background image
deploy femtocells with relatively low up-front
costs, as well as cause a low level of disruption to
the existing operations of an operator's core and
macro networks.
3. Ability to Scale to Support Mass-Market
Deployments ­ The approach must be able to
cost-effectively scale to support the deployment
of hundreds of thousands (if not millions) of
femtocells.
4. Rapid Path to Standardization ­ The approach
must allow for a rapid path to standardization
for the interface between femtocells and the core
network in order to support a robust femtocell
market.
"rnc"-BASeD SoLUTionS
The earliest
femtocell solutions looked to leverage the existing 3GPP
defined Iu-b interface that exists between 3G Radio
Network Controller (RNCs) and 3G radio antennas (Node
Bs). Primarily proposed by RNC vendors, these solutions
allowed operators to leverage the same RNC to support
femtocells in addition to macro network NodeBs.
As RNC-based solutions enable operators to leverage
their existing core networks through standard interfaces
(Iu-CS and Iu-PS), they meet the operator requirement for
full-service transparency, as well as the requirement for
low initial deployment cost and network disruption. As a
result, these solutions were initially viewed positively.
However, they eventually fell out of favor as they were
not able to cost-effectively scale. The challenge with
scaling this approach is in the basic design of RNCs, which
are typically optimized to support a relatively low number
of very high-capacity macro base stations.
Unfortunately, femtocells place the exact opposite design
requirement on a network controller, which is support for
a very large number of low-capacity base stations.
In addition, while 3GPP does define the Iu-b interface,
it is largely a proprietary interface specific to each RAN
vendor. As a result, there has been hesitancy to fully
standardize the Iu-b interface in order to open up the
femtocell market.
SiP/MSc-BASeD SoLUTionS
One alternative
proposal for femtocell device-to-core network connectivity
is to use a new SIP-based protocol between the mobile core
network and the femtocells. Operators would deploy a new
SIP-based MSC (mobile switching center) that operates
in parallel with their existing circuit and packet-based
core network. When a handset is connected to a femtocell,
it receives all of its services from the new SIP-MSC core
network.
As many operators believe they will eventually
transition their core networks toward an IMS and SIP-
based infrastructure, these solutions were also initially
viewed positively. SIP-based MSC approaches also held
the promise of cost-effective support for large-scale
deployments.
However, as handsets are served by a different core
network when connected to femtocells versus when they
are on the macro radio network, service transparency
becomes a significant challenge for SIP-based MSC
solutions. For every service an operator makes available on
its macro network, it now needs to replicate that service
on its new SIP-based MSC network.
Further, as the SIP-based MSC approach requires
operators to acquire and integrate a new core service
network, the initial deployment costs and overall
SPRING 2008
UMA ToDAY
09
Figure 2: Comparison of Device-to-Core Network Connectivity Approaches
Service Transparency
Excellent
Excellent
Poor
Initial cost/Disruption
Good
Excellent
Poor
Ability to Scale
Poor
Excellent
Excellent
Path to Standards
Poor
Excellent
Good
RNc
RAN Gateway
SIP/MSc