPL: IPv6 routing protocol for low-power
and lossy networks.
[11] Sheng, Z., Yang, S., Yu, Y., Vasilakos, A., Mccann, J., &
Leung, K. (2013). A survey on the ietf protocol suite for the
internet of things: Standards, challenges, and opportunities.
IEEE Wireless Communications, 20(6), 91-98.
[12] Tripathi, J., de Oliveira, J. C., & Vasseur, J. P. (2010). A
performance evaluation study of rpl: Routing protocol for low
power and lossy networks. Information Sciences and Systems
(CISS) 44th Annual Conference on (pp. 1-6). IEEE.
[13] Accettura, N., Grieco, L. A., Boggia, G., & Camarda, P.
(2011). Performance analysis of the RPL routing protocol. In
Mechatronics (ICM) IEEE International Conference on (pp.
767-772). IEEE.
[14] Sehgal, A., Mayzaud, A., Badonnel, R., Chrisment, I., &
Schonwalder, J. (2014). Addressing DODDAG inconsistency
attacks in RPL networks. In Proceedings of Global Informa-
tion Infrastructure and Networking Symposium (GIIS’14), 1-8.
[15] Mayzaud, A., Badonnel, R., & Chrisment, I. (2016). A Taxon-
omy of Attacks in RPL-based Internet of Things. International
Journal of Network Security, 18(3), 459-473.
[16] Seeber, S., Sehgal, A., Stelte, B., Rodosek, G. D., & Schon-
walder, J. (2013, October). Towards a trust computing ar-
chitecture for RPL in Cyber Physical Systems. In Network
and Service Management (CNSM), 2013 9th International
Conference on (pp. 134-137). IEEE.
[17] Djedjig, N., Tandjaoui, D., & Medjek, F. (2015, July). Trust-
based RPL for the Internet of Things. In Computers and
Communication (ISCC), 2015 IEEE Symposium on (pp. 962-
967). IEEE.
[18] Airehrour, D., Jairo, G., & Sayan, K.R. (2016). A Lightweight
Trust Design for IoT Routing. Dependable, Autonomic and
Secure Computing, 14th Intl Conf on Pervasive Intelligence
and Computing, 2nd Intl Conf on Big Data Intelligence
and Computing and Cyber Science and Technology Congress
(DASC/PiCom/DataCom/CyberSciTech), 2016 IEEE 14th
Intl C. IEEE.