|
Andy Berquist - Chairman, Magenta Netlogic
Partial Private Circuits
Hidden Benefits of Local Loop Unbundling
The unbundling of local loop networks is a significant step towards a
more efficient communications market. But the resultant competition in DSL
is not the only benefit of this progress. Andy Berquist investigates a new
form of interconnect tariff emerging from Europe's great unbundling.
With much of the hype surrounding Local
Loop Unbundling focused on DSL services, the advantages that carriers and
service providers can achieve through use of Partial Private Circuits in
Europe is receiving relatively little press. For connections from customer
sites to an Other Licensed Operator's (OLO's) Point of Presence, Partial
Private Circuits offer service providers potentially cost-effective methods
to extend their national networks through use of the national incumbent's
networks. Where local access circuits are typically between 25% and 50% of
the total bid price in Europe for ISP's and network providers, use of PPCs
can deliver a very welcome benefit.
As shown in ECTA's Leased Line Scorecard
this month, significant cost savings are already a reality in a number of
European countries. Partial Private Circuit tariffs that have been added to
interconnect tariff offerings by British Telecom, France Telecom, Telecom
Italia, Belgacom, and Telefonica have brought carrier access circuit prices
in these countries down to some of the most competitive rates in Europe.
Table
1: Leased Line Scorecard - ECTA/Magenta netLogic
Interconnect vs. Retail - 2 Mbps Monthly Recurring Charges (April 2002)
What
are Partial Private Circuits?
Partial Private Circuits (PPCs) are connections from a third party,
typically a customer site of a OLO, to the service provider's Point of
Presence (PoP). Most European carriers have offered interconnect tariffs
that connect OLO's to the incumbent carriers for some time. Partial Private
Circuits now support connections from the OLO direct to third party or
customer sites, so that OLOs can take advantage of the reduced rates of
wholesale-level interconnect tariffs instead of having to use standard
retail-level leased line tariffs for customer site circuits.
From the PoP, the PPC's use the OLO's
network to complete the connection to the Internet or other offices through
networks such as ATM, Frame Relay, IP/VPN. The structure of PPC's vary
between incumbents for tariffing purposes.
Figure 1 shows the structure employed by
British Telecom which mirrors the actual PPC network architecture.
Figure 1: British Telecom's Partial
Private Circuit Configuration
The elements of a Partial Private Circuit include:
Third Party Link: Connects the OLO's customer to the incumbent's
serving exchange. Typically this is the local exchange to the customer's
site or the nearest exchange with interconnect capability.
Main Link: Connects the incumbent carrier's serving exchange
to the carrier's serving node.
Handover Link: The handover link connects the carrier's serving
node to OLO's PoP. If the OLO's has facilities installed to the Serving Node,
the handover link can be provided by a comparatively in-expensive
"In-span Handover". Alternatively, a "Customer Sited
Handover" can be used that connects to the OLO's premises.
Potential Cost Advantages of Partial
Private Circuits
The key to gaining cost advantages with
Partial Private Circuits is with the main link and the handover link and
the equipment at the Serving Nodes and PoPs. If the OLO purchases retail
links, they must purchase individual end-to-end links to each customer
site. PPC's allow consolidation of customer circuits within the incumbent's
network to the Serving Nodes. The main link extends the OLO's reach to the
incumbent carrier's local exchanges. There are opportunities for main link
rates to be reduced as they are part of bulk or wholesale commitments made
by the OLO. The Handover Link is typically a high-capacity link for
handling several customer's access circuits, e.g. starting at STM-1/155
Mbps, which attracts reduced rates when compared to purchasing multiple 2
Mbps circuits. If equipment is shared at the Serving Node and PoP, further
cost savings can be achieved. Note that the Third Party Link connects to
the incumbent's local exchange and typically attracts standard retail
rates.
The end result with using PPC's is that
the incumbent carrier's Serving Nodes can be considered as virtual PoPs for
the OLO - carriers outside the country, ISP's and network providers - see
Figure 2.
With this configuration, the handover links can now be considered to be
part of the OLO's network where the costs can be incorporated in their
network service charges, such as for access to the Internet or usage of
Frame Relay, ATM, IP/VPN services. As the network charges are shared across
the service provider's customers, the local access charges for each
customer can be greatly reduced.
Figure 2: Extending Reach via Partial
Private Circuits
Realising Cost Advantages of Partial
Private Circuits
The bad news is that these potential cost
advantages of shared links has not been realised with all of the initial
offerings due to high initial equipment charges and the inability to share
equipment among several OLO's. Sonia Hilton of KPNQwest notes that it is
true that BT's annual rental charges for PPCs are lower than the
interconnect prices set out in the European Commission's benchmarks.
However, BT has employed an 'element-based' charging structure that
requires an up-front payment of £134,000 for starting a level SMA-4
multiplexer required at each PoP, which can go as high as £198,000 for an
SMA-16. These charges are for the multiplexer 'carcass', where additional
connection cards are needed at a cost of £2,000-£8,000 per card for the
ability to share equipment. Annual rental rates of £2,500 for the SMA-16,
£1,600 for the SMA-4, and £50-£250 for the tributary cards add to the
overall charge. At the customer end, multiplexers are also required, which
range from £3,000 for a slot in an existing 16x2 mux to similar charges for
the SMA-4/16 at the PoPs.
Sonia says that these up-front PPC
charges are anti-competitive and represent a significant investment
required at entry level. The best method to resolve this situation is to
allow the multiplexers to be shared between operators. BT does not allow
this at present and sharing between retail and PPC circuits of the same
operator. There are further 'deal killers', including BT's failure to offer
PPC equivalents of all retail products and no retail equivalent SLA
commitment level.
On the basis of all this, Hilton and
KPNQwest would like to see Europe avoid the 'element-based' approach to PPC
tariffing. Instead, she would encourage the adoption of a traditional voice
interconnect 'service-based' model whereby charges are based on recovery of
costs for services provided.
Reviewing the new PPC offerings in other European countries, it appears
that this service-based model is being followed, although the details are
just emerging. If this is the case, as European incumbent carriers rollout
Partial Private Circuits, there should be a steady reduction in the costs
for accessing pan-European and international networks, which should have a
knock-on effect of improving in the cost-effectiveness of advanced network
services such as high-speed Internet Access, ATM, Frame Relay, IP/VPN.
Based on the initial experiences, though, some rationalisation for PPC
charging is needed for PPCs to be a major factor in future carrier
networks. Future versions of ECTA's Leased Line Scorecard are being
designed to track this promising area for improvements in our beleaguered
industry.
top
|
