3G Deployment Scenario

As mobile operators approach the evolution towards 3G, many are examining
the continued use of circuit-switched technology within their core networks.
Due to the current global economic recession, market uncertainties,
and what is today regarded as significant business risks involved with 3G,
most operators are trying to reduce and optimize their capital and operational
investment in their next generation networks. Even though a lot of
money has been invested in the 3G licenses across Europe (i.e., sunk cost),
hesitance is seen in quickly investing too much money in what is a relative new technology where the standardization is barely stabilized. It is very interesting
to realize that most 3G investments have been made before GPRS has
become a commercial success. This is due to the relatively late deployment of
commercial volume GPRS handsets/mobile stations and the delayed introduction
of mobile packet data into the market. This uncertainty alone puts a
tremendous risk in all 3G business plans and should drive the 3G deployment
scenario.
The 3G deployment scenarios are driven by two important assumptions:
• There will be a significant shift from voice-centric to data- and
multimedia-centric services, which make up the traffic mix within
these networks. This is illustrated by the current growth of SMS and
expected growth in Wireless Application Protocol (WAP) usage in
conjunction with GPRS.
• Operators can earn a profit selling 3G services.
Thus, it is clear that current radio access and circuit CN architectures
and technologies do not provide an appropriate and efficient infrastructure for the delivery of bursty packet-based data services such as the Internet, mobile
commerce (m-commerce), and corporate virtual private networks (VPNs).
To date, some mobile operators have created portfolios of traditional
and next-generation services by building multiple networks. It is common to
find a mobile operator using a time division multiplex (TDM) network to
support voice, an ATM or frame relay network to support GPRS, and an IP
network to support new features. Of course, using multiple network infrastructures
to support multiple services is costly. In addition, building services
that require combining diverse network technologies becomes exceptionally
difficult and tedious because they must be manually provisioned.
The use of a single homogenous network based on IP becomes the logical
choice for the delivery of seamless services within mobile networks. It
allows these services to span the voice, data, and video domains—thus
migrating the mobile network towards a true multimedia capability. The
migration from circuits to packets is achieved by migrating all of the services
and applications within the mobile network onto packet based (IP or
IP+ATM) network. This is most logically achieved from the core network
with migration outwards towards the RAN.
In the UMTS network architecture, ATM adaptation layer
(AAL)2-AAL5/asynchronous transfer mode (ATM) (R99), or IP (R00) are
used. Today it is not easy to determine at what point in time the all-IP model
will be introduced. Furthermore, it is likely that R3/4 and R5 will coexist in a
large number of mobile operators’ networks for quite some time. This is one
of the reasons for deploying network architecture based on multiprotocol label
switching (MPLS), as both IP and ATM are perfectly supported. Figure 2.8
shows a possible network deployment scenario for a 2G mobile operator.
With MPLS it is possible to utilize a single physical infrastructure, switches,
and transmission links, for both types of traffic: native ATM and IP. Network
resource allocation, namely switching capacities and transmission bandwidths,
is controlled by a few network management commands. No forklift
upgrades will be necessary when gradually moving from an R3 to R5 architecture.
This will protect the initial investments of mobile operators and give
them total flexibility for the introduction of new services.