Real broadband real growth Dirk van der Woude

Real broadband, real growth Dirk van der Woude Ontwikkelingsbedrijf Gemeente Amsterdam

2 Effect of infrastructure… § …like that of the British steam railways… § 1838 Netherlands: choice between speed barges or steam…

3 Fiber?

4 And Wi. Fi was important too… 1792 Lille => Paris: • 15 stations • 36 characters in 32 minutes • all records broken, huge success • And up to 1848 cause for the French to forgo investing into a copper telegraphe network

http: //www. generation-nt. com/actualites/25551/carte-fibre-optique-paris-couverture/ 5

6 Each industrial revolution is underpinned by new infrastructure 1971 1908 1875 1829 RAILWAYS, PENNY POST AND TELEGRAPH 1771 Source: Carlota Pérez

7 Old shoes are nice (some are quite old…) 1971 1908 Optical chip throughput: 1, 6 Tbs (april 2006) Fiber speed: 8 Gb/s over 30, 000 km (march 2006) First trans atlantic fiber (1988), First usable fiber optic cable (1970) First TV Transmission through coax Berlin Games (1934); First pilot in USA, AT&T (1936) 1875 Coaxial cable patented in Germany 1829 First demonstration of electric telephone Antonio Santi Giuseppe Meucci (Havana 1849, New. York 1854) 1771 Semaphore Ernst Werner von Siemens (1884) (portable version: Napoleon Bonaparte, 1802)

8 Imagine a world without having the age of steam… § So let us be thankful for copper and coax! § however…

9 1999 - 2001: awareness Amsterdam § Can we be sure copper & cox networks are future proof – Practical speed – Parallel use (video) § Not able to sustain the city’s economic & social needs – Like continuing strong old and new media & ICT sector – Like citizens overwhelming web use (2006: > 85%) § “Living at a dirt road? Don’t buy a Ferrari” – Services follow infrastructure – Creative sector & citizens will find out use – Local government facilitates § Time for new network?

10 Amsterdam – employment (total 412. 000 working persons)

11 Amsterdam internet exchange: growth Worldwide peak speed records: July ‘ 05: > 50 Gb/s October ’ 05: > 100 Gb/s May ’ 06: > 154 Gb/s June’ 07: > 282 Gb/s

12 • And of course there is this ‘contention’ thing… http: //www. lightreading. com/document. asp? doc_id=93103&page_number=4 VDSL 2 – laboratory conditions

13 x. DSL 8 - 14 Mb/s down, 0. 7 – 0. 8 Mb/s up Leaves for web use: ? Mb Two phones: 2 x 0, 2 Mb IP TV SD low = 3 to 4 Mb SD normal = 4 to 6 Mb HD low = 8 to 12 Mb HD = 18 to 20 Mb So: • Better live next to the central office… • And of course still there is this ‘contention’ thing… • Parallel speed? Ams-IX per sub: UPC 40 kbits Essent: 60 kbits XS 4 All: 67 kbits http: //www. ams-ix. net/connected/? expanded=1 Shared capacity…

14 First user test of Amsterdam Ftt. H • Advertised speed: 100 Mb up, 100 Mb down • In practice (first days): 88 Mb down

15 Japan: 270, 000 new Ftt. H… per month

16 Oct 1, 2008

17 Ftt. H, elsewhere in Europe just a pick (1) § France, the battle for & around Paris – Iliad: 1 million Ftt. H at 1 billion Euro investment (City of Paris supporting) – Neuf Cegetel: in 2009 we want 250, 000 Ftt. H (50 Mbits) at euro 29, 90 – France Telecom, 2007: Ftt. H (100/10 Mbits at euro 44, 90) in Paris, Poitiers, Marseille, Lille, Toulouse & Lyon. Target: 200, 000 subs end 2008 – Noos Numericable, march 2007: “Massive investment towards Ftt. H” – over 100 broadband projects France with communal participation – Nov. 2006, French government: 4 million Ftt. H in 2012 § Germany: competitive teleco’s announcing Ftt. H – f. e. Net. Cologne: all of Cologne, to be followed by Bonn, Aachen? (Net. Cologne = 100% GEW Köln AG = 100% City of Cologne) – Other projects in Schwerte, Norderstedt, Hamburg, Gelsenkirchen, Dessau, Magdeburg

18 Ftt. H, elsewhere in Europe just a pick (2) § Vienna, Zürich (muni energy corp’s): – Ftt. H in whole city § Norway: Hafslund energi (53% owned by City of Oslo) – Ftt. H to half of Norwegian population § Sweden: – 200 of 289 communities own a fiber network § Denmark: energy corps doing Ftt. H – 2008 – 2009 to 35% of homes (= 50% of population) § UK: – Oxfordshire project, part Ftt. H (? ), part Fiber/VDSL – Rest of UK, according to BT & DSG: “ 12/1 Mb is enough for all & ever. ” “No fiber please, we ‘re British” – However, Ofcom report march 2007: “Last mile will have to be fiber”

19 Hauts-de-Seine: 700, 000 connections § Western part of Greater Paris – pop. 1. 5 million, 100. 000 SME’s, 880. 000 jobs, 85% in services – Per capita highest income of France § 2005, adoption of proposal by the Chairman of the Conseil General – Ftt. H to all population and companies – With a maximum subsidy of euro 70 million § Said Sarkozy: “Copper is not gonna cut it, we need fiber”

20

21 Ftt. H in the EU, some examples Köln Vienna Paris, Iliad Net. Cologne 200, 000 Ftt. H 1 million Ftt. H Hauts-de. Seine Milan Stockholm Amsterdam Ftt. H Ftt. B Dark fiber 40, 000 Ftt. H Up to 70 million 100 million 6 million of 30 million (passive layer) n. a. No Yes n. a. Municipal financial participation Open network? 250 million No n. a. Aim: yes 75 to 100 million worth of support Yes? subsidy Problem for EU?

22 Three layers, three types of investors 40, 000 homes passed now – more than 450, 000 later on Service providers 100% market Wholesale transmission provider 100% market Passive access network provider 20% municipal consumer/ SME Rent

23 Three kinds of financial dynamics 40, 000 homes passed now – more than 450, 000 later on Services High Op. Ex, Low Cap. Ex Quickly profitable Transmission Attractive Op. Ex, acceptable Cap. Ex – profitable in few years Passive network Real estate like investment (Highly? ) Profitable on long term

24 Translated into investments (millions of euro)

25 Fiber-from-the-Home 40, 000 meter boxes, 10% of Amsterdam Boroughs of Zeeburg (100%), Oost (part) & Osdorp (part)

26 Architecture (1) § three-layer model • Passive fibre infrastructure: Point-to-Point • Unbundled local loop of fiber = maximum competition at upstream level in value chain • Largest capacity for future growth • Active layer: Active Ethernet • Applications services layer, Service providers are being offered transparent access: • with discrete virtual LANs (VLANs) for each service on a per user basis • allowing multiple services to be delivered and invoiced to each home in parallel (i. e. multiple ISP’s, Citywide Intranet, closed circuit IP-based surveillance, IP-TV, care and medical services etc. )

27 Architecture (2) § Open network – Concession of 8 years for operator/investor, after that more operators possible – Operator is non discriminatory wholesale seller of capacity § Why not x-PON – Avert risk of having to (expensive) re-digging – Labor costs dominate, will rise with inflation – Short distances, so savings on cost of material (fiber) are small – No lock-in of equipment supplier which stifles innovation – Impossible to have different technology/supplier per subscriber line

www. glasvezelamsterdam. nl & www. citynet. nl Vragen?