Network Virtualization: New Challenges

Abstract- Network virtualization has been considered as an enabling technology for

future network, through which multiple heterogeneous virtual networks can run on a
shared infrastructure. In order to study and test the network management mechanism
of future network, we develop and implement a network operation simulation platform
of the network virtualization environment. It is the method of splitting
up network resources into separate bandwidth channels that are each independent from
one another, and that can be assigned and reassigned to servers or
other network devices. Its purpose is to optimize the speed, reliability, and flexibility of
the network. But it can have a high cost of implementation and creates a security risk.
Basically it requires several links in a chain that must work together cohesively and
creates an availability issue. Initially, network virtualization only involved the
separation of the control plane and the forwarding plane (management and packet
transmission) within networking devices like switches. This has shifted to include the
totality of virtualizing a network, including how the network is programmed,
administered, and deployed, be it hardware, software, or management and packet
transmission. This paper evaluates the potential of network virtualization from an
operator’s perspective and based on possible scenarios for implementing and using
network virtualization, new business roles and models are examined. Open issues and
topics for further evaluation are identified. In summary, the objective is to identify the
challenges raised by network virtualization.

I. INTRODUCTION
Network virtualization is a technology which enables combining computer
network resources into a single platform appearing as a single network. In this
form of virtualization all hardware and software in the virtual network appear as
a single collection of resource. There is another form of network virtualization
in which logically isolated network partitions are created over the shared
physical network infrastructure. Such logical partitioning, as shown in Figure 1,
resulting in multiple heterogeneous virtual networks can coexist simultaneously
over the shared infrastructure.

Network virtualization is a useful technology to achieve better utilization of

infrastructures in terms of reusing a single physical or logical resource for
multiple other network instances, or to aggregate multiples of these resources to
obtain more functionality. These resources can be network components, such as
routers, switches, hosts, virtual machines, etc. Hence, Network virtualization
can reduce the total cost by sharing network resources.

Figure1: Logical Network partitions on a shared network infrastructure

Shared Physical Network Infrastructure

Although virtual networks are only logically isolated networks, over a common
infrastructure, but each such virtual network can provide the user, of logical
virtual network, network services similar to those provided by the common
infrastructure as a non-virtualized network. Network virtualization is not an
completely new concept but it has been in use to limited extent such as virtual
private networks (VPNs) & virtual local area networks (VLANs).

 Key features of Network Virtualization

Some of the key features of network virtualization are given below:
a) Partitioning: Network virtualization allows creation of logical network
partitions with a programmable control plane so that users can use
Protocols, network topologies, and functions as per their requirements.
The re-configurability capability of Network virtualization make the
logical partitions capable of easily and rapidly creating network
topologies and reconfiguration according to users requirements, the
status of networks, policies of resource owners, etc.
b) Isolation: The network virtualization technology is capable of providing
the isolation, among logical network partitions, so that there is no
interference which may affect network performance. This includes the
isolation to deterioration of the performance of a logical partition due to
exhaustion of network resource by a malicious logical partition.
c) Abstraction: In a logical network partition the network elements should be
controlled as abstract resources, in which a given virtual resource need not
directly correspond to its component resources. Network abstraction allows
hiding the underlying characteristics of network elements from the way in
which other network elements, applications, or users interact with those network
elements and separates infrastructure instances and control frameworks of
network virtualization.

d) Aggregation: Network virtualization makes it possible to provide high

performance resources for users by logically aggregating multiple resources into
single resource. To illustrate some of the key features, we may take Server
Virtualization as an example. Server virtualization software allows the user to
run multiple guest computers on a single host computer with those guest
computers believing they are running on their own hardware. By doing this,
user gain all the benefits of any type of virtualization such as portability of guest
virtual machines, reduced operating costs, reduced administrative overhead,
server consolidation, testing & training, and disaster recovery benefits.
Examples of server virtualization products are VMware Server, Workstation,
Player, and ESX Server, Microsoft Virtual PC and Virtual Server, Xen, Virtual
Iron, etc. Following are the different levels of server virtualization commonly
offered:
 Full virtualization: guest OS is unmodified and believes it is
running on the same hardware as the host OS.
 Para virtualization: guest OS is modified.
 Emulation: guest OS is unmodified but it is running on a
software emulated CPU.

 Benefits of Network Virtualization

Virtualization technology offers a unique opportunity for organizations to

improve efficiency and scalability and reduce overall operational costs and
complexity. It provides an efficient way to centrally manage network
resources, simplifying provisioning and maintenance tasks.
Virtual Networks attempt to better utilize networking infrastructure by reusing
individual routers or links (i.e., either physical or logical networking resource)
for multiple concurrent network instances, or to aggregate multiple such
resources to obtain increased capabilities. These resources can be any network
component, including routers, hosts, links, and services, (e.g., name mapping
services). Increased capability can refer to aggregate capacity provided by
bundles of links or groups of routers, or increased fault tolerance of a cluster of
primary and backup service systems.

Network virtualization when done at the device level reduces the number of
physical network devices, or when done at the network level, by creating
multiple logical networks, enables full utilization of one physical network.
 In brief, following are some of the benefits of network virtualization:
 It helps in de-ossifying the current network architectures.
 It allows multiple virtual networks to coexist over a shared physical
infrastructure.
 It provides paths to the future networks approaches.
 It allows the deployment of new business roles and players.
 It reduces/shares cost of ownership.
 It optimizes the resource (network infrastructure) usage.

Architecture of Network Virtualization

Network virtualization is required to be capable of providing multiple partition
of network appearing to be isolated from each other. These partitions, also
referred to as Logically Isolated Network Partitions (LINP),may be created over
the single physical infrastructure. Figure 2 shows multiple LINPs created in a
network virtualization framework. Each LINP is isolated from each other and is
programmable to satisfy the user’s demand on the functionality and amount.
Users’ demand is conveyed to an entity known as LINP manager which
coordinates infrastructures resource so that appropriate LINP is provided to the
user as per the user’s demand.
 Figure 2: Network virtualization

II. LITERATURE SURVEY

[1] IEEE 2015, Network Functions Virtualization: The Portability Challenge

From this paper ,it can inferred that Network Virtualization is a shift in the
telecom industry, where the functions traditionally hosted in purpose-built
equipment become virtualized and run on pools of standard servers. However,
delivering high performance while being portable across commodity servers is
challenging , while this article proposes an approach to address both
requirements at the same time, relying on the notion of synthetic devices
abstracting hardware-dependent acceleration functions.

Some of the performance challenges are - NFV acceleration, Software

acceleration and Hardware acceleration.

The portability challenges include –

Software Dependencies: In NFV framework use of hardware acceleration

techniques can a real issue, unlike in non-virtualized environments. Though, the
NFV framework provides provision to load, execute, and move VNFs across
different servers.
NFV management and orchestration aspects: This approach defeats the key
benefits of NFV. Network operators need to resort to stopgap measures such as
using a homogeneous server pool. So, the major challenge here is how to
achieve high performance by leveraging hardware accelerators and at the same
time have hardware-independent VNFs.

[2] IEEE 2015, Network Function Virtualization: Challenges and Opportunities

for Innovations

From this paper, we get an overview of Network Function Virtualization (NFV)

from its room for innovation to the challenges faced. NFV is proposed to
improve the network flexibility of network service provisioning and reduce the
time to market new services. NFV separates software implementation of
network functions from the hardware. The challenges it faces are about
virtualized network functions(VNF) in their network performance, it is shown
that virtualization may lead to abnormal latency variations and significant
instability throughout even when the underlying network is lightly used. Further
,another major challenge network carriers have to face is how to smoothly
migrate from the existing network infrastructure to NFV-based solutions.

This paper describes two use cases, including their benefits and challenges.

a. Virtualization of mobile core network

Benefits-It enables flexible distribution of hardware resources to remove any

performance bottlenecks and rapid launch of innovative services. Challenges-
SDN could enable the service chaining of various components in cellular core
networks. However, as SDN has primarily focused on data centre networking in
the past, it’s still not clear how existing SDN controllers perform in the wide
area in terms of scalability and manageability.

b. Virtualization of home network

Benefits-It reduces the operating expenses by avoiding the constant

maintenance and updating of the CPE devices, and reducing the call centre and
product return burdens. Furthermore , it improves the quality of experience by
offering unlimited storage capacity and enabling access to all services and
allows content sharing from different locations and multiple devices like
smartphones, tablets, etc.
Challenges-The biggest challenge in this area is performance of virtualized
packet processing on standard high volume servers. Also, the security and
privacy issue is also another research challenge.

Some of the other challenges in NFV are -

 Network Performance of VNF

 Placement of Virtual Appliances
 Instantiation and migration of virtual appliances
 VNF outsourcing .

[3] IEEE 2015 ,Network Service Chaining Challenges for VNF Outsourcing in
Network Function Virtualization.

This paper’s primary objective is to elaborate challenges of network service

chain under VNF outsourcing ,fundamentally .NFV allows VNFs to get
outsourced by a third party because it separates network functions from their
locations. But ,VNF outsourcing raises significant challenges.

NFV alters the way that the networks are built and are operated by leveraging
standard IT virtualization technology and consolidating network equipment
types onto commodity servers. There are other challenges that NFV brings like
network performance of virtualization, the placement, instantiation and
migration of virtual appliances .In this paper, outsourcing of virtual network
functions (VNFs) is of primary importance.

This paper illustrates the NFV architectural framework and also summarizes the
service chaining technology, it describes the deployment of composite
networking services and approach to network service provisioning.

In network service chaining challenges in VNF outsourcing -

a. Maintaining multiple network service chains per traffic flow : When VNFs
are outsourced to third parties, the single network service chain must be divided
into a series of small network service chains.

b. Network Service Chain Dependency Management : When NSPs outsource

VNFs ,they maintain multiple network service chains to be applied to the same
traffic flow ,mutual dependence between those network service chains must also
be defined.
c. Network Service Chain Outgoing Point Management : When we consider
VNFs are created in an infrastructure domain managed by an NSP, outgoing
traffic flow on end of a network service chain is always forwarded to WAN in
order to arrive at their intended destinations. Even if we exploit VNF
outsourcing, there is more than one outgoing point to forward the traffic flow.

d. Establishing Data Plane between VNF Service Providers : There can be

multiple network service chains for a single traffic flow and thus the traffic flow
must be delivered between domains deploying network service chains.