International Journal of Computer Applications (0975 – 8887)

Volume 160 – No 2, February 2017

A Survey on Network Routing Protocols in Internet of

Things (IOT)
Hanumat Prasad Alahari Suresh Babu Yalavarthi, PhD
Department of Computer Science & Prof. in Computer Science,
Engineering,Research Scholar, Department of Computer Science,
AcharyaNagarjuna University, Guntur. India. J.K.C. College, Guntur,
Andhra Pradesh, India.

ABSTRACT path/network and any service. Internet of Things is a new

Internet of Things (IoT) is a paradigm that has gained more revolution of the Internet. Objects make themselves
popularity in recent years. At a conceptual level, IoT refers to recognizable and they obtain intelligence by making or
the interconnectivity among our everyday devices such as enabling context related decisions thanks to the fact that they
personal computers, laptops, tablets, smart phones, PDAs, and can communicate information about themselves and they can
other hand-held embedded devices. These devices now access information that has been aggregated by other things,
communicate smartly to each other. The goal of the Internet or they can be components of complex services. Six years
of Things is to enable things to be connected anytime, ago, for the first time, the number of “things” connected to the
anyplace, with anything and anyone ideally using any Internet surpassed the number of people. Yet we are still at
path/network and any service. Internet of Things is a new the beginning of this technology trend. Experts estimate that,
revolution of the Internet. Objects make themselves as of this year, there will be 25 billion connected devices, and
recognizable and they obtain intelligence by making or by 2020,50 billion. Some estimate that by 2020, 90% of
enabling context related decisions thanks to the fact that they consumer cars will have an Internet connection, up from less
can communicate information about themselves and they can than 15 percent in 2015 Three and one-half billion sensors
access information that has been aggregated by other things, already are in the marketplace, and some experts expect that
or they can be components of complex services. This paper number to increase to trillions within the next decade. All of
surveys some of the standard and non-standard protocols that these connected machines mean much more data will be
are used for network routing in IoT applications. It should be generated: globally, by2018, mobile data traffic will exceed
noted that we have partitioned the network layer in two sub fifteen Exabyte’s – about 15 quintillion bytes – each month.
layers: routing layer which handles the transfer the packets By comparison, according to one estimate, an Exabyte of
from source to destination, and an encapsulation layer that storage could contain 50,000 years’ worth of DVD-quality
forms the packets. Encapsulation mechanisms will be out of video. Further, research in IoT relies on underlying
scope of this paper. Six network layer routing protocols of technologies such as real-time computing, machine learning,
IoT were discussed in this paper. security, privacy, signal processing, big data, and others.
Consequently, the smart vision of the world involves much of
Keywords computer science, computer engineering, and electrical
IoT, PDAs , Routing Layer engineering. Greater interactions among these communities
will speed progress to design the smart devices, Smart
1. INTRODUCTION phones, Smart cars, Smart homes, Smart cities and a smart
The Internet of Things (IoT) is a paradigm that has gained world. This paper surveys some of the standard and non-
more popularity in recent years. At a conceptual level, IoT standard protocols that are used for network routing in IoT
refers to the interconnectivity applications. It should be noted that we have partitioned the
network layer in two sub layers: routing layer which handles
among our everyday devices such as personal computers, the transfer the packets from source to destination, and an
laptops, tablets, smart phones, PDAs, and other hand-held encapsulation layer that forms the packets. Encapsulation
embedded devices[1,2] as shown in figure-1. These devices mechanisms will be out of scope of this paper. Six routing
now communicate smartly to each other. Moreover, protocols in IoT were discussed in this paper .
connected devices equipped with sensors and/or actuators
perceive their surroundings, understand what is going on, and
perform accordingly. These interconnected device networks
can lead to a large number of intelligent and autonomous
applications and services that can bring significant personal,
professional, and economic benefits resulting in the
emergence of more data centric businesses. IoT devices have
to make their data accessible to interested parties, which can
be web services, smart phone, cloud resource, etc. Hence ,
IoT can’t be seen as individual systems, but as a critical,
integrated infrastructure upon which many applications and
services can run. Some applications will be personalized such
as digitizing daily life activities, others will be citywide such
as efficient, delay-free transportation, and others will be
worldwide such as global delivery systems. The goal of the
Internet of Things is to enable things to be connected anytime,
anyplace, with anything and anyone ideally using any Figure-1 :Internet of Things

18
 International Journal of Computer Applications (0975 – 8887)
Volume 160 – No 2, February 2017

2. CHARACTERISTICS OF INTERNET  Retail And Marketing: Usage of the IOT in the

marketing and the retail field is very beneficial. For
OF THINGS(IOT) example knowing the amount of stock present and for
Some of the characteristics[3] of IOT that communicates with billing the amount through the usage of the RFID tags.
different varieties of physical devices. Device heterogeneity
plays a very vital role in the technology of the IOT.  Environment Monitoring: Usage of sensors and RFID
where we can know more about the nature and where we
 Scalability : It must have the scalability as it is having can know the pollution caused to the nature.
many number of devices which are being over the vast
network all the devices must be identified uniquely and  Transportation Industry: Where we can use them in
the and the tags for the devices should be given properly. the trains and buses for the collection of the fares and
then we can avoid the toll gate system through the IOT.
 Information and knowledge management : when we
consider IOT we don’t need to give instructions every  Agriculture And Breeding: Through the IOT we can
time to the machine the device is provided with the know use the high technology into the farming we can use the
ledge and information before it starts functioning and it drone technology in to the agriculture and feeding the
takes decisions and find solutions on its own basing on animals in the cattle.
the knowledge.
4. CHALLENGING ISSUES
 Ubiquitous data exchange: In IOT where the devices With advance of internet technology and development of
are being connected through the internet and where the social network, it is reasonable to expect that a new
information is transferred. We have ubiquitous sensors generation of Internet (also called future Internet) that will
where these are the intelligent sensors gather the appear in the near future. In, several key technical issues[4] of
information and transfer based on the given input. IoT were pointed out. These challenges and open issues
 Optimized energy solution: We must be able to track clarify that the dilemma of current Internet architecture
even a low powered device and the consumer must be requires great efforts to change.
able to get the most optimal outcome.  Security: Security providing might be difficult as the
 Localization and tracking capabilities: Must be able to automation of the devices has been increased which
track the devices and locate them with in a less duration. created new security issues.

 Self –Organization: It is needed to restore the services  Data management: As the communication between the
provided by the devices and to maintain the network devices is being done, every day between the devices lot
connectivity. of data is being generated and there is lot of information
to be transferred from one place to another. Should check
3. APPLICATIONS OF INTERNET OF whether the exact data is being transferred or not. Data
management plays a very important role in IOT.
THINGS (IOT)
The world of IoT includes a wide variety of devices and  Storage management: As there is large amount of data
diverse applications[8], which call for different deployment generated. When the devices are being connected there
scenarios and requirements. Most of applications were used in would be a large amount of multimedia data which is
our daily life being transferred they occupy a large amount of data and
the other kind is random files where the it contains data
 Traffic Management: Where we don’t require a traffic regarding the devices these files doesn’t occupy a huge
police to manage the traffic on roads and all the signal amount of space but they are large in number they must
lights are automated where we don’t require the use of be accessible very quickly whenever necessary.
the manual signals and where we can avoid the
accidents.  Server technologies: as the number of devices over the
network area increases the request and the number of
 Driverless Cars: We don’t require a driver in a car responses of the device also increases at the same time it
where all the and where is car is automatically controlled totally depends on the server where we are running the
by the sensors of the car. interface. Response of the server to the request of the
 Earthquake Detection: All the disasters like device should be done quickly. There should be no delay
earthquake, tsunami etc. can be identified before it’s in the response to the client.
going to happen through the IOT devices.  Insecure authentication/authorization: We generally
 Connected Medicine: Without going to the clinic the provide authentication to provide permission for the user
doctor can monitor the patients. to access the information and authorization is used to
edit or change the data for that particular application and
 Automotive Industry: Advanced Sensors are being permission will be given by the administrator.
equipped to all the cars, trains, buses etc. here we can use
the RFID technology to increase the production and to 5. NETWORK LAYER ROUTING
satisfy the customer by providing more number of PROTOCOLS
services. This section discusses some of the standard and non-standard
 Independent Living: IOT application can be very useful protocols[6] that are used for routing in IoT applications. It
for the aging people for reminding there medicines and should be noted that we have partitioned the network layer in
activities on time and giving them support at the end of two sub layers: routing layer which handles the transfer the
the lives. packets from source to destination, and an encapsulation layer
that forms the packets. Encapsulation mechanisms will be out
of scope of this paper. Six routing protocols in IoT were

19
 International Journal of Computer Applications (0975 – 8887)
Volume 160 – No 2, February 2017

discussed in this section. RPL is the most commonly used Lightweight on-demand ad hoc distance-vector routing
one. It is a distance vector protocol designed by IETF in 2012. protocol-next generation LOADng : The Lightweight on-
CTP is a distance-vector routing algorithm that was developed demand ad hoc distance-vector routing protocol-next
as a solution to routing in WSNs. It stands as a predecessor to generation or LOADng [11] is a lightweight variation of
RPL and was considered the de-facto routing standard for AODV for LLNs. It is designed based on the idea that LLNs
Tiny OS. The Lightweight on-demand ad hoc distance- are idle most of the time. Hence instead of adopting a
vector routing protocol-next generation or LOADng is a proactive approach would generate unnecessary overhead,
lightweight variation of AODV for LLNs. It is designed based LOADng follows a reactive approach in which routes are
on the idea that LLNs are idle most of the time. Hence instead established towards destinations only when there is some data
of adopting a proactive approach would generate unnecessary to send. LOADng is a reactive routing protocol, and found
overhead, CORPL is a non-standard extension of RPL that is suitable for a more general traffic pattern. It does not have any
designed for cognitive networks and utilizes the opportunistic node that performs special functions like the root and is hence
forwarding to forward packets at each hop. On the other hand, not subjected to the subsequent problems that arise due to
CARP and E-CARP is the only distributed hop based routing such a consideration. Also, due to its compressed and flexible
protocol that is designed for IoT sensor network applications. data format, there is no possibility of fragmentation. It does
CARP and E-CARP is used for underwater communication not impose any strict source routing rules, hence it can
mostly. Since it is not standardized and just proposed in accommodate applications which require a fixed MTU.
literature, it is not yet used in other IoT applications. However, LOADng might have a higher delay in the route
discovery phase and might have higher control traffic
RPL Protocol : RPL[7] is a distance-vector and a source overhead if the traffic flows are predominantly P2P.
routing protocol that is designed to operate on top of several
link layer mechanisms including IEEE 802.15.4 PHY and CORPL Routing Protocol: In [12] CORPL will retain the
MAC layers. These link layers could be constrained, Directed Acyclic Graph (DAG) based approach of RPL and at
potentially lossy, or typically utilized in conjunction with the same time introduce novel modifications to allow its
highly constrained host or router devices, such as but not application in Cognitive Radio environments. CORPL uses
limited to, low-power wireless or PLC (Power Line an opportunistic forwarding approach that consists of two key
Communication) technologies. RPL mainly targets collection- steps: selection of a forwarder set i.e., each node in the
based networks, where nodes periodically send measurements network selects multiple next hop neighbors, and a
to a collection point. A key feature of RPL is that it represents coordination scheme to ensure that only the best receiver of
a specific routing solution for low power and lossy networks. each packet forwards it (unique forwarder selection). In
The protocol was designed to be highly adaptive to network CORPL, each node maintains a forwarder set such that the
conditions and to provide alternate routes, whenever default forwarding node (next hop) is opportunistically selected. The
routes are inaccessible. RPL provides a mechanism to DAG construction process in CORPL follows a similar
disseminate information over the dynamically formed procedure as in RPL. After detecting a vacant channel, the
network topology. This mechanism uses Trickle to optimize gateway node transmits a Destination Information Object
the dissemination of control messages (DIO) message. The forwarder set is constructed in such a
way that the forwarding nodes are within the transmission
Collection Tree Protocol (CTP) : In [5] CTP is a distance- range of each other. During the DIO transmission, each node
vector routing algorithm that was developed as a solution to also reports some additional information using the Option
routing in WSNs. It stands as a predecessor to RPL and was field of the DIO message . Each node updates the
considered the de-facto routing standard for Tiny OS. It builds neighborhood information through the DIO message
a tree-based topology with the root at the sink of the network, transmission. Based upon the neighborhood information, each
CTP uses adaptive beaconing mechanism to broadcast routing node dynamically prioritizes its neighbors in order to
control messages. Moreover, CTP relied on a specific link- construct the forwarder list.
layer technology for topology formation, CTP was earlier
known for its efficient energy consumption and high Packet
Reception Ratio (PRR).
Table-1: Study on various routing Protocols on Internet of Things
Routing Protocol Main Results Simulation Used

 RPL showed better PRR and Energy consumption

 RPL showed lesser churn
 RPL showed high PRR Contiki/
RPL
 RPL had higher control-traffic overhead Cooja
 RPL able to cater to variety of traffic patterns,
 RPL is link-layer independent

 In smaller networks, CTP showed better PRR. In larger

networks, Contiki/
CTP
 CTP showed high PRR Cooja
 CTP is only collection-based

20
 International Journal of Computer Applications (0975 – 8887)
Volume 160 – No 2, February 2017

 LOADng caters to more general traffic pattern

 LOADng has flexible and compressible
 packet format Contiki/
LOADng
 No single point of failure in LOADng Cooja
 Longer route discovery phase in LOADng
 More control traffic in LOADng if traffic is predominantly P2P

 In LOAD, control traffic / data traffic

 LOAD routes longer than RPL routes
LOAD NS2
 Higher delay in LOAD due to buffering during route-discovery
 More collisions in LOAD due to flooding

 CORPL make use of Directed Acyclic Graph (DAG) like RPL Contiki/
CORPL
 CORPL uses an opportunistic forwarding approach Cooja

 CARP is a multi-hop delivery of data to the sink for WSN.

CARP  CARP takes care of link quality while selecting the next-hop Real Time Test-bed
node on a route to the sink

 E-CARP is an enhancement upon CARP,

 E-CARP is a location-free and greedy hop-by-hop routing
E-CARP protocol for forwarding packets from sensor nodes to the sink Test-bed
node in an energy efficient manner
 E-CARP does not differentiate the priority of different attributes

CARP Routing Protocol: In[9] Channel-aware Routing drawbacks such as link quality is explicitly taken into account
Protocol (CARP) is a multi-hop delivery of data to the sink for for selecting the next-hop node on a route to the sink.
WSN. CARP obviates to the
Table-2: Comparison of Various protocols supports the Challenges Issues

Storage Data management

Sno Protocol Type Server Technologies Security
Management

.
1. RPL
Yes NO Yes Yes

2 NO NO NO
CTP
2. Yes

. LOADng
3. Yes NO Yes Yes

4 LOAD NO
4. Yes Yes Yes

5 CORPL
5. Yes NO NO Yes

6 CARP NO
6. NO Yes Yes

CARP quickly varying conditions of the underwater channel, have unnecessary control packets to be forwarded in CARP
the fact that two nodes can exchange short control packets when selecting relay nodes for packet forwarding, and these
correctly, may not be sufficient to guarantee that longer data control packets may be avoided in certain situations. Other
packets are also going to be safely delivered ” . Generally, characteristics that make CARP relay selection particularly
CARP is a location-free and greedy hop-by-hop routing suitable for implementing multi-hop routing in UWSNs include
protocol, whose performance is proved better than FBR[],and of the following: (i) The use of simple topology information (hop
its enhanced version EFlood. Link quality is explicitly count) for routing around connectivity holes and shadow zones,
considered when selecting a relay node for packet forwarding. thus avoiding the well-known pitfalls of geographic routing; (ii)
The performance and applicability of CARP have been considering residual energy and buffer space, and (iii) taking
evaluated in the real ocean environment. However, there may advantage of power control, if available, for selecting

21
 International Journal of Computer Applications (0975 – 8887)
Volume 160 – No 2, February 2017

transmission powers so that shorter control packets experience a failure in LOADng, Longer route discovery phase in LOADng,
similar Packet Error Rate (PER) of longer data packets. More control traffic in LOADng if traffic is predominantly P2P.
E-CARP Routing Protocol: In [10] E-CARP, which is an 7. REFERENCES
enhancement upon CARP, to develop a location-free and greedy [1] Gnawali, Omprakash, et al. "Collection tree
hop-by-hop routing protocol for forwarding packets from sensor protocol." Proceedings of the 7th ACM conference on
nodes to the sink node in an energy efficient manner. Generally, embedded networked sensor systems. ACM, 2009.
CARP does not consider the reusability of sensory data collected
previously by domain applications in the following time points, [2] T. Winter, et al, "RPL: IPv6 Routing Protocol for Low-
which induces sensory data packets forwarding which may not Power and Lossy Networks," IETF RFC 6550, Mar. 2012.
be beneficial to certain applications. Therefore, E-CARP allows
[3] IEEE 1905.1-2013, “IEEE Standard for a Convergent
the caching of sensory data at the sink node, for avoiding these
Digital Home Network for Heterogeneous Technologies,"
data packets forwarding in the network. CARP requires to reply
93 pp., April 12 2013.
a PONG control packet whenever receiving a PING control
packet, when selecting the most appropriate relay node for [4] A. Aijaz and A. Aghvami, "Cognitive machine-to-machine
packet forwarding. This PING-PONG strategy may not be communications for internet-of-things: A protocol stack
mandatory when the network topology is relatively steady. This perspective," IEEE Internet of Things Journal, vol. 2, no. 2,
observation drives us to improve the relay node selection pp. 103-112, April 2015.
strategy in CARP, and the relay node adopted previously is
given a higher priority to be reused at this moment. Simulation [5] Al-Fuqaha, M. Guizani, M. Mohammadi, M. Aledhari, and
results validate that our E-CARP can decrease the M. Ayyash, "Internet of things: A survey on enabling
communication cost and increase the network capability to a technologies, protocols and applications," IEEE
large extent, especially when the ratio of packet size between Communications Surveys Tutorials, vol. PP, no. 99, 2015.
control packets and sensory data packets is relatively large. E- [6] J. Granjal, E. Monteiro, and J. Sa Silva, "Security for the
CARP does not differentiate the priority of different attributes. internet of things: A survey of existing protocols and open
In fact, sensory data of attributes of more importance should be research issues," IEEE Communications Surveys Tutorials,
routed to SN with a higher priority. Besides, sensory data of a vol. 17, no. 3, pp. 1294-1312, 2015,
certain sensor node may vary following a spatial and/or temporal
discipline. [7] Salman, Tara. "Internet of Things Protocols and
Standards.", 2015.
6. CONCLUSION [8] Basagni, S., Petrioli, C., Petroccia, R., & Spaccini, D.
This paper surveys some of the standard and non-standard
(2015). CARP: A Channel-aware routing protocol for
protocols that are used for network routing in IoT applications.
underwater acoustic wireless networks. Ad Hoc
Six routing protocols in IoT were studied in this paper. RPL is
Networks, 34, 92-104.
the most commonly used
[9] Aijaz, Adnan, Hongjia Su, and Abdol-Hamid Aghvami.
one. It is a distance vector protocol. CORPL is a non-standard
"CORPL: A routing protocol for cognitive radio enabled
extension of RPL that is designed for cognitive networks and
AMI networks." IEEE Transactions on Smart Grid 6.1
utilizes the opportunistic forwarding to forward packets at each
(2015): 477-485.
hop. On the other hand, CARP is the only distributed hop based
routing protocol that is designed for IoT sensor network [10] Zhou, Z., Yao, B., Xing, R., Shu, L., & Bu, S. (2015). E-
applications. CARP is used for underwater communication CARP: an energy efficient routing protocol for UWSNs in
mostly. Since it is not standardized and just proposed in the internet of underwater things. IEEE Sensors
literature, it is not yet used in other IoT applications. E-CARP is Journal, 16(11), 4072-4082.
an enhancement upon CARP, E-CARP is a location-free and
greedy hop-by-hop routing protocol for forwarding packets from [11] . Verdiere, Axel, et al. "The Lightweight On-demand Ad
sensor nodes to the sink node in an energy efficient manner. E- hoc Distance-vector Routing Protocol-Next Generation
CARP does not differentiate the priority of different attributes. (LOADng)." (2016).
LOADng caters to more general traffic pattern LOADng has
flexible and compressible packet format, No single point of

IJCATM : www.ijcaonline.org
22