See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/321866375

New Networking Era: Software Defined Networking

Article · November 2013

CITATIONS READS

9 3,513

3 authors:

Furqan Alam Iyad Katib

King Abdulaziz University King Abdulaziz University
14 PUBLICATIONS 1,076 CITATIONS 118 PUBLICATIONS 2,979 CITATIONS

SEE PROFILE SEE PROFILE

Ahmed Alzahrani
King Abdulaziz University
40 PUBLICATIONS 760 CITATIONS

SEE PROFILE

All content following this page was uploaded by Furqan Alam on 17 December 2017.

The user has requested enhancement of the downloaded file.

 Volume 3, Issue 11, November 2013 ISSN: 2277 128X
International Journal of Advanced Research in
Computer Science and Software Engineering
Research Paper
Available online at: www.ijarcsse.com
New Networking Era: Software Defined Networking
Furqan Alam, Iyad Katib, Ahmed Saeed Alzahrani
Computer Science Department, Faculty of Computing & IT
King Abdulaziz University Jeddah, Saudi Arabia

Abstract: The fundamental problem with traditional network architecture is that it is static. Users don’t have much
controlling power and it is just trying to accommodate current networking needs of the user. Therefore which means
window for innovation is too narrow with respect to network controlling, virtualization, automation, scalability,
programmability and fault tolerance. A revolutionize network technology which can put all the above aspects into
practice in future networks is Software Defined Networking. In the study we review a brief history of Software
Defined Networking that can help to provide evidential clarity with respect to the understanding of the new technology
and more significantly, in understanding the motive that why we need Software Defined Networking.

Keywords: Software Defined Networking, SDN, SDN models, Control Plane, Data Plane, OpenFlow.

I. Introduction
Networking a field of vast interests and studies. Yet not mature enough due to no hardcore principles like other computer
science domains. Today networks and internet are far more developed than ever before but we can clearly see all new
dimensions for future networking and internet, yet not functional enough on large scale. Current or in-practice network
architectures don’t match to the actual demands of today’s network users and businesses. Closeness to “abstraction”
principle is becoming popular idea behind why users need a new network architecture which revolutionize this field and
concentrate on abstraction rather complexity. For instance, today networks are huge and they support many services from
simple text messages to various multi-media services in parallel to several network users opting for same service at a
time. This creates traffic within a traffic and add more complexity to network architecture hence things look messy in the
end and hard to handle at some point of time. The present network architecture is consisting of various hardware like
firewalls, routers and switches and we can see all of them together as large complex system and thus a minute change at
any level can cause network failure. This gives researchers a motive to move to such an architecture which has centralize
control, dynamic, easy to manage, and highly scalable. The answer to the above is Software Defined-Networking which
has all essences to revolutionize and to take networking to completely different environment.
A. Contributions
The following are the contributions of this literature survey:
 The study answers why there is a need to move to a different networking architecture from the traditional one.
 The study tried to establish the facts why SDN is far better than the current network architecture. This is
achieved by studying what exactly SDN offers us.

II. Traditional Networks

Data flow is controlled by routers and switches in traditional networks. Current network architecture is static if compared
to SDN. Users don't have any control over packets. Generally this job is done my switches whereas in SDN, user can
shape his own traffic because of the centralize control mechanism. In the study [16], [17] comparison of traditional
networks and SDN are done. Furthermore in this new design user can change and introduce his own traffic policies over
the network which is lacking in traditional approach. This is accomplished with help of SDN provided software interface.
This is cost effective too as now we are not much dependent on costly switches and their vendors [4], [6]. Traditional
networks are being over used to their limit. The design principles behind them is not matching today’s astronomical data
growth, bandwidth, speed, security and scalability issues. In general networking devices consist of control plane. The
control plane is an important part of a router. The duty of it is to concentrate on how that one individual box
communicates with neighbors and what the state exchange. Data plane whose task is to consult the forwarding table. The
network device uses forwarding table to decide where they have to send the packets. Both resides on networking
hardware devices [5]. User can’t put his innovation in practice with traditional networks due to the design limitations of it
but this is possible by software Defined networking. A traditional network architecture consists of a data plane which
carries data physically in form of packets from one node to the other by following certain protocols, control plane which
consist of logic that devices are using therefore to forward packets over the network then comes Management plane
which behave as an administrator [14]. The basic fundamental protocols behind the design of internet is still a success. If
we see carefully the initial architecture of the network is same as before. But the new application points out new
requirements which the network is not able to satisfy in an efficient manner, it is just accommodating it currently.
Therefore reviewing the network architecture is a need of hour [9].

© 2013, IJARCSSE All Rights Reserved Page | 349

 Furqan et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(11),
November - 2013, pp. 349-353

Figure 1. Traditional Switch Architecture. Source: IBM Systems and Technology Thought Leadership White Paper

III. Software Defined Networking

What SDN is all about? The basic idea of SDN is the separation of control plane and packet forwarding plane [1], [2]. In
SDN control plane is programmable hence giving rise to innovation [13]. Software Defined Networking (SDN) assure to
strikingly minimize the current complex networking environment and its management. Innovation feature is making the
networking environment more attractive to work on. User can differentiate between the service providers because of
various dimensions which are not there in traditional networks. This will help drastically to maximize the network ability
to deliver to the best and more efficiently. SDN is seen as area of attraction for researchers and hardware manufacturers
in recent times. For example Ericsson, Net-gear, Nokia Siemens Networks and Verizon are in the list of current members
of Open Networking Foundation (ONF). Along with what discussed above, various opportunities which should be
trapped with help of implementing the SDN to wireless infrastructure-less networks are stated in the study [3]. SDN is
based on abstraction. Separation between control plane and forwarding plane is done. An interactive application
environment is provided with control plane. Network topology information plays critical role with the help of which
decision can be changed, modified and designed in completely new manner with help of SDN application interface [7].
Control plane can also be termed as the operating system of the network. Refer Figure.2 for more detailed SDN design
concept.
What are the architectural Models available for SDN? A very critical task is to select a SDN model or models SDN can
support. Basically there are three types of models which are network virtualization model, evolutionary model and the
OpenFlow model of SDN. The most important goal of network virtualization is to get rid of restrictions on LAN
partitioning that resides in Ethernet Virtual LAN standards and solve the issues like scalability & multicasting in the
network architectures. One highly popular advantage of network virtualization is that it is affirmative towards multi-
tenant clouds without making any change in the network but the most significant negative of virtualization is that it adds
complexity and extra overheads. Evolutionary model objective is to maximize software control within the boundaries of
network topologies. Integration of standards in the devices by vendors is a major problem in this model. Some do it some
won’t there compatibility issue arises. Last one is OpenFlow is the model, on which the basic idea on SDN is built [8].
The paper [20] concentrated on three problems in networks. Firstly, there must be a mechanism for frequent changes to a
network. These change will be both in conditions and network state. Secondly, there must be a high level language for
network configuration support. Lastly, there must be better visibility and control on network troubleshooting tasks as
well as on its diagnosis. For the solution a wide range of high level language network policies and correct scan of
network problem. Later study explained how SDN can help is efficient network management.

Figure 2. SDN Layered Architecture. Source: ONF White Paper

© 2013, IJARCSSE All Rights Reserved Page | 350
 Furqan et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(11),
November - 2013, pp. 349-353
One of the major output in giving future Internet design is Software Defined Networking (SDN) [1], [10]. With the help
of SDN user can design and defined network protocols that can be tested on production systems. OpenFlow, with the
objective of implementing innovation in computer science networking field is the major advances [11]. What OpenFlow
does is to make differentiation by creating separation among the control plane and data plane. SDN represents a great
opportunity to the scientist and engineers to rethink networks with the view point of separating the software that controls
the network from the individual network entities. OpenFlow concentrated on Ethernet technologies but its functional
ability with wireless and optical networks is still in question yet works are going on in this area because it will open a
wide environment for SDN [9]. In study [12] authors stated that one of the motive of SDN based architecture is to give
programming capability whereas in ICN gaining popularity among the computer science experts who wants to change
present IP based layer with a complete new idea. With emergence of Information Centric Networking (ICN) and SDN as
one of topics of research communities also attracting corporate businesses and investments to gain from it. It’s too
important how ICN and SDN will combined in parallel, implemented and tested.

IV. Cutting Edge Feature of Software Defined Networking

Software Defined Networking (SDN) provides very interesting features which will revolutionize the future networks like
centralize control mechanism [33], cost efficiency, innovation, programmability, scalability [21],[26], security,
virtualization, cloud support, automation, reliability [26] and efficient environment to support Big-Data. In paper [15]
what SDN offers us is stated. It gives us a centralized control mechanism of various networking hardware devices from
multiple vendors, further network automation is greatly improved with the help of APIs which are used to abstract
network details. One of the main objective of SDN is to provide innovation over the internet or the network, this ensure
very less vendor dependence. Programmability by operators is feature of Software Defined Networking which helps to
put innovation into practice. Security and reliability are the major issues in today’s networks, SDN enhance these two
aspects to the maximum. Simple operation and cost efficiency are key factors with openness to design and invent in
SDN. Yet this openness with respect to APIs not yet fully explored but in study [13] this aspect has been studied. SDN
makes it easy to share resources in an efficient manner with the feature of network intelligence. Further paper stated SDN
is helpful in service aware networking too. With SDN, Network Management has improved vastly which also give it
cutting edge on traditional networking architecture [27] and important determinant in future works.
Networks based on Software Defined Networking are being implemented both on testbed and production networks. Fault
tolerance property is a key for the production networks and a most desirable & a must for SDN networks. It is noted that
with respect to fault tolerance in SDN there are not much researches. In the paper [18] CORONET, a SDN fault-tolerant
system is discussed. A fault tolerance SDN architecture is stated which quickly recover from the occurred faults and can
work on highly scalable networks. The architecture describes the recovery from multiple links failures. SDN based
networks are more and more been deployed both on testbed and production networking environments. For example,
Google exhibit its deployment of OpenFlow and the most famous SDN protocol, On April 2012 at the OpenFlow
Networking summit [19].

Figure 3. CORONET Architecture. Source:[18]

V. SDN Concept to Enhance Big-Data management and Optical Networks

Databases are now real-time and distributed with parallelism, they support thousands of queries at a time. The data
volume in various engineering and science disciplines is assumed to be increase by a factor of 10 or even more in coming
years [22], [23]. Companies are moving from SQL databases to No-SQL databases. We have seen amazing data
production over internet and by many other science projects like Large Hadron Collider. All this has given birth to Big-
data concept which means processing of large variety of data in huge volume with high velocity. Current networks are
lacking in providing efficient support for this purpose but SDN is seen as an answer to this problem. With the
architectural idea of Software-Defined networking a Research Sandbox demonstrator has been implemented in paper [24]
where virtual switch network abstraction is studied with combination to SDN. This is taken as an example of adaptable

© 2013, IJARCSSE All Rights Reserved Page | 351

 Furqan et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(11),
November - 2013, pp. 349-353
networks according to demand and requirements. Internet is becoming important part of our everyday life. There is a
necessity of providing disaster resilient wide area networks which is reliable [26]. SDN or OpenFlow is maturing
technologies in this respect which tends to solve several problems of operation management in WANs. The study [25]
represent a disaster resilient WAN system. SDN based networks are also getting popular among wireless and optical
networks and hence shown bright prospects for future work in this area. Enabling high bandwidth, flexibility, efficient
resource utilization, error control, data security and efficient traffic management requirements of wireless and optical
networks are achieved with SDN based wireless and optical networks. Complex issues like traffic grooming can be
efficiently handle with SDN. Further it will be a cost efficient and highly reliable solution for these networks [28], [29],
[30], [31] and [32].
VI. Conclusion
The traditional networking architecture can’t accommodate the advance user requirements efficiently. Increase of mobile
devices, virtualization, high automation & security, efficient Big-Data management and high quality with variety of
services, SDN is a promising architecture. For sure current network is not dynamic if compared to SDN. Though yet
SDN, needs some time to mature and industry also need time to synchronize the devices according to it. SDN can also
manage optical and wireless networks fruitfully. To understand why SDN will play a critical role in future in shaping
various technologies, we have to think what actually SDN providing us apart from all technical advances? SDN gives
user or operator a feel of nearness to the network. He won’t feel that he is a distinct entity when operating on SDN based
networks. Hence it is true to say that future networks will revolve around SDN based networking.

References
[1] K. Greene. 10 Emerging Technologies, TR10: Software Defined Networking, 2009. Available at:
http://www.technologyreview.com/biotech/22120/.
[2] B. Lantz, B. Heller, and N. McKeown. A network in a laptop: rapid prototyping for software-Defined networks. In
Proceedings of the Ninth ACM SIGCOMM Workshop on Hot Topics in Networks, Hotnets ’10, pages 19:1–19:6,
New York, NY, USA, 2010. ACM.
[3] S. Costanzo, L. Galluccio, G. Morabito, S. Palazzo. "Software Defined Wireless Networks:Unbridling SDNs".
European Workshop on Software Defined Networking. 2012.
[4] http://www.zdnet.com/10-key-questions-about-software-Definedd-networking-sdn-7000015822/
[5] http://globalconfig.net/software-Definedd-networking-vs-traditional/
[6] http://readwrite.com/2013/04/23/software-Definedd-networking-dn#awesm=~omPg0fn3rysfHX
[7] Vijay K. Gurbani, M Scharf, T.V. Lakshman and V. Hilt. Bell Laboratories, Alcatel-Lucent"Abstracting network
state in Software Defined Networks (SDN) for rendezvous services"." Communications (ICC), 2012 IEEE
International Conference on Software Defined Networks. (2012).
[8] http://searchcloudprovider.techtarget.com/tip/Three-models-of-SDN-explained
[9] Fl´avio de Oliveira Silva, Jo˜ao Henrique de Souza Pereira, Pedro Frosi Rosa†, and Sergio Takeo Kofuji. "Enabling
Future Internet Architecture Research and Experimentation by Using Software Defined Networking". European
Workshop on Software Defined Networking. 2012.
[10] A. Greenberg, G. Hjalmtysson, D. A. Maltz, A. Myers, J. Rexford, G. Xie, H. Yan, J. Zhan, and H. Zhang, “A clean
slate 4D approach to network control and management,” SIGCOMM Comput. Commun. Rev., vol. 35, no. 5, p. 41–
54, Oct. 2005. [Online]. Available: http://doi.acm.org/10.1145/ 1096536.1096541
[11] N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson, J. Rexford, S. Shenker, and J. Turner,
“OpenFlow: enabling innovation in campus networks,” SIGCOMM Comput. Commun. Rev., vol. 38, no. 2, p. 69–
74, Mar. 2008, ACM ID: 1355746.
[12] Dimitris Syrivelis, George Parisis, Dirk Trossen, Paris Flegkas, Vasilis Sourlas, Thanasis Korakis and Leandros
Tassiulas. "Pursuing a Software Defined Information-Centric Network". European Workshop on Software Defined
Networking. 2012.
[13] Myung-Ki Shin, Ki-Hyuk Nam; Hyoung-Jun Kim. "Software-Defined networking (SDN): A reference architecture
and open APIs". ICTC. 2012.
[14] IBM. "IBM Systems and Technology Thought Leadership White Paper". 2012
[15] Open Networking Foundation. "Software-Defined Networking: The New Norm for Networks". ONF White Paper.
April 13, 2012.
[16] OpenFlow switch specification Version 1.3. Open Networking Foundation. Avaible
at:Http://www.opennetworking.org/. 2012.
[17] Open Networking Foundation. “OpenFlow /Software Defined-networking (SDN)”. http://www.
www.opennetworking.org/.
[18] Hyojoon Kim, Jose Renato Santos, Yoshio Turner, Mike Schlansker, Jean Tourrilhes, Nick Feamster.
"CORONET: Fault Tolerance for Software Defined Networks". IEEE. 2012.
[19] OpenFlow Networking Summit. http://opennetsummit.org/, Apr. 2012.
[20] Hyojoon Kim and Nick Feamster, Georgia Institute of Technology. "Improving Network Management with
Software Defined Networking". IEEE Communications Magazine. Feb, 2013.
[21] Soheil Hassas Yeganeh, Amin Tootoonchian, and Yashar Ganjali. "On Scalability of Software-Defined
Networking". IEEE Communications Magazine. February, 2013.

© 2013, IJARCSSE All Rights Reserved Page | 352

 Furqan et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(11),
November - 2013, pp. 349-353
[22] Eli Dart, Brian Tierney, editors, “HEP (High Energy Physics) Network Requirements Workshop, August 2009 -
Final Report”, ESnet Network Requirements Workshop, August 27, 2009, LBNL LBNL-3397E.
[23] Eli Dart, Brian Tierney, editors, “BER (Biological and Environmental Research) Network Requirements
Workshop, April 2010 - Final Report”, ESnet Network Requirements Workshop, April 29, 2010.
[24] Inder Monga, Eric Pouyoul, Chin Guok. "Software-Defined networking for big-data science: Architectural models
from campus to the WAN". High Performance Computing, Networking, Storage and Analysis (SCC). 2012.
[25] Kien Nguyen, Quang Tran Minh, and Shigeki Yamada. "A Software-Defined Networking approach for Disaster-
Resilient WANs". IEEE. 2013.
[26] Xinjie Guan, Baek-Young Choi and Sejun Song. "Reliability and Scalability Issues in Software Defined Network
Frameworks".Second GENI Research and Educational Experiment Workshop. 2013.
[27] Hyojoon Kim and Nick Feamster. "Improving Network Management with Software Defined Networking". IEEE
Communications Magazine. February, 2013.
[28] J.-P. Elbers, A. Autenrieth. "From Static to Software-Definedd Optical Networks".
[29] Salvatore Costanzo, Laura Galluccio, Giacomo Morabito, Sergio Palazzo. "Software Defined Wireless Networks:
Unbridling SDNs". European Workshop on Software Definedd Networking. 2012.
[30] Brandon Collings. "New Devices Enabling Software-Defined Optical Networks". IEEE Communications
Magazine. March. 2013.
[31] N. Amaya, S. Yan, M. Channegowda, B. R. Rofoee, Y. Shu, M. Rashidi, Y. Ou, G. Zervas, R. Nejabati, D.
Simeonidou, B.J. Puttnam, W. Klaus, J. Sakaguchi, T. Miyazawa, Y. Awaji, H. Harai, N. Wada."First
Demonstration of Software Defined Networking (SDN) over Space Division Multiplexing (SDM) Optical
Networks".
[32] Channegowda et al. "Software-Defined Optical Networks Technology and Infrastructure: Enabling Software-
Defined Optical Network Operations". OPT. COMMUN. NETW./VOL. 5, NO. 10/OCTOBER 2013.
[33] Open Networking Foundation. Software-Defined Networking: The New Norm for Networks. 2013.

© 2013, IJARCSSE All Rights Reserved Page | 353

View publication stats