Effect of Conflicting Flows on TCP and UDP Data Transfer Rates in
OpenFlow Switch for Software-Defined Networks (SDN)
MUTAZ HAMED HUSSIEN, MOHAMED KHALAFALLA HASSAN
Faculty of Engineering,
Future University,
Khartoum,
SUDAN
Abstract: - Software-defined networking (SDN) is a framework that enhances scalability and agility in
network simplicity and control. It is characterized by logical centralization, improving control, and data planes.
Nevertheless, additional investigation is needed to fully understand the influence of flow conflict on the transfer
rate variable. The present research aims to analyze the impact of flow conflict on the efficacy of the two
protocols, the transmission control protocol (TCP) and the user datagram protocol (UDP), respectively, by
utilizing throughput as a measure of efficiency. The measurements are verified through SDN OpenFlow
networking modeling using MININET. The findings reveal a significant average alteration in transfer rate for
TCP and UDP when SDN conflict rules are present, ultimately impacting network and operational efficiency.
Key-Words: - Software defines the network, OpenFlow Table, conflict flow, transfer rate, TCP, and UDP.
Received: August 12, 2023. Revised: October 27, 2023. Accepted: November 29, 2023. Published: December 31, 2023.
1 Introduction
SDN is a networking infrastructure technology that
makes networks more efficient, adaptable, and
programmable, [1], [2]. This approach has resulted
in numerous benefits. Administrators and systems
engineers can apply modifications through a
centralized management interface to address new
business demands, [3]. Software-defined networking
(SDN) significantly advances network agility by
integrating specialized technological paradigms.
The strategic decoupling of the network's control
and data-forwarding elements is central to the SDN
architecture. This separation facilitates the
independent configuration of the network's primary
controller, commonly referred to as the kernel, thus
empowering network architects with comprehensive
authority over the foundational structure of the
network. SDN stands out in its ability to effectively
govern and regulate network behavior, exhibiting
remarkable reactivity and cost-efficiency, [4].
OpenFlow (OF) is a critical Software-Defined
Networking (SDN) protocol. In the context of SDN,
the controller is another crucial component of
OpenFlow. The controller's primary role is to
integrate platforms and enable the creation of
applications using a northbound Application
Programming Interface (API). OpenFlow facilitates
the establishment of an interface among controllers
and the forwarding plane, allowing a more effective
and adaptive response to changing company
demands, [5], [6]. The OpenFlow architecture of
Software-Defined Networking (SDN) is depicted in
Figure 1.
The OpenFlow protocol is the cornerstone for
communication between OpenFlow switches and
controllers. This protocol operates through flow
tables embedded within the switching devices,
which are instrumental in forging the connections.
Utilizing these flow tables as a conduit, the
OpenFlow protocol facilitates the establishment of
interconnectivity. These tables play a pivotal role in
ensuring that data transmissions are accurately
conveyed, executed, and dispersed, adhering to the
specific flow entries. This mechanism is critical in
maintaining efficient and effective network
operations, [7], [8].
Fig. 1. OpenFlow SDN Architecture, [9]
WSEAS TRANSACTIONS on COMPUTERS
DOI: 10.37394/23205.2023.22.35
Mutaz Hamed Hussien, Mohamed Khalafalla Hassan
E-ISSN: 2224-2872
311
Volume 22, 2023
2 Problem Formulation
Different types of flow conflicts can affect both
traditional networks and software-defined networks
(SDNs). making it difficult for networks to connect.
These conflicts in SDNs can be categorized into two
main types based on their adherence to SDN rules
and their impact on networks, [10]. In the realm of
Software-Defined Networking (SDN), flow entry
conflicts can arise under many scenarios, notably
during the establishment of an SDN network
wherein a switch interfaces with several virtual
networks. These conflicts may emerge whether the
network is under the purview of a single controller
or a group of controllers, leading to potential
discrepancies in flow management. Additionally, a
distinct category of flow conflict is observed when
packets of ambiguous importance correspond with
existing flow entries, further complicating the
network's operational dynamics. These scenarios
underscore the complex interplay of elements within
SDN configurations, necessitating meticulous
oversight for optimal network functionality, [11],
[12], [13].
Flow entries in an open flow table can cause
security module malfunctions. Correlating packets
with their priorities helps, but interconnected flow
entries can cause conflicts, [14], [15].
3 Problem Solution
TCP and UDP are commonly utilized protocols for
sending and receiving packets and data between
networked devices inside a network, [16]. Assessing
network throughput is essential for improving
performance. Consequently, numerous notable
research has concentrated on analyzing and
assessing the data transmission capacities of TCP
and UDP protocols in Software-Defined Networks
(SDN). These investigations have thoroughly
analyzed the efficacy of TCP and UDP protocols in
terms of transfer rates. In addition, they have
utilized a specialized model to evaluate the
effectiveness of these protocols when incorporated
into an SDN framework.
This research is crucial in identifying the most
efficient network architectures for SDNs,
guaranteeing optimal data transmission and overall
network performance, [10], [17].
The primary objective of this study is to assess
and track the TCP and UDP throughput metrics in
software-defined networking (SDN). The transfer
rate is evaluated under two circumstances. The
normal flows are operational in the network
architecture in the first case. In the second case,
additional rules are introduced to the OpenFlow
switch to create conflicting flows that run parallel to
the normal flow. The execution of this project
involves the utilization of the Ryu controller, which
employs the simple and fat tree topology network
that has been built and tested within the Mininet
software. Figure 2 depicts the suggested approach.
Fig. 2. Proposed Method Process
The initiation of the process begins with the
establishment of a Software-Defined Network
(SDN) within the Mininet environment. Following
this, a specific network topology is selected, with
the example here including both simple-tree and fat-
tree topologies. The next critical step is the precise
configuration of the controller and the OpenFlow
switch. This stage involves the transmission of
network packets and a detailed examination of the
characteristics of TCP and UDP protocols. The
process then progresses to the implementation of
conflict rules in the OpenFlow switches. This leads
to the retransmission of packets within the network,
necessitating a reevaluation based on their TCP and
UDP properties. After the data capture is completed,
it undergoes rigorous analysis and further
evaluation. Notably, the conflict rules are
systematically applied to the OpenFlow switch
across both topologies at varying intervals, ranging
from 0 to 3600 seconds. This systematic approach
allows for a comprehensive understanding of the
network dynamics under different configurations
and timeframes.
Record Data
Remeasure the TCP/UDP(Transfer Rate)
Regenerate the Packets in Network
Implement and apply Conflict Rules in Flow table
Measure TCP/UDP(Transfer Rate)
Generate Normal Packets in Topology
Define Ryu Controller and OpenFlow Switch 1.3
Select Type of Topology
Create SDN Topology
WSEAS TRANSACTIONS on COMPUTERS
DOI: 10.37394/23205.2023.22.35
Mutaz Hamed Hussien, Mohamed Khalafalla Hassan
E-ISSN: 2224-2872
312
Volume 22, 2023
4 Methodology
This section provides a thorough explanation of
configuring SDN platforms and creating simulations
for research purposes. It also includes a complete
description of the frameworks and methods utilized
in the conducted tests.
4.1 Configuration Established for SDN
Architecture
The objective of this investigation is to generate,
gather, and safeguard contradictory and consistent
data streams using a simulation software-defined
networking (SDN) architecture. To accomplish this,
we are using the Mininet framework, coupled with
the Ryu controller and the Mininet simulator. The
Mininet framework has VirtualBox installed, which
is a virtual communication network. For a detailed
overview of the simulation settings in the simulation
environment, please refer to Table 1.
Table 1. System and environment specification for
Mininet simulation
Software and Hardware
Specifications
Processor
Core i7
RAM
16 GB
Operation System
Ubuntu 18.04
Controller
Ryu
Programming
Python 2.7
OpenFlow Switch
Version 1.3
4.2 Model Development
The SDN module that forwards the simulation
comprises two types of constraints. The initial
configuration enables regular operation and
functionality, while the subsequent configuration
enforces conflict rules in the OpenFlow switch.
OpenFlow is a protocol that allows the SDN
controller to interact with the switch to determine
the path of data packets. This SDN network is
created within the Mininet framework, a versatile
emulator capable of simulating intricate network
environments comprising multiple hosts, switches,
and controllers. Mininet is widely utilized for testing
a variety of network applications, protocols, and
topologies under diverse operational conditions.
Customization of Mininet topologies is achieved
through Python scripts, enabling the design of
network structures tailored to specific simulation
scenarios. The study in question delves into the
topologies deployed within a Mininet network,
particularly focusing on their connection to the Ryu
controller. Ryu stands out as a component-based
SDN controller, offering network administrators the
flexibility to devise custom traffic management
rules. The simple tree topology explored includes
three switches and four hosts connected to the Ryu
controller. In contrast, the fat-tree topology
comprises a more complex arrangement of seven
switches and eight hosts, all integrated with the Ryu
controller. This comprehensive exploration provides
valuable insights into the functionalities and
adaptabilities of different network architectures
within an SDN environment.
5 Result and Discussion
The results of two protocols, TCP and UDP, are
presented in Figure 3 and Figure 4, respectively.
The graphs show the throughputs of the two
architectures used in this research. The numbers in
the graphs are generated by simulating transmitted
data collected over 3600 seconds, with
measurements taken at 360-second intervals. The
illustrations depict the usual flows in the primary
tree topology and fat-tree topology using dark blue
and dark brown lines. However, the introduction of
conflict rules is represented by white blue, and
yellow lines, which show the altered flows. The
results indicate that the conflict rules have a
negative impact on network forwarding
transmission, resulting in a reduced transfer rate for
conflict flows as compared to expected flows during
all test durations.
Figure 3 compares TCP transfer rates in simple
tree and fat-tree networks. The evaluation was done
for 0-3600 seconds, and two cases were examined:
with and without conflict rules. Without conflict
rules, simple tree topology showed varied transfer
rates from 600-771 Gb/s, and fat-tree had 564-647
Gb/s. With conflict rules, transfer rates decreased
for both topologies. For simple tree topology, the
results fluctuated between 478-673 Gb/s, and for the
fat-tree topology, the results ranged from 213-488
Gb/s. During the 0-3600 seconds period, the TCP
transfer rate decreased by 10% and 31% for simple
tree and fat-tree topologies, respectively.
Fig. 3. TCP Transfer Rate for Two Topologies
WSEAS TRANSACTIONS on COMPUTERS
DOI: 10.37394/23205.2023.22.35
Mutaz Hamed Hussien, Mohamed Khalafalla Hassan
E-ISSN: 2224-2872
313
Volume 22, 2023
Figure 4 compares UDP transfer rates for simple
tree and fat-tree topologies with and without conflict
rules. The bare tree had transfer rates of 19.1-20
Gb/s, while the fat tree had rates of 18.1-21 Gb/s.
With conflict rules, both topologies experienced
declines, with introductory tree rates at 17.1-18.2
Gb/s and fat-tree rates at 10.1-13 Gb/s. UDP
transmission decreased by 11% for simple trees and
44% for fat trees over 0-3600 seconds.
Fig. 4. UDP Ttransfer Rrate for Two Topologies
The evaluation of two scenarios, one employing
the TCP protocol and the other utilizing the UDP
protocol, is conducted by assessing the quantity of
data transmitted. Figure 3 displays a decline in TCP
transferred data within time intervals ranging from 0
to 3600 seconds. The simple tree topology shows a
loss of 10%, while the fat-tree topology shows a
drop of 31%. Similarly, Figure 4 compares UDP
data transmission over extended time intervals
ranging from 0 to 3600 seconds. The conflicting
flows in UDP protocol also lead to a reduction in
data transfer, specifically by 11% for a basic tree
topology and 44% for a fat-tree topology. Lastly,
Table 2 presents a comparison of the mean decrease
in TCP and UDP protocols under both implemented
conditions.
Table 2. Comparison of the average drop-in transfer
rate with flow conflict.
TCP
UDP
10%
11%
31%
44%
6 Conclusion
This investigation focuses on evaluating the
message transmission variable in Software-Defined
Networking (SDN) for both regular and conflict
flows in a standard forwarded SDN environment. To
achieve this, we utilized the Mininet software
simulator to execute an algorithmic model. Our
findings demonstrate that the implementation of
conflict rules significantly impacts data transport for
both TCP and UDP protocols. Additionally, the data
highlights an apparent difficulty with conflict flows
in OpenFlow switches, suggesting that further
research is necessary to address this issue by
resolving the observed losses. One of the key
takeaways from this research is that while
implementing software-defined networking (SDN),
it is essential to consider the issue of conflict flows,
especially if the programs or services that are
utilized demand an improvement in transfer rate. As
such, subsequent investigators ought to emphasize
the improvement of safety attributes in the SDN
network while maintaining the network's quality of
service.
References:
[1] Pisharody, S., et al., Brew: A security policy
analysis framework for distributed SDN-
based cloud environments. IEEE
Transactions on Dependable and Secure
Computing, 2017. 16(6): p. 1011-1025.
[2] Khairi, M.H.H., et al., Detection and
classification of conflict flows in SDN using
machine learning algorithms. IEEE Access,
2021. 9: p. 76024-76037.
[3] Cui, J., et al. Transaction-based flow rule
conflict detection and resolution in SDN. In
2018 27th International Conference on
Computer Communication and Networks
(ICCCN), 2018, IEEE.
[4] Khairi, M.H.H., et al., A Review of Flow
Conflicts and Solutions in Software Defined
Networks (SDN). IIUM Engineering
Journal, 2021. 22(2): p. 178-187.
[5] McKeown, N., et al., OpenFlow: enabling
innovation in campus networks. ACM
SIGCOMM Computer Communication
Review, 2008. 38(2): p. 69-74.
[6] Adeniji Oluwashola David, O.I.O., "Scalable
Flow based Management Scheme in
Software Define Network (SDN) using
sFlow," WSEAS Transactions on Computers,
vol. 22, pp. 64-69, 2023,
https://doi.org/10.37394/23205.2023.22.7.
[7] Lo, C.-C., P.-Y. Wu, and Y.-H. Kuo. Flow
entry conflict detection scheme for software-
defined network. In 2015 International
Telecommunication Networks and
Applications Conference (ITNAC). 2015.
IEEE.
[8] Hauser, F., et al., P4-MACsec: Dynamic
topology monitoring and data layer
protection with MACsec in P4-SDN. arXiv
preprint arXiv:1904.07088, 2019.
WSEAS TRANSACTIONS on COMPUTERS
DOI: 10.37394/23205.2023.22.35
Mutaz Hamed Hussien, Mohamed Khalafalla Hassan
E-ISSN: 2224-2872
314
Volume 22, 2023
[9] Khairi, M.H.H., et al., et al., A Review of
Anomaly Detection Techniques and
Distributed Denial of Service (DDoS) on
Software Defined Network (SDN).
Engineering, Technology & Applied Science
Research, 2018. 8(2), p.2724-2730.
[10] Wang, M.-H., et al., SDUDP: A reliable
UDP-Based transmission protocol over SDN.
IEEE Access, 2017. 5, p.5904-5916.
[11] Khairi, M.H., et al., Generation and
collection of data for normal and conflicting
flows in software defined network flow table.
Indonesian J. Electr. Eng. Comput. Sci.,
2021. 22(1), p.307.
[12] Zaw, H.T. and A. Maw, Traffic management
with elephant flow detection in software
defined networks (SDN). International
Journal of Electrical and Computer
Engineering, 2019, 9(4), p.3203.
[13] Wael Hosny Fouad Aly, H.K., Nour Mostafa,
Samer Alabed, "Towards Securing
OpenFlow Controllers for SDNs using
ARMA Models," International Journal of
Applied Mathematics, Computational
Science and Systems Engineering, vol. 4, pp.
21-29, 2022, DOI:
10.37394/232026.2022.4.3.
[14] Ali, T.E., A.H. Morad, and M.A. Abdala,
Traffic management inside software-defined
data center networking. Bulletin of Electrical
Engineering and Informatics, 2020, 9(5),
p.2045-2054.
[15] Monika, P., R.M. Negara, and D.D. Sanjoyo,
Performance analysis of software defined
network using intent monitor and reroute
method on ONOS controller. Bulletin of
Electrical Engineering and Informatics,
2020, 9(5), p.2065-2073.
[16] Khairi, H., et al., The impact of firewall on
TCP and UDP throughput in an OpenFlow
software defined network. Indonesian
Journal of Electrical Engineering and
Computer Science, 2020, 20(1), p.256-263.
[17] Gu, Y. and R.L. Grossman, UDT: UDP-
based data transfer for high-speed wide area
networks. Computer Networks, 2007, 51(7),
p.1777-1799.
Contribution of Individual Authors to the
Creation of a Scientific Article (Ghostwriting
Policy)
Mutaz Hamed Hussien, Mohamed Khalafalla,
carried out the simulation, and the optimization
Manuscript writing.
Sources of Funding for Research Presented in a
Scientific Article or Scientific Article Itself
No funding was received for conducting this study.
Conflict of Interest
The authors have no conflicts of interest to declare.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the
Creative Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en
_US
WSEAS TRANSACTIONS on COMPUTERS
DOI: 10.37394/23205.2023.22.35
Mutaz Hamed Hussien, Mohamed Khalafalla Hassan
E-ISSN: 2224-2872
315
Volume 22, 2023