3rd International Conference of EETT “Innovation in Broadband Networks & Services” – Part D

Day 2 – Session 2: Innovation in Networks: Next Generation Access Networks

Nicolas Curien, member of the Board of ARCEP gave interesting insight to ARCEP’s future regulation priorities regarding NGANs. If I got this straight, ARCEP is currently analyzing the French market to evaluate the best type of regulation for allowing access to the FT’s ducts (civil engineering). Another important issue is the form of access & entry of fiber/operators to the buildings (fears of monopolizing access to the buildings by the first entrant are expressed).

French are separating copper/traditional network and fiber networks. This is evident on the regulation approach followed regarding these networks. They are working on an asymmetric regulation regime for FT’s ducts (inherited from the traditional monopoly) and on a symmetric regulation regime for ALL operators that are building fiber access (the last mile).

Crister Mattsson, Ericsson Senior Advisor, former Market and Information Director of Stokab (City Network of Stockholm) presented the Stokab’s strategy (interesting input for municipal broadband network owners – i.e. local governments – in Greece that discuss operational structure and pricing strategies). In a nutshell:

• Provision of dark fiber BELOW operator’s cost
• Network neutrality in all respects
• Open Access in all respects
• Ease customer for provider selection

He also indicated that fiber offerings in Sweden are cheaper than ADSL and cable alternatives! He also explained the business model of Stokab where a Communications Op is operating below ISPs/SPs layer and above Network owners in the network value chain, thus ensuring open access on the dark fiber and the service level. Comm Op revenues are down streamed to network owners and equipment vendors.

Wolfang Fischer, Business Development Manager, CISCO presented the case of point to point Ethernet, its benefits in respect of deployment cost and open access regulation [see previous post].

Julio Villalobos, Managing Director, SVP Advisors stated that rushing into NGAN investments may not be the wiser thing to do as many uncertainties are yet to be resolved (technological choices, costs, future access and core network requirements). These uncertainties provide a potential advantage to being a follower. Whatsoever, on a macro perspective, countries that have successfully promoted NGAN have done in only with the cooperation and strong support of either central government (Japan, Korea) or local governments (Netherlands, Sweden).

Peter Alexiadis, Partner, Gibson, Dunn& Crutcher LLP summarized all available options and proposals for regulation measures and remedies. His speech main topics were:

• NGA is likely to lead to re-monopolization. (Technology choice for instance – GPON – may lead to monopolizing users/buildings/cities/regions making the network sharing difficult).
• A good part of NRA job is to seek investments but a clear regulatory framework is required. The often proposed “regulatory holidays” is not the answer. Investment takers value more the clarity of regulation framework than any other (temporary) kind of regulation.
• We should promote both service competition and infrastructure competition. Service competition can assist operators to provide additional value to end-users but retail prices drop, massive investments required for NGA has lead many other countries to alternative infrastructure models.
• We need to rethink about wholesale pricing in the case of NGAs. NRA might want to give greater flexibility on wholesale pricing to the incumbent. Incumbents will need that pricing strategy to experiment with what can be done with the products.
• NGANs will not require less regulation. It will probably require more intense and smarter regulation. The cost characteristics and some forms of technology choices of NGA hide the inevitable risk of customers’ monopolization by one provide. The monopolization may include buildings, cities and/or regions.
• Regulation process must start with duct sharing and follow up on sub-loop unbundling and then bit-stream obligations, in that order, or if possible the three altogether. Functional separation may be a condition at the end of the road.