Fiber or Radio? Yes!

Proponents of some level of public investment in telecom infrastructure in the US are at risk of tying themselves in knots over the non-issue of whether radio or fiber technology should be used.  The very short answer is that both are needed. In many cases radio is by far the faster and cheaper way to provide the last mile of Internet access and that only radio supports roaming while only fiber (of today’s alternatives) is appropriate for high capacity backbone.

San Francisco has an ambitious plan to provide WiFi citywide.  Some people have said “Wait.  We should be doing fiber instead.”  Vermont Governor Jim Douglas has proposed that Vermont become the first e-state by leveraging the proceeds from a $40 million dollar bond issue into true  “anywhere and everywhere” statewide fixed and roaming broadband and cellular coverage by putting much of the money into towers on which the state will lease space to cellular and broadband providers for radios.

In Vermont, Chittenden Community Television (dba The Center for Media and Democracy) said:  “It is our opinion (the Center for Media & Democracy and other Vermont public telecom advocates) that Vermont should use the GTA [their name for the bonding authority the Governor proposed]  to leverage the funds that will be necessary to build a public fiber backbone—like the highway system—capable of true broadband speeds (NOT DSL speeds of 786Kbps BUT 100 Mbps) that is OPEN TO ALL: Comcast, Verizon, FairPoint, SoverNet or the ACME internet company. We urge the state to build public fiber backbone serves [sic] the communities of Vermont. Model it on the open networks of Burlington Telecom (serving Burlington but capable of reaching across the state) or Northlink (built to serve the northern and northwest counties of Vermont by the Northern Vermont Economic Development Council).”

CCTV seems to be confused about the difference between a fiber backbone and fiber to the home. A fiber backbone IS essential to providing excellent Internet service in Vermont or San Francisco or anywhere else.  There is a lot of backbone fiber in Vermont but it doesn’t go everywhere it needs to.  North-Link (cited above) is one of a number of publicly funded projects looking to extend fiber into underserved regions of the State; but North-Link does NOT propose to bring fiber to the home.  From their website: “North-Link is a project that will provide backbone service to northern Vermont through fiber optic cable. Smaller private firms and public entities will then be able to provide wired last mile connections or wireless services off of the North-Link infrastructure…”

As I understand it, what the Governor has proposed is completely complimentary to this effort.  The proposed towers need to be connected to a fiber backbone.  They are also customers for a fiber backbone.  There is a much higher likelihood that small providers will deliver the wireless last mile connections envisioned by North-Link if they can rent space on existing towers than if they have to raise the capital to build towers.  The backbone fiber is needed to bring the Internet to the neighborhood; the towers are needed to help get it the last mile to businesses and residences and provide roaming capability.

Burlington Telecom (also cited above) HAS run fiber to many of the residences in Vermont’s largest city.  It plans to be citywide soon.  They’ve done a very good job providing an alternative to Adelphia cable (now Comcast) and Verizon who were not investing in this kind of infrastructure for Burlington.  BT provides not only Internet access but also telephone and the equivalent of cable TV service.  Although BT is open to having other providers sell services over its network, so far none have come forward to do that.

So why not just replicate the Burlington Telecom model statewide in Vermont rather than build radio towers?  Why not just bring fiber to every residence in San Francisco and forget this WiFi stuff?

Four main reasons:  Speed of deployment, support for roaming, cost, and need for competition.

Digging trenches and stringing fiber on poles takes a long time compared to building towers and putting radios on them.  In San Francisco there is the problem of digging up city streets.  In Vermont there is both a relatively high cost of construction per household and a need for environmental protection.  The Examiner.com quotes fifteen years as the time required for a full fiber rollout in SF (which I think is overly pessimistic). Common estimates are that the WiFi project proposed by Google could be done in a year if the City Council would ever stop arguing over it.

Fiber will be built to many premises in Vermont and in San Francisco.  Telling people that they shouldn’t have any service (the case for too many Vermonters) until they can have 100 meg fiber is the equivalent of telling them to “eat cake”.

But fiber, even when it comes, will not provide roaming service.  Unless we’re all going to wear fiber leashes, we will continue to get roaming access in our cars and on our streets and in the country by radio.  The towers for radio transmitters may well gain in value as fiber connections below them make connectivity more and more essential to the life of the community.  No real conflict between fiber and radio here.

Again from Examiner.com, the expected cost to bring fiber to every residence and business in the city is $560 million.  The San Francisco Chronicle estimates the cost of WiFi covering all San Francisco at $12 million. These are very different services but it’s a lot easier to come up with $12 million than $560 million.

But what about the speed issue that CCTV raises?  Well, my wireless ISP provides me with 3 meg up and down for $39.95/month.  That’s a much better speed than basic DSL which IS becoming obsolete.  At the moment, the broadest residential bandwidth offered by Burlington Telecom for Internet access is 5 meg in both directions for $40/month. Radio is on the verge of getting much faster; BT could offer more bandwidth today if it saw the demand for it.  Point is radio can easily offer people the bandwidth they need today and tomorrow; fiber may eventually supplant radio on many last miles but that’s no reason to leave these last miles unserved.

According to its filing with the Vermont Public Service Board,  Burlington Telecom obtained commitments of almost $20 million in lease financing to buildout Burlington.  This includes a few million to build out the headend which would not have to be replicated if BT expands to other locations.  On the other hand, Burlington is much more thickly settled than most of Vermont.  So far this expenditure has NOT resulted in private risk capital coming in to build services on top of this infrastructure.

There are many precedents in the US for different providers sharing and paying for space on towers owned by a third party and making their own investment in radios to put on those towers.  I’m not aware of any working instances in this country of last mile fiber-sharing.  In Burlington BT provides needed competition to Comcast and Verizon (and vice versa).  In the areas which are still unserved in Vermont, competition will develop more quickly on multi-provider towers than on a fiber network to each residence owned by a single entity – no matter how benign that entity may be.

I’m a fan both of Burlington Telecom and of  Jim Douglas’ e-state proposal.  There’s a proposal before the Burlington City Council for a committee to help determine whether Burlington Telecom should be allowed to offer services to other locations in Vermont.  I hope the council approves the committee and the committee answers “yes”.  And I hope the Vermont legislature approves the e-state plan and funding so Vermonters everywhere in the state can quickly have the connectivity at home and roaming access they need.  As a Vermonter, I have no suggestion for San Francisco.