Nguyen, Brunstrom, Grinnemo, and Taheri
46
Although NFV has been proven as the key enabler for the development of 5G,
especially the core part, there are many challenges that need to be further studied in
future work, such as optimization of network functions placement, resource allocation,
management and orchestration, and network performance [29].
2.3.2.3 Cloud Computing
As described in the previous section, cloud computing has been considered as an ideal
solution for re‐designing the current RAN architecture. With its anticipated benefits.
such as on‐demand and elastic provisioning of services and resources over the Internet,
cloud computing has made it as one of the key enablers for designing 5G core networks.
In this case, 5G core network functions will be realized as virtual machines or containers
controlled by the cloud manager. The capability of providing resources in a multi‐tenant
model of cloud computing allows mobile operators to implement the concept of mobile
virtual network operators (MVNO) much more easily than in the past. In addition, a
pay‐as‐you‐use business model and the ability to move and consolidate the resources
offered by cloud computing. can help the mobile operators reduce their capital and optimize operational expenses. Compared to NFV, cloud computing was born to virtualize
the commodity IT hardware. while NFV refers to the inspiration of cloud computing to
virtualize network functions. In fact, in the recent development of NFV, many cloud technologies such as OpenStack
1
or VMware
2
are serving as the resource backend for virtual
network functions.
Centralizing all resources will ease the management and provisioning process. but it
will result in the long delay for end‐to‐end communication, which may not be suitable
for some of the newly defined 5G services. Therefore, combining cloud computing and
other computing paradigms, such as mobile cloud computing, fog and edge computing,
will be a promising direction to investigate in future research.
2.3.3 5G End‐to‐End System
The key technology enablers for constructing a 5G end‐to‐end system include network
slicing, and management and orchestration.
2.3.3.1 Network Slicing
Today’s 4G system have been optimized mostly for serving human‐to‐human communication where mobile phones are the main players. However, in the future,
the 5G system is expected to support diverse services and applications with various
characteristics and requirements, where IoT devices will become dominant. Such IoT‐
related services will require different types of features and network capabilities in terms
of latency, data rate, mobility, reliability, security, etc. Therefore, in order to guarantee
these requirements and improve the network performance and resource utilization,
each service type should be provided as an end‐to‐end, isolated, and infrastructural
environment to operate on. In this sense, the network slicing concept will become the
foundation of all capabilities.
1 OpenStack. https://www.openstack.org/
2 VMware. http://www.vmware.com/
46
Although NFV has been proven as the key enabler for the development of 5G,
especially the core part, there are many challenges that need to be further studied in
future work, such as optimization of network functions placement, resource allocation,
management and orchestration, and network performance [29].
2.3.2.3 Cloud Computing
As described in the previous section, cloud computing has been considered as an ideal
solution for re‐designing the current RAN architecture. With its anticipated benefits.
such as on‐demand and elastic provisioning of services and resources over the Internet,
cloud computing has made it as one of the key enablers for designing 5G core networks.
In this case, 5G core network functions will be realized as virtual machines or containers
controlled by the cloud manager. The capability of providing resources in a multi‐tenant
model of cloud computing allows mobile operators to implement the concept of mobile
virtual network operators (MVNO) much more easily than in the past. In addition, a
pay‐as‐you‐use business model and the ability to move and consolidate the resources
offered by cloud computing. can help the mobile operators reduce their capital and optimize operational expenses. Compared to NFV, cloud computing was born to virtualize
the commodity IT hardware. while NFV refers to the inspiration of cloud computing to
virtualize network functions. In fact, in the recent development of NFV, many cloud technologies such as OpenStack
1
or VMware
2
are serving as the resource backend for virtual
network functions.
Centralizing all resources will ease the management and provisioning process. but it
will result in the long delay for end‐to‐end communication, which may not be suitable
for some of the newly defined 5G services. Therefore, combining cloud computing and
other computing paradigms, such as mobile cloud computing, fog and edge computing,
will be a promising direction to investigate in future research.
2.3.3 5G End‐to‐End System
The key technology enablers for constructing a 5G end‐to‐end system include network
slicing, and management and orchestration.
2.3.3.1 Network Slicing
Today’s 4G system have been optimized mostly for serving human‐to‐human communication where mobile phones are the main players. However, in the future,
the 5G system is expected to support diverse services and applications with various
characteristics and requirements, where IoT devices will become dominant. Such IoT‐
related services will require different types of features and network capabilities in terms
of latency, data rate, mobility, reliability, security, etc. Therefore, in order to guarantee
these requirements and improve the network performance and resource utilization,
each service type should be provided as an end‐to‐end, isolated, and infrastructural
environment to operate on. In this sense, the network slicing concept will become the
foundation of all capabilities.
1 OpenStack. https://www.openstack.org/
2 VMware. http://www.vmware.com/