Tuesday, March 24, 2009

The Government Solicits Broadband Policy Proposals...

During the presidential campaign, Barack Obama called for America to lead the world in internet access with "true broadband in every community... [and] better use of the nation's wireless spectrum."

How exactly this would be achieved has been a great mystery. As is always the case with telecommunications policy, the devil is in the details, and therein resides the controversy. What should be the role of government? Should government provide broadband directly to citizens, offer incentives to private companies to develop the infrastructure and extend access to underserved areas, or avoid interference and leave it up to the private sector entirely?

Interestingly, the government itself has yet to determine its preferred answer, and is seeking advice from - you guessed it - the public. As Wired reports,

The FCC announced Thursday that at its next open meeting, April 8, it will discuss a notice of inquiry (.pdf) — essentially an open call for comment on what the government's role, if any, should be in making the country richly wired and unwired. From there the FCC will draw out a broadband road-map report — due to Congress in less than a year under the rules of the stimulus package.

So if you were advising the president on how to make the United States the global leader in broadband internet access, what would you say? What would be your vision for America's 21st-century digital infrastructure?

Here are a few of my humble suggestions...

  1. Everyone who wants broadband internet, and is willing to pay for it, should be able to get it.

    Some of us living in major metropolitan areas forget that millions of Americans still have no choice but to use 56k dial-up connections, not because they're cheapskates, but because the infrastructure doesn't yet exist in many rural areas. This is simply unacceptable. Just as the government ensured with the telephone network a hundred years ago, broadband infrastructure must extend to every remote corner of the country. The Feds should offer subsidies to private telecom companies to build it, and if they refuse, then the government ought to do it themselves, accepting bids for contracts.

  2. Pursue municipal wireless strategies.

    Physical broadband infrastructure is expensive to build, which is why some type of public-private partnership is necessary. However, wireless infrastructure is light-years cheaper, and also less problematic operationally since, legally, the public already owns the airwaves. The Feds should offer financial assistance to municipalities that build viable Wi-Fi alternatives to their wired counterparts. Also, free Wi-Fi should be implemented in all public places - such as parks, libraries, government buildings, and, of course, schools. This is already occurring in places like Bryant Park and the New York City Public Library, but so much can be done to speed the process up and make it a universal principle. Basically, government should seek to expand wireless internet access through every avenue possible.

  3. Adopt IPv6.

    This is more on the technical side, but the current IP address system is projected to reach its capacity by 2011; and that's a major potential problem. A solution already exists, known as IPv6, however, there has been a paradoxical problem in rolling it out - Nobody wants to adopt IPv6 until it becomes widely supported, and administrators don't want to support it until it becomes more widely adopted. The Feds can give everyone a swift kick in the rear-end and get the ball rolling simply by using their procurement power and requiring that any recipients of stimulus money adopt native IPv6 support. No 21st-century infrastructure will be complete without it, and this is a great excuse to take the initiative.

  4. Keep broadband affordable.

    There's no point in making sure every American has broadband access if no one can afford it. I'm not suggesting that's the case now, however there is an omnipresent danger of pricing ourselves out of the game if this principle gets taken for granted. To improve on our existing system, there needs to be greater market competition. While nationally things seem rosy, the fact is that most of us have only one or two broadband providers in the community where we live, thus giving us no alternative but to pay whatever prices they ask, and giving those providers little motivation to innovate or improve their quality of service. That has to change.

    To reach this goal of increased market competition, (as an anonymous commenter suggests) the government should disallow infrastructure companies from providing services on their own network, or failing that, they should be required to offer identical access at identical prices to what they provide to themselves. Having a single company own/control the infrastructure as well as the services on that infrastructure means that it's against their business interests to cooperate with and compete fairly with competing services, and that vertical integration needs to stop.

  5. Increase funding for science and engineering research.

    Admittedly, I may be biased due to the fact that I would likely be a recipient of such funding, but the fact remains that if America wants to be a global leader in digital technology, it has to actually invest in technology research. This is a common sense principle. We don't want to be forced into a situation 30 years from now where we have to lease technology from India and China just to be competitive. Grants involving newer technologies, such as deploying fiber-based and WiMax infrastructure projects, could fall under this umbrella as well.

Thus, any vision of a 21st-century America ought to include being at the forefront of technological development, vibrant market competition, and widespread broadband penetration. While the private sector will rightfully be the primary agent for deployment, there is nevertheless an awful lot that the government can, and should, do. And there's no time like the present.


At 6:40 PM, Anonymous Anonymous said...

Switzerland runs four dark fibers to every home so that every service provider (voice, TV, Internet, security) must compete without monopoly. Japan relied on powercos and broadband over power lines (BPL) to connect rural subscribers and allowing any AC outlet to carry many megabits to every single device.

Natural monopolies such as power grids must be regulated anyway, so require powercos to both deploy and maintain at least four dark fibers to every town and let homeowners own the fibre "tails" connecting them to the hubs: Set one federal tariff to access them regardless of service mix offered, with a higher tariff and stringent clear rules to access "smart grids" (a very secure VPN) that monitor every AC device (G.hn, P1901).

This way, communications services subsidize rollout of power saving & resilience technologies (medical monitoring, etc.) and wireless transceivers can be put up at fibre endpoints. Initial deployment relying on WiFi/WiMax to connect smarter power meters minimizes fibre expenses & allows adoption of smarter home devices relying on sub-megabit connections right away - for less than $100/home.

Consumers can then upgrade seamlessly from fixed wireless to BPL to FTTH bandwidth to improve communications & displace legacy satellite, coaxial & PSTN connections with VoIP, very secure voice & VPNs running as peers to the powerco's own grid monitor system, more flexible VPNs running as peers to commercial cable TV & Internet offerings delivered on G.hn & fibre.

Any universal broadband strategy that does not build on the smart grid and existing power lines to provide its first hundred megabits into the home is necessarily slower than any strategy that does. We're seeing smart meters deployed anyway and they are perfectly capable of routing entry level broadband and high quality voice and many secure services that fiber can't support because it can't talk to AC devices universally.

So the fastest upgrade path is:

1. Dark fiber backbone exclusive to powercos supporting many wireless endpoints used to collect reports from smart meters and also to make rural fixed wireless ubiquitous. Require access for qualified secure services to do power conservation and medical monitoring and so on - put powerco share of this into fund to fund further dark fiber rollouts & improve power network monitoring. Keep this all under federal tariff.

Require new homes to have chargers for electric cars (which must be off peak) in garages - increasing demand for electric cars and saving the US auto industry (now committed to moving to electric cars quickly)

Require the federal government to actually meet its telework rules, in place since 2000, and support other companies to support more telework to cut commuting, traffic jams and gasoline use - increasing demand for bandwidth to more areas.

2. Open up multiple dark fibers to any service provider and require a powerco to use income and savings from smart meter services (the fund) to continue to expand back bone and security and resilience, and help consumers buy efficient and more controllable devices (as Berkeley CA's power utility does).

At this stage any communications or entertainment provider will be able to get high bandwidth services to hubs in every town. Homeowners can be allowed to own fiber "tails" to these hubs to access any service of interest, or public funds can be used to connect outlying areas. If incumbent providers make coaxial or their own fiber available for home owners or public agencies to buy, they should get very favorable tax and access incentives that make it more profitable for them to become commodity providers of services at the hubs rather than monopolies to the home. A mix of fixed wireless, DSL, coaxial, fiber can be used in this interim period to communicate, but BPL/G.hn/P1901 will be deployed universally to AC devices in homes.

Require new buildings to have at least two dark fiber "tails" that connect them to the nearest federal tariffed hub - owned as part of the building. Set a schedule for every building in the country to have at least shared fiber access to hubs that is not controlled by monopoly.

3. As broadband demand grows due to familiarity with services, simplify the network by expanding dark fiber backbones and improving robustness of the "smart grid" infrastructure so more services can be supported as peers to the powerco's own, and competition develops in monitoring, conservation, fire/security/medical and other services using AC outlets to connect & control home devices.

Federal tariffs can require lower costs of access and longer periods of amortization - allowing powercos to only charge for new rollouts and ensuring that communication revenue continues to subsidize the advanced smart grid services required to cut peak energy demand using demand side management (DSM) and subsidize rollout to most remote rural areas. This was the model that worked for universal power, phone, natural gas and so on.

Homeowners will continue to pay the cost of upgrading from existing copper or basic BPL level bandwidth to fiber - not subsidized by public funds at all. When they pay, they will own what they have paid for, and will be paying cost and easily able to pay and own collectively with neighbors on favorable terms.

4. Final phase: Set a sunset date by which all fiber and copper on any public rights of way will be owned by the public agencies that own fibre, and by which time every incumbent will be providing service to hubs only. Separate businesses can continue to run dark fiber or hooking up remote communities with collectively owned bundles of tails back to the nearest federal tariff hub but they cannot be controlled by any service provider. Also set a sunset date for control of smart grid services by the powerco itself as its monitoring needs become very clear and security challenges are elaborated. A public agency that is capable of managing the secure network connecting all AC devices to critical infrastructure standard and capable of making emergency power and communication rationing decisions should be in charge, but not until the medical, safety, fire and power monitoring systems are sufficiently mature to write its mandate. Powercos then return to specializing in power equipment maintenance and managing diverse sources on the supply side, and are not required to make complex trade offs between medical and resilient community and powercos' own needs.

Service markets by now will have developed so that persons who are medically vulnerable (or their insurance companies) have choices, and a wide variety of providers can be allowed access to AC outlets if they can meet the extremely strict standards set by the federal agency including web of trust signatures, universal challenge/response codes and other aspects of a very secure critical infrastructure.

The cost of upgrading all buildings to hard-to-tap fiber connections or swapping AC devices for lower-power powered Ethernet (PoE/IEEE 802.3at) DC devices can accordingly be paid for from military and "Homeland Security" budgets as they reduce vulnerability of communities to terrorist attacks, natural disaster & military scale Internet assaults.

Even simultaneous destruction of all the federally tariffed hubs and major power stations would not be able to take out VoIP over "smart grid" grade networks or LED light strands lighting up stairwells on DC PoE. Battery backup for a full 72 hours could keep every essential service running in every building, albeit without as much bandwidth for video, music, P2P. So attacks can cut off entertainment news for a few days (this will make a lot of people very mad and mobilize them!) but they can't prevent you from being able to reach 911 or fire services, or even 311 or 211 local municipal services. Inconvenience rather than disaster.

Meanwhile the economy will be so efficient, relying on telework and collaboration technology rather than separately lit, heated, powered office buildings and on jammed commuter routes wasting gas on idling, that this economy will be almost immune to recessions too.

At 10:58 AM, Blogger Robert J. Domanski said...

Wow, these are some very intelligent and detailed ideas. I believe the two most intriguing ones you raise are 1) to let the homeowners own the fiber tails connecting them to the hubs, for this will decentralize control over the infrastructure and give the consumer renewed ownership rights concerning things that actually reside inside their homes (this would also be easy to sell politically), and 2) deploying wireless transceivers at fiber endpoints would be a relatively cheap and fast way to upgrade to a faster network - most individuals would probably make the capital investment on their own. BPL (Broadband Over Power Lines) also has a lot of potential to become a viable alternative, however it has yet to gain much traction.

Thank you for taking the time and contributing some intelligence to the discussion :-)

At 11:14 AM, Anonymous Anonymous said...

Here's some BPL trials in the US that highlight the advantages over any other option:


Here's a blog that focuses on where IT meets energy. Notice Cisco and IBM have committed to the "smart grid"/Internet fusion:


(St. Arnaud is also a good source on the "homes with tails" model).

And here's another big company, Alcatel-Lucent, listing the ways that a G.9960=G.hn over BPL deployment can pay for itself:

Here's the contents of slide 7:

"Home Area Network : Part of the drive to Home Automation

Security :

Video surveillance

Technical security

Gas, Flood or Fire alert

Tele-assistance :

Health monitoring

Elderly support

Assisted living

Home energy management

& Climate control :

Remote monitoring and control

of heating, A/C,

appliances, lights,

Shutters and shades motors …

Home automation

& User remote control :

Control thermostats, windows, doors, appliances, sprinklers, entertainment, media from anywhere

Vacation home monitoring

Pet feeding


Smart-metering : Power, water …

A nascent market: full of opportunity"

Add all that up and it's easy to see how fibre to the transformer with a G.hn gateway there could be paid for. So G.9960 gateways as part of the smart meter would seem to be the key to doing this *fast* and getting everyone a gigabit to every AC outlet.


Post a Comment

<< Home