Advanced Network
In order to meet the 5G key requirements such
as latency and the large number of simultaneous
connections, and to support new business models
and scenarios for operators, technologies above
the radio access level should also be considered.
Therefore, we need to develop technologies at the
system architecture level from the network point of
view. To accomplish the 5G key requirements and
support the increased data rate of new 5G radio
access, 5G network will have to evolve towards a
distributed and flat architecture.
Current mobile network architectures designate a
dedicated node in the core network (e.g., PGW -
Packet Data Network Gateway in 3GPP) as a mobility
anchor that allocates an IP address to the terminal,
tracks terminal location in the IP topology, and
ensures terminal’s reachability by tunneling its traffic
to wherever it goes. All terminal traffic is tunneled
through the centralized node in the mobile core network.
However, the undesirable consequences of
this design include the following:
- Increase in end-to-end transmission latency
due to elongated data paths.
- Additional load of backhauling and network
processing in the core networks.
- Low network reliability due to introduction of a
single point of failures.
In the 5G flat network architecture, as illustrated in
Figure 19, user mobility is managed efficiently and
in a dynamically scalable fashion by pushing the
functionality to the edges of the network and even
onto the mobile terminals [20].
The three important benefits of this approach include
the followings:
First, such a distributed mobility management always
provides the shortest data path between MS
and the Internet without traversing the core network.
This distributed mobility management leads to a
significant reduction of signaling and data transmission
delay. Also, low end-to-end latency requirements
of ‘less than 5 ms’ for new 5G services such
as immersive UHD video streaming, cloud gaming,
and virtual reality, cannot be archived solely by reducing
radio access latency but would also require
a flatter network architecture design. In the flat network
architecture, services which require low latency
transmission can be provided by Edge Servers
and they can benefit from advanced network features
which utilizes network information for optimal
operations.
Second, it provides a highly scalable solution compared
to the centralized architecture, in which a
single core network gateway maintains the whole
traffic from MSs or to MSs.
Third, it easily avoids the risk of having a single point
of failure. In flat network architecture, the breakdown
of one network gateway would not significantly interfere
with the operations of the other gateways.
Flexibility is considered as another key requirement
of 5G network architecture. Software-Defined Networking
(SDN) and Network Function Virtualization
(NFV) provide promising examples of programmable
design technologies for realizing a flexible 5G
network architecture.