Software Defined Networking and Virtualization

The assignment discusses the importance of software defined networking (SDN) and virtualized networking in cloud computing. It highlights the benefits of using SDN and NFV, including scalability, resource isolation, fast resource allocation, and centralized control. The document also explores the limitations of geography-dependent network design and the role of L2 and L3 networks in rewriting applications. Overall, this assignment provides a comprehensive understanding of cloud networking techniques and their applications.

Introduction
Cloud computing revolutionized the online availability of on-demand computing
resources such as data storage, especially Computing powers, etc. Cloud
computing is considered a Software as a service (SAAS), Platforms as a service
(PAAS), and Infrastructure as a service. SAAS provides software resources for
developers, whereas (PAAS) provides the platform services on clouds and (IAAS)
delivers the most central VM servers and related Infrastructure utilizes by cloud
customers. To provides all these services, providers maintain and manage large
scale data centers where thousands of host servers are connected through numerous
network switches.
For decades, Cloud computing services are managed by service providers
traditionally, where the data centers require high administration. This can be very
costly and time consuming from your business point of view. Due to the
heterogeneity in resources, these cloud-based platforms become more complex
while management and many other issues were also arising.
To deal with all shortcomings of traditional cloud computing, software-defined
networking (SDN) concept is adopted by data cloud centers. Software-defined
networking (SDN) concept is a network architecture used to manage the network
through centralized control and programmed intelligentially using software
applications [1]. SDN manages or controls the network flow efficiently because the
software controller observes the whole network and makes decisions more quickly
than traditional networks.
In addition to these major concepts of cloud computing and SDN used with cloud
computing, many other concepts like network virtualization, network layers, SDN
enabled cloud computing, SDN impact, and future directions will be discussed
below.
Related work
There are various criteria set for networking that requires to be followed [3]. There
are different networking basic bodies; some of them are International Organization
for Standardization (ISO), International Telegraph and Telephone Consultative
Committee (CCITT), and so on... OSI model layers are the following:

 The first layer is defined as a physical layer that sends and receives
packets.
 The data link layer uses logical link control and media access control.
 The third layer is the network layer, which receives packets by using IP
protocols.
 The fourth layer is the transport layer, which uses TCP protocols.
 The fifth layer is a session layer that is used for authentication and
authorization.
 The sixth layer is a presentation layer for WWW and (HTTP).
 The seventh layer is an application that interacts with the application
through GUI.
In the early stages of cloud computing, service providers manage cloud data
centers, which consisted of thousands of host servers requiring high administration
in managing the data centers. Due to this fatiguing task's complexities, providers
faced numerous problems while providing different types of cloud services to
users. In the traditional network approach in cloud computing, the control plane is
used over or on top of forwarded planes. The decision-making process is done by
each of the switches in the network according to the information he gathers from
its neighbor devices in a distributed state. As each device makes the network's
decision, this can increase the complexity and bring unpredictability in the
network.
On the other hand, SDN provides centralized control into a software controller and
observe the network entirely. As the information is gathered centrally, the
controller can easily decide according to the situation. Providers can easily
program control logic into the controller and send these packets to network
switches in the entire network through SDN protocols. Any control login will be
designed and implemented into the SDN as shown below in figure 1 [2].

Figure 1
The world of cloud computing has acknowledged SDN as the most
advantageous software in cloud computing due to its helpful features like
easily programmable control planes, security enhancements, frequently
changing workloads, etc. The basic technologies used to develop the SDN
enabled cloud computing are the virtualization of servers, SDN, and virtual
middleboxes, which are like the network appliances like firewalls to filter,
transform and maintain the packet forwarding. Moreover, OpenFlow
provides us the interface to cache the packet forwarding in switch tables and
notification for topology change.
SDN architecture is based on cloud managers, centralized SDN controller,
East-west APIs, cloud resources, hypervisors, resource management
switches, and SDN controllers, etc. All these components of SDN
architecture are managed the data cloud centers (DCN) collaboratively as
shown in figure 2 [3].

Figure 2
SDN in data centers has many objectives studied by researchers, and due to
these objectives, ' providers deploy SDN enabled cloud computing. The
most common objective is energy efficiency because cloud data centers
consume numerous amounts of energy every year. Through SDN
deployment, it is quite possible because the packets pass-through through a
smaller number of switches, and other resources remained off. Performance
is another objective of SDN, including the availability, network latency, and
quality of services (QoS), which is defined as providing a specific
bandwidth to the user. For inter and intra-DCN environment SDN use
quality of services (QoS) and virtual network environment for virtual
machine hosts. Virtualization is another objective of cloud computing.
Various virtualization technologies used in SDN enabled cloud computing
like a hypervisor, enabling multiple VM's on a single host. Network function
virtualization (NFV) is another example of SDN virtualization [2]. Security
objectives are also included in SDN to protect SDN resources and are more

important for protecting against Distributed Daniel of services attacks
(DDoS).
Critical analysis
Networking is one of the primary reasons for the development of cloud
computing and also its services. To execute networking in cloud computing,
various approaches and versions are styles to outcome optimum wanted
results. One of the versions is referred to as The Basic Reference Model for
Open Systems, which is the mix of ISO and CCITT models, which is based
upon the seven layers of the OSI version. This version supplies the basis of
cloud networking because most cloud computing services are based on
networking like scalability, virtualization, and so on.
Another method that is being complied with currently a day is Software
Defined Networking (SDN), in which networking is being controlled by the
software program and is made network smart. A network is managed by a
software application or a developed program, particularly for this objective.
One more used technique is Network Functions Virtualization (NFV), in
which network features are being virtualized into lots of blocks connected
via a network [4]. The additional improvement in networking is with
virtualization. The advantages of using the virtualized network are if there
are massive or small resources offered, it is scalable; the one user sources
are isolated from each other, taking care of the sources become much
simpler. The resource allotment comes to be fast as well as dependable.
Taking about SDN (Software Defined Networking), it also has various
approaches and benefits because it supplies different information and a
control board via which administration ends up being very easy. Also, SDN
has a systematized and programmatically regulated control panel. The
benefit of utilizing networking in a cloud computer is that it makes the cloud
solutions available to every user, managing the source pool becomes simple,
and resources are allocated to the customer according to their requirement.
The user is billed just for the solutions they use. SDN's style consists of
SDN supervisor that takes care of the entire renter, SDN controller controls
the geography, virtual network, network tracking, and network setup. These
elements are combined with a full design of SDN will be created.

Next to advantages, there are also some limitations of networking styles that
are dependent on geography. The design of geography relies on how L2 and
L3 networks are created, re-writing of application that has two issues
broadcast domain name lack as cloud, appointed IP addresses.
Conclusion
Advancement of cloud computing and execution of its services, provided to
the clients interconnected through a network, includes a key role of
software-defined networking (SDN) and virtualized networking. Networking
resources are now allocating to multiple users dynamically at the same time
by using networking methodologies. The efficiency and preciseness in SDN
services are due to the complete centralized control of software, which is
programmed lively. SDN also provides the control panel, which observes the
whole network for straightforward decision-making and data panel. Network
virtualization allows us to allocate every user a separate resource that can't
be accessed by any other users. In this way, it provides privacy to the user.
References
[1] A.Sill, “Standards underlying cloud networking”, IEEE Cloud
Computing, 3(3), pp. 76-80, 2016.
[2] R.Jain, and S.Paul, “Network Virtualization and software
defined networking for cloud computing: a survey”, IEEE
Communication Magazine, 51(11), pp. 24-31, 2013.
[3] J.Son and R.Buyya, “A taxonomy of software defined
networking (SDN)-enabled cloud computing”, ACM
Computing Surveys (CSUR), 51(3), pp. 1-26, 2018
[4] S. Azodolmolky, P. Wieder, and R. Yahyapour, “Cloud
computing networking: Challenges and opportunities for

innovations,” IEEE Commun. Mag., vol. 51, no. 7, pp. 54–62,
2013, doi: 10.1109/MCOM.2013.6553678.