Cross Layer Design in Software Defined Networking (SDN): Integration of two technologies

Abstract

The current era of the internet has high demands considering the need of high bandwidth and faster internet speeds. In this challenging environment where the Datacenters are running out of space and getting cramped up with new networks every day the onus of creating more efficient networks and providing the aforementioned aspects without sacrificing the quality of service and security of the networks is of our key interest. In today's date this is possible due to technologies such as Software defined networking where in the physical networks are virtualized to make the networks more scalable and less restricted. Collaborating with such a technology is another approach of Cross layer design which will bring in the flexibility in the network architecture. The information exchange between the layers can be very useful making the overall network more secure, scalable and efficient. This paper throws light on both the technologies and their fusion which brings better and stronger networks. Also this paper will discuss the shortcomings and methods to curb them. Index Terms-Cross layer design, SDN ,

Full text

Cross Layer Design in Software Defined Networking
(SDN): Integration of two technologies
Neha Jadhav (300037254)
Department of Electrical and Computer
Engineering
University Of Ottawa
ON, Canada, K1N6N5
Email: njadh059@uottawa.ca
Abstract – The current era of the internet has high
demands considering the need of high bandwidth and faster
internet speeds. In this challenging environment where the
Datacenters are running out of space and getting cramped up
with new networks every day the onus of creating more efficient
networks and providing the aforementioned aspects without
sacrificing the quality of service and security of the networks is of
our key interest. In today’s date this is possible due to
technologies such as Software defined networking where in the
physical networks are virtualized to make the networks more
scalable and less restricted. Collaborating with such a technology
is another approach of Cross layer design which will bring in the
flexibility in the network architecture. The information exchange
between the layers can be very useful making the overall network
more secure, scalable and efficient. This paper throws light on
both the technologies and their fusion which brings better and
stronger networks. Also this paper will discuss the shortcomings
and methods to curb them.
Index Terms –Cross layer design, SDN ,
I. INTRODUCTION
Technological advancements are on the rise exponentially
and this obviously calls for the increased use of hardware and
additional networks for the processing of the data. These
networks which are deployed newly can take a lot of time to
deploy the architecture completely which will incur extra cost
and efforts. Here is when the rising technology of Software
Defined Networking (SDN) comes in to picture which gives
the flexibility to virtualize the network by eliminating the
hardware part to certain extent. This is possible due the
decoupling of the data plane and control plane.
The modern architecture of the networks can have mixture of
multiple technologies just to make the network architecture
more reliable, resilient and more organized. In the SDN
architecture the controller is logically centralized given us the
pliability to program and optimize the entire network. SDN
gives the developer the ownership to program the entire
network. This also means that the security of the whole
network can be mended and monitored centrally with help of
any SDN controller. [1] The current network architecture is
coupled vertically which makes it even more difficult to handle
but SDN architecture. This vertical paradigm of the traditional
networks is broken by the SDN technology. Abstraction of the
components also helps in simplifying the overall network
making it more organized and evolving. [13]
Additionally to the software defined networking technology
which deployed in the modern networks the cross layer design
brings in its own positive attributes to the whole networks
aspect. As the name suggests cross-layer architecture does not
follow the strict layered approach but it communicates through
the layers. In traditional networks following strict OSI layered
architecture it impossible to communicate between the layers
making the structure restrictive and monotonous. [44]
The performance of the network is enhanced to a greater level.
The current level performing any task will take inputs from the
layers below and above additionally the performance of the
TCP/IP will also be improved when the information is
exchanged by the layers the redundancy is reduced between
the layers. [4] All these processes are inter-related the
amalgamation of the two technologies can be useful for not
one but many kind of communication technologies such as
Wireless networks, 5G telecom, Inter net of things etc. So this
mixture and deployment of these two technologies can come
up with really strong networks. The current scenarios with the
traditional networks are very restrictive. Bringing in Cross
layer design in SDN will give us the privilege to treat the
networks according to our needs. [29]
This paper will highlight both technologies separately and
then the positive attributes of both the technologies together.
Also the shortcomings faced by SDN due to decoupling of the
planes can be solved with the help of cross layer approach.
II. INTRODUCTION TO CROSSLAYER DESIGN
A.Overview of Cross layer design
The cross layer architecture allows direct communication
between the non-adjacent layers unlike the traditional OSI
model. The basic functioning of cross layer is such that there
should be communication between all the layers and one layer

can be accessed by other layers. The need of higher data rate
for telecom purposes, video streaming a, video calling
entertainment etc. Faster throughput is possible because of the
cross layer design.[14] The complexity in the cross layer
architecture is that it is difficult to identify the interactions
between the layers. The issue of network congestion can also
be solved when the interrelation between the layers. The Basic
performance of cross layer design can be referred in two
ways[13].
 Tightly coupled cross layer design: Here the
optimization is done as the collection of all the layers
and the optimization problem is treated as one.
 Loosely coupled cross layer design: The loosely
coupled design highlights the optimization of one
single layer at a time.
B. Basics of Cross layer design
The flexibility of this technology makes it possible for varied
use in various technologies such as UMTS/WLAN, 5g
networks, cognitive radios and various other wirless
technologies.[2][3][5]The basics of the cross layered design is
the information flow that the layers accept. The proposed four
designs are these designs were taken in consideration the
TCP/IP model. [6][7]
1. Upward information flow: This method proposes that
that the information flow is from downwards. The
lower layer transfers information to the upper layer in
the stack.[14]
Fig.1 The upward information flow in cross layer design[14]
2. Down ward information flow: In very rare cases this
take place when the information streamed to the
above is flown down. This happens when the
application layer wants the lower levels to know
about how the data is treated. This flow might be
restricted and the interface of such flows can be set to
a lower layer parameter. These information flows are
usually the caution notification from the upper layers
to the lower levels. This action can be restrictive as
this is one to one communication between just two
layers. And other layers can’t impart the information
to the lower level.
Fig 2. Downward information flow in cross layer design[14]
3. Back and forth information flows: In this way the
layers responsible for performing tasks differently can
be mixed together. And they can communicate
simultaneously in stack protocol. The information
loop is shown in the figure below.
Fig.3 The back and forth information flow[14]
4. Merging of adjacent layers: Two or more adjacent
can be merged together and can be treated as one
super layer. This particular approach won’t need any
specific integration method. Also this is done for a
very specific reason where the application needs to be

handled in a certain way and it needs to integrate two
layers.
As the fundamentals of any networking starts with the OSI
layer structure and the performance of any network is tested
with the help of TCP/IP model the layered structure can be
improvised using cross layer design it gives the ease of
communication in various scenarios.[14][16]
Fig 4 system design framework for cross layer design [15][20]
The overall framework for the cross layer design is
something as described above. This basically tells that
complex systems can be broken down in straightforward
systems and can be easily interpreted. The individual layers
can be improvised and updated as we need it. As described
earlier loosely coupled and tightly coupled systems can be
developed. This eventually reduces the cost of entire network.
Few of the shortcomings of this design architecture is that
the flow of the information can vary depending on the layered
boundaries. As described previously the layered architecture
has its inefficiencies and redundancies when two or more
layers perform same job. This can result in tradeoff between
quality of service and performance. When optimizing the
layers jointly this architecture might result in complicated
algorithms. The modularity between the networks can be
hampered.
III. INTRODUCTION TO SOFTWARE DEFINED
NETWORKING (SDN)
Software defined networking has the architecture of traditional
networks fundamentally but due to SDN it becomes possible to
decouple the control and data plane. This is the main reason it
brings in the flexibility in the overall network. The entire
network is divided between network plane and data plane and
the communication between them takes place with the help of
Southbound API and Northbound API. The detailed
architecture of SDN is shown in Figure5. To have a better
understanding of the SDN architecture the figure explains the
division between the planes and layers.[1][47]
Fig. 5 Software defined architecture in planes and layers
The primary components of SDN architecture are [1]
 Forwarding Devices: These are the devices which are
software or hardware based performing initial actions
to forward the flows in a particular direction. The
incoming information should be forwarded to the
desired location. The instructions are specified by the
south bound API which is mainly OpenFlow.[1]
 Data plane: All the physical devices are placed in this
layer and then later these are abstracted to the control
plane. The network infrastructure has interconnected
devices and comprises of the data plane.[1]
 Southbound API: The communication which takes
place between the data plane and the control plane is
with the help of Southbound API’s. The instructions
given to the forwarding devices are decide by the
southbound API.[1]
 Control plane: This is one of the crucial areas in any
SDN architecture. Control plane is where the real
action takes place. The logically abled controllers are
set in this plane. This is considered as the “Network
Brain”. This is the most important part where all the
logic is rested and forwarded to the entire network.
The scalability of the network is dependent on how
strong the controller is [51]. The survivability of the
network is based on the fact the controller should be
able to take the network attacks and withstand it. The

controllers are made string enough that it can take
care of the issues in the whole network [1]
 Northbound API: The main function of a
northbound API is to present the developer with an
interface to program the applications. It helps in
abstracting the underlying instructions by southbound
API.[1]
 Management plane: The management plane looks
after the network functions needed for the network to
go on. These functions include functions such as
firewall applications; load balancing, routing and
monitoring of the data in the network.[1][11]
The network hypervisors are responsible to bring in the
flexibility in the abstraction of the physical components to
create virtual machines. This is very important in the
decoupled networks. This helps in virtualizing switches,
routers, firewalls and also servers.[1][47]
Fig.6 System design architecture.[1]
IV. CROSS LAYER DESIGN IN SOFTWARE DEFINED
NETWORKING (SDN)
As we are aware that cross layer design and Software defined
networking are two of the latest and most famous technologies
which are being employed in today’s world to create modern
networks. We shall see the progress which is being propelled
until now where these two technologies are merged for the
betterment of the technology. There are few applications
which have employed these technologies jointly for excelled
results. [49][25]
A. Cross layer service design to network orchestration.
The SDN mechanisms are widely used and are very
helpful for various applications. [50] These applications such
as cloud computing. Functions such as network virtualization
and abstractions give the developer the privilege to make the
network more scalable and bring in new aspects to the same
network according to the changing needs of the services.[51]
But this dynamic behavior of the various connections getting
connected to one prime network can introduce complexities.
Network traffic may increase.[19] So the increasing demands
of the networks should be provisioned in real time. And for
this we need to have a more strong and robust approach
towards the system. The main highlight of this particular topic
is curbing to network congestion issues caused by the dynamic
nature of the network. [19][23][16]
The current architecture of the cloud environments which
incorporates software defined networking. This has
connectivity throughout the different cloud infrastructures for
better communication. In the current scenario the real time
applications the network layer using SDN is shown in a certain
way. [19]The dotted lines are the protocol which is needed for
the transmission of the data. These are used to exchange the
information and then the videos and audios calls are set up.
The connection is set up and then the media is interchanged
which is shown by solid line.[19][41]
Fig.7 Present Application and network architecture.[19]
In this current approach the network elements won’t be able
to notice the requirements of the real time and it will need to
come up with modern algorithm. Therefore to solve this issue
the Cross layer Orchestrator is introduced which will form an
interaction between the application and the network layer
which will give the functionalities to keep up with the network
requirements. [25] The traffic flows will be improved after this

as there will be better communication between control and
data plane. [19][23] [41]
Fig.8 Control and network layer interface [19]
i. Functioning of Cross-layer orchestration architecture.
From the security and reliability perspective the data
centers and the information on datacenters are distributed over
cloud networks and then they are connected by a wide area
network [48]. So these cloud infrastructures are based on their
requirements and then service provision according to nature of
the services. All the components are SDN based; the virtual
networks in different datacenters are SDN enabled which
makes the whole setup more scalable.[19][23][45]
Fig.9 Functional Architecture of CLO (Cross layer
Orchestrator) [19]
The main architecture is of the Cross Layer orchestration
architecture is divided in three subsections concerning the
Application Layer, Control layer and Infrastructure layer as
they are the main building blocks of the CLO.[19][26]
 Application Layer: The integration of the application
layer and control layer. The presence of CLO will
make the. The services required will have to expose
themselves to the CLO which will detect the
resource requirement and allocate the resources
from the Virtual machines and other virtual
components. The CLO-app will be the reference
point.[19][24]
 Control layer: This is the most crucial part of the
CLO the main problem with traffic congestion and
issues related to network flow are taken care of in
this section.[19] This particular layer needs to have
the view of the overall network which will then help
in traffic steering in all the cloud infrastructures.
This layer contains three prime components of
Virtual infrastructure manager, open flow controller
and cross layer orchestrator. [19][18][16]
The CLO is capable of the having the overview if
the service topologies and the infrastructure. The
openVirtual infrastructure manager component
CLO-vim provides Infrastructure as service which
makes computation easy. The reference points helps
in calculating the traffic path. This will regulate the
traffic flow increasing the efficiency.[19]
 Infrastructure layer: Hardware components are
resided in this layer which helps in establishing
connectivity between different hosted services of
different data centers which are connected by WAN.
The only requirement here is to transfer the flows
from the northbound API to the CLO with the help
OpenFlow management and configuration
protocol.[19][46]
The general mechanism to restrain the traffic problems is the
priority queueing which are represented with the help of
configuration rules. The packets which enter the network are
matched to the CLO defined classifiers [19] and then it’s
pushed ahead to the destination according to the CLO defined
forwarding protocols.[43] The acceptance of these Cross layer
orchestrator the implementation is done using a real-time
multi-service scenarios. The services used here are IMS 3GPP
infrastructure, RTC service with a multi-party call control
service which is scalable in nature also web RTC is included
[26][19].
To know the worth of the network different scenarios are
conducted such as 3-party video call, peer-to-peer video call
and peer-to-peer call with medium quality but with guaranteed

quality of service.[27]the multimedia traffic generated can be
relevantly transferred[9] With the mentioned cross layer
architecture it is then possible for us to instantiate on-demand
services with the help of SDN and NFV rules[24]. This is a
SDN cross layer design architecture with application. This will
increase the performance for service oriented services. This is
possible because there is a proper integration between the
SDN and cross layer design architecture which gives the
flexibility to program the services like RTC.[19] [16]
B. Software defined radios based on Cross layer architecture
with SDN principles.
The second example we shall dis is of Software defined
radios where the physical communication system components
of modulators, demodulators, mixers, amplifiers detectors etc
which are mainly used in signal processing units are executed
using Software.[22] This is done so that the radio can make
use of the physical layer protocols specially. The radio
architecture needs cumbersome hardware architecture. To
eliminate this Software is introduced. This automatically
increases the flexibility to program and in real time
situations.[22]
The dynamic nature of the modern networks wants the SDR
to configure rapidly this is possible using Software defined
networking to have an overall idea of the network. As
integration of Software defined Radios and Software defined
networking is crucial the key highlight of this topic is to come
up with the cross-layer architecture. The important aspect of
the architecture is a platform called Cross Flow. The
applications included in this platform are [22]
[29]
 Adaptive Routing: A global controller is used to
have a view of the entire network. This is used to
permit equal crossflow throughout the controller.
Depending on the nature of the network and
application requirements adaptive routing is used
toggle between routing protocols.[22][30][11]
 Cross-layer control: for the joint optimization of
different protocols, algorithms at different
layers.[22]
 Quality of Service (QoS): Admission control,
scheduling are the policies which are needed in
provisioning of the data. ARQ, FEC are few of the
examples of it [22].
 Physical Layer Adaptation: For proper resource
provisioning and manage switching scenarios
Frequency hopping, adaptive modulation and
frequency hopping is used for better spectrum
utilization.[22][12]
To test the crossflow platform a software platform of GNU
radio is used which has signal processing environment. This
architecture is basically used to provide a platform to manage
high configurable SDR’s and to have effective integration by
using SDN.
i.. Cross flow architecture and design
The AEtherflow [32] process as discussed before talks about a
technique where wireless networks and systems are mixed with
SDN processes in a systematic way. It is done to introduce
ability of programming the network like the networks which
are wired. The concept model of SDN for AEtherflow is
derived from the [33]. Every single component of SDN in
AEtherflow has interfaces of 4 types such as events, statistics,
configurations and capabilities.[22]
TinyNBI model is used for categorization of ports of
radio. First thing it goes is for Configuration. By using this
interface, properties of USRP device can be configured by the
controller [22]. The properties are sample rate, frequency,
channel and gain. The other interface used is Capabilities. By
using this interface, the controller gets authority to perform a
query function and can get information about the capabilities
of the radio. [33]
The capabilities of radio can be identified in terms of
numbers of channel, id of device, which sensors are supported,
and bandwidth supported. One more Interfaces used is
Statistics. It allows the controller to obtain the statistics of the
board such as Received Signal Strength Indication and also
temperature obtained by sensor.[8] One More Interface called
as Events is used which gives the advantage of making a
decision based on occurrence of events. The events can be for
example related to low Received Signal Strength Indication or
high temperature sensed by sensor. B[22]. Extensions of
Control Plane and Messages. OpenFlow methodology is used
in the design which we are implementing for SDN
methodology. Openflow is mainly concentrated into wired
devices even if it’s a robust model.[31] The OpenFlow
protocol needs some changes and extensions so that it is able
to support radio interfaces. Therefore, for the implementation
of CrossFlow, messages will be carried in OpenFlow protocol.
These messages are then utilized by application for controlling
the processing of blocks and for carrying out the
implementation of complex functionality.[22] [46]
Furthermore research is done in field of definition of
message, structure of message. The CrossFlow model is shown
in Figure 1 which is used to comprehend expectations of our
goal to control the heterogeneous network [22].
The hardware platform for our model is USRP 310 which is
from Etus Research. E310 makes use of Zynq 7020 which is
employed with dual core ARM Cortex A9 processor and
FPGA on same chip. In the device, both then systems are

loaded for more processing power. Namely GNU radio blocks
and the CPqD SoftSwitch software. [28]
OpenFlow agent mechanism will be carried out by
SoftSwitch because of the extension. SoftSwitch will also
serve as an interface between the GNU Radio Blocks and Ryu
SDN Controller. An IPC channel is created by opening sockets
inorder to establish communication between GNU Radio
Blocks and softswitch. There will be a custom message block
implementation in GNU Radio domain whose main task is
translation to message. The message translation is done from
softswitch to a format which is understood by block. The
message format which is used by USRP Source and USRP
Sink Block is Polymorphic Types(PMT). The main function of
source and sink is to send samples from and to the hardware
driver of USRP(UHD). Figure 1 shows the mechanism
discussed above.
Fig .10 Implementation of Crossflow [22]
V. ISSUES IN SDN AND CROSS LAYER DESIGN
As we discussed the novel approach of amalgamation of two
technologies which comes with added advantages of both the
technologies. There are certain areas which need to be
discussed which concerns with the issues of this integration.
The issues are concerned with the layers specifically. The
issues discussed can be divided layer wise. [34][35][45]
The Application layer problems can be unauthorized
authentication. As SDN has open source approach there can be
unauthorized authentication. Then the other types of issues
which are occurred are in regards to Accountability and access
control which might result in malicious activities to take place
which can be a threat to the whole network.[34] Then when the
flows are considered there can be a problem of payloads which
also is unsafe for the network as it might bypass the network
control policies. To solve such issues the software components
such as Perm OpenFLow, FRESCO, Flover can be used which
enables the flow supplication according to the rules of the
controller. The application like FRESCO [37] highlights the
security issue helps in encryption to safeguard the data [36].
As the SDN controller is treated as the brain of the computer
the logical unit can be exposed to many threats related to
controller specific attacks and scalability threats[51].
Controller specific attacks can be very damaging as it will
affect the entire system.[36][38] The solution to these issues
can be having multiple controllers which will act as backup if
one fails and also dividing the network in subdomains. Hyper
flow and flow checker can be used to deploy multiple
controllers which will be minimizing the scalability threat as
well as the controller attack threat.[21] Few of the
misconnection in the network due to external attacks can be
treated with the help of Flow checker.[38][45][50]
The data plane attacks are centralized to the data considering
that the form of the data is untraceable making it unreliable for
the networks. The SDN system can be at risk due to the data
plane vulnerabilities [36]. The solution to such issues in the
data plane can be minimized with the measures such as
Isolation of Data plane. Differentiating the data packets and
treating them with same infrastructure can minimize the issue
up to certain level.[38] Another way to avoid loss of data is
randomization which will identify the security breach. For the
overall security of the network OpenFlow provides
mechanisms which results in transport layer security. Mutual
authentication is required to have trust factor between the
components of the SDN architectures. [40][17][50]
VI. CONCLUSION
This paper shows the importance of integration of the two
newest and rising technologies Software defined Networking
and Cross layer Design. These imminent technologies have
high capabilities of taking traditional networks to new heights.
We saw two examples of integration of these technologies
which clarified the potential of these two technologies. With
the advances taking place in the internet fields it becomes
important to come up with some solid solutions to the network
problems. Even though there is a lot of scope considering the
current condition of the integration of the aforementioned
technologies, they hold a lot of potential to improve the basic
networks. We also discussed the issues which are faced
currently by these technologies and also discussed solution for
the same. More research is being done to formulate these

technologies and then applying to network technologies
making the internetworking bigger and btter.
VII.REFERENCES
[1] D. Kreutz, F. M. V. Ramos, P. E. Veríssimo, C. E.
Rothenberg, S. Azodolmolky and S. Uhlig, "Software-
Defined Networking: A Comprehensive Survey," in
Proceedings of the IEEE, vol. 103, no. 1, pp. 14-76, Jan.
2015.
[2] S. Shakkottai, T. S. Rappaport and P. C. Karlsson, "Cross-
layer design for wireless networks," in IEEE
Communications Magazine, vol. 41, no. 10, pp. 74-80, Oct.
2003.
[3] B. Fu, Y. Xiao, H. J. Deng and H. Zeng, "A Survey of
Cross-Layer Designs in Wireless Networks," in IEEE
Communications Surveys & Tutorials, vol. 16, no. 1, pp.
110-126, First Quarter 2014.
[4] C. Luo, F. R. Yu, H. Ji and V. C. M. Leung, "Cross-Layer
Design for TCP Performance Improvement in Cognitive
Radio Networks," in IEEE Transactions on Vehicular
Technology, vol. 59, no. 5, pp. 2485-2495, Jun 2010.
[5] Li Ma, Fei Yu, V. Leung and T. Randhawa, "A new
method to support UMTS/WLAN vertical handover using
SCTP," 2003 IEEE 58th Vehicular Technology
Conference. VTC 2003-Fall (IEEE Cat. No.03CH37484),
Orlando, FL, 2003, pp. 1788-1792 Vol.3.
[6] Q. Guan, F. R. Yu, S. Jiang and G. Wei, "Prediction-Based
Topology Control and Routing in Cognitive Radio Mobile
Ad Hoc Networks," in IEEE Transactions on Vehicular
Technology, vol. 59, no. 9, pp. 4443-4452, Nov. 2010.
[7] R. Xie, F. R. Yu and H. Ji, "Dynamic Resource Allocation
for Heterogeneous Services in Cognitive Radio Networks
With Imperfect Channel Sensing," in IEEE Transactions
on Vehicular Technology, vol. 61, no. 2, pp. 770-780, Feb.
2012.
[8] Fei Yu and V. C. M. Leung, "Mobility-based predictive
call admission control and bandwidth reservation in
wireless cellular networks," Proceedings IEEE INFOCOM
2001. Conference on Computer Communications.
Twentieth Annual Joint Conference of the IEEE Computer
and Communications Society (Cat. No.01CH37213),
Anchorage, AK, USA, 2001, pp. 518-526 vol.1.
[9] Fei Yu, V. Krishnamurthy and V. C. M. Leung, "Cross-
Layer optimal connection admission control for variable bit
rate multimedia traffic in packet wireless CDMA
networks," in IEEE Transactions on Signal Processing,
vol. 54, no. 2, pp. 542-555, Feb. 2006.
[10] L. Zhu, F. R. Yu, B. Ning and T. Tang, "Cross-Layer
Handoff Design in MIMO-Enabled WLANs for
Communication-Based Train Control (CBTC) Systems," in
IEEE Journal on Selected Areas in Communications, vol.
30, no. 4, pp. 719-728, May 2012.
[11] J. Zuo, C. Dong, S. X. Ng, L. Yang and L. Hanzo,
"Cross-Layer Aided Energy-Efficient Routing Design for
Ad Hoc Networks," in IEEE Communications Surveys &
Tutorials, vol. 17, no. 3, pp. 1214-1238, thirdquarter 2015.
[12] R. Xie, F. R. Yu, H. Ji and Y. Li, "Energy-Efficient
Resource Allocation for Heterogeneous Cognitive Radio
Networks with Femtocells," in IEEE Transactions on
Wireless Communications, vol. 11, no. 11, pp. 3910-3920,
November 2012.
[13] http://kliazovich.org/papers/whycross-layer.pdf
[14]http://shodhganga.inflibnet.ac.in/bitstream/10603/122077/
10/10_chapter2.pdf
[15] http://msw3.stanford.edu/~zhuxq/adhoc/
[16] B. Chang, Y. Liang and J. Jin, "Adaptive cross-layer-
based TCP congestion control for 4G wireless mobile
cloud access," 2016 IEEE International Conference on
Consumer Electronics-Taiwan (ICCE-TW), Nantou, 2016,
pp. 1-2.
[17] K. Wrona, M. Amanowicz, S. Szwaczyk and K.
Gierłowski, "SDN testbed for validation of cross-layer
data-centric security policies," 2017 International
Conference on Military Communications and Information
Systems (ICMCIS), Oulu, 2017, pp. 1-6.
[18] Y. Cheng, T. A. N. Nguyen, M. M. Rahman, S.
Gangadhar and J. P. G. Sterbenz, "Cross-layer geodiverse
protocol stack for resilient multipath transport and routing
using OpenFlow," 2016 12th International Conference on
the Design of Reliable Communication Networks (DRCN),
Paris, 2016, pp. 103-105.
[19] G. Carella et al., "Cross-layer service to network
orchestration," 2015 IEEE International Conference on
Communications (ICC), London, 2015, pp. 6829-6835.
[20] V. Karyotis, E. Stai and S. Papavassiliou, "A Component-
Based Cross-Layer Framework for Software Defined
Wireless Networks," 2016 8th IFIP International
Conference on New Technologies, Mobility and Security
(NTMS), Larnaca, 2016, pp. 1-6.
[21] K. Hasan, S. Shetty and T. Oyedare, "Cross layer attacks
on GSM mobile networks using software defined radios,"
2017 14th IEEE Annual Consumer Communications &
Networking Conference (CCNC), Las Vegas, NV, 2017,
pp. 357-360.
[22] P. Shome, M. Yan, S. M. Najafabad, N. Mastronarde and
A. Sprintson, "CrossFlow: A cross-layer architecture for
SDR using SDN principles," 2015 IEEE Conference on
Network Function Virtualization and Software Defined
Network (NFV-SDN), San Francisco, CA, 2015, pp. 37-39.
[23] Wolfgang John, Kostas Pentikousis, George Agapiou,
et.al., Research Directions in Network Service Chaining,
Presented at the 2013"Software Defined Networking for
Future Networks and Services"(SDN4FNS) conference in
Trento, Italy, published athttp://arxiv.org/abs/1312.5080
on 18 Dec. 2013
[24] ETSI, Network Function Virtualization (NFV),
ArchitecturalFramework,http://www.etsi.org/deliver/etsi_g
s/NFV/001_099/002/01.01.01_60/gs_NFV002v010101pdf

[25] N. McKeown et al., ”OpenFlow: Enabling Innovation in
CampusNetworks,” ACM Comp. Commun. Rev., Apr.
2008.
[26] Hyojoon Kim; Feamster, N., "Improving network
management withsoftware defined networking,"
Communications Magazine, IEEE, vol.51, no.2,
pp.114,119, February 2013
[27] Civanlar, S.; Parlakisik, M.; Tekalp, A.M.; Gorkemli, B.;
Kaytaz, B.;Onem, E., "A QoS-enabled OpenFlow
environment for Scalable Videostreaming,"GLOBECOM
Workshops (GC Wkshps), 2010 IEEE , vol.,no.,
pp.351,356, 6-10 Dec. 2010
[28] CPqD OpenFlow 1.3 Software
Switch.http://cpqd.github.io/ofsoftswitch13
[29] H.-H. Cho, C.-F. Lai, T.K. Shih, and H.-C. Chao.
Integration of SDR and SDN for 5G. Access, IEEE,
2:1196–1204, 2014.
[30] V. Mancuso, C. Vitale, R. Gupta, K. Rathi, and A.
Morelli. A prototyping methodology for SDN-controlled
LTE using SDR. 2014.
[31] N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar,
L. Peterson, J. Rexford, S. Shenker, and J. Turner.
Openflow: Enabling Innovation in Campus Networks.
ACM SIGCOMM Computer Communication
Review,38(2):69–74, 2008
[32] M. Yan, J. Casey, P. Shome, A. Sprintson, and A. Sutton.
Aetherflow: Principled wireless support in SDN. In
Proceedings of the ICNP 2015 Workshop on Control,
Cooperation, and Applications in SDN protocols
(CoolSDN ’15), 2015.
[33]C. J. Casey, A. Sutton, and A. Sprintson. TinyNBI:
Distilling an API from Essential OpenFlow Abstractions.
In Proceedings of the Third Workshop on Hot Topics in
Software Defined Networking, HotSDN ’14, pages 37–42,
New York, NY, USA, 2014. ACM
[34] D. Migault, “Identifying and addressing the
vulnerabilities and security issues of SDN,” no. January
2016, 2015
[35] Q. Yan, F. R. Yu, Q. Gong, and J. Li, “Software-defined
networking (SDN) and distributed denial of service
(DDOS) attacks in cloud computing environments: A
survey, some research issues, and challenges,” IEEE
Commun. Surv. Tutorials, vol. 18, no. 1, pp. 602–622,
2016
[36] I. Ahmad, S. Namal, M. Ylianttila, and A. Gurtov,
“Security in Software Defined Networks: A Survey,” IEEE
Commun. Surv. Tutorials, vol. 17, no. 4, pp. 2317–2346,
2015
[37]S. Shin, P. Porras, V. Yegneswaran, M. Fong, and G. Gu,
“FRESCO: Modular Composable Security Services for
Software-Defined Networks,” Netw. Distrib. Syst. Secur.
Symp. (NDSS 2013), vol. 2, no. February, pp. 1–16, 2013
[38] S. T. Ali, V. Sivaraman, A. Radford, and S. Jha,
“Securing Networks Using Software Defined Networking :
A Survey,” IEEE Trans. Reliab., vol. 64, no. 3, pp. 1–12,
2013
[39] A. Science, “A Cross-Layer Design for Large Data
Transfers,” no. May, 2016.
[40] F. R. L. S. Ell, “Software-Defined Networking Security:
Pros And Cons,” no. October, pp. 40–45, 2012
[41] W. Cerroni, M. Gharbaoui, B. Martini, A. Campi, P.
Castoldi, and F. Callegati, “Cross-layer resource
orchestration for cloud service delivery: A seamless SDN
approach,” Comput. Networks, vol. 87, pp. 16–32, 2015.
[42] H. Yang, L. Cheng, J. Deng, Y. Zhao, J. Zhang, and Y.
Lee, “Cross-layer restoration with software defined
networking based on IP over optical transport networks,”
Opt. Fiber Technol., vol. 25, pp. 80–87, 2015.
[43] S. Nanda, K. Balachandran, and S. Kumar, “Adaptation
techniques in wireless packet data services,” IEEE
Commun. Mag., vol. 38, no. 1, pp. 54–64, 2000
[44] H. Zimmermann, “OSI Reference Model-The ISO Model
of Architecture for Open Systems Interconnection,” IEEE
Trans. Commun., vol. 28, no. 4, pp. 425–432, 1980
[45] V. Shamugam, I. Murray, J. A. Leong, and A. S. Sidhu,
“Software defined networking challenges and future
direction: A case study of implementing SDN features on
OpenStack private cloud,” IOP Conf. Ser. Mater. Sci.
Eng., vol. 121, no. 1, 2016.
[46] William Stallings, “Software-Defined Networks and
OpenFlow - The Internet Protocol Journal, Volume 16, No.
1,” vol. 16, no. 1, p. 11.
[47] R. Jain, “Introduction to Software Defined Networking
( SDN ),” Washingt. Univ. Saint Louis, vol. 11, no. 7, pp.
1–44, 2013.
[48] Vmware, “Software-Defined Datacenter – In Depth,” 9.
[Online]. Available:
https://www.vmware.com/ca/solutions/software-defined-
datacenter/in-depth.html
[49] B. Raghavan et al., ‘‘Software-definedinternet
architecture: Decouplingarchitecture from infrastructure,’’
inProc. 11th ACM Workshop Hot TopicsNetw., 2012, pp.
43–48.
[50] H. Kim and N. Feamster, ‘‘Improving network
management with softwaredefined networking,’’ IEEE
Commun. Mag.,vol. 51, no. 2, pp. 114–119, Feb. 2013.
[51] S. Yeganeh, A. Tootoonchian, and Y. Ganjali,‘‘On
scalability of software-definednetworking,’’ IEEE
Commun. Mag.,vol. 51, no. 2, pp. 136–141, Feb. 2013.