Wednesday, March 18, 2015

Greg Wyler reports OneWeb progress

Think about the possibility of a WiFi network with a low-latency, 50 mbps back-haul link to the Internet in every school or rural clinic in the world.

I've been tracking Greg Wyler and Elon Musk's satellite Internet projects for some time. Both have been relatively quiet (most of what I know of Musk's SpaceX project came from an unauthorized cell phone video of a recruiting talk he gave), but Wyler talked about his company OneWeb in a keynote at the Satellite 2015 Conference yesterday.

Wyler plans a constellation of about 650 satellites in low-earth orbit (about 1,200 kilometers). He said that they plan to launch satellites in 2017 and hope to begin offering service in 2019. (It seems that OneWeb is ahead of the SpaceX schedule).

They will offer 50 mbps, 30 ms latency connectivity to $250 ground stations that will also serve as hot-spots, providing WiFi, LTE, 3G or 2G connectivity.

As shown below, a terrestrial route between Los Angeles and the tip of Chile requires 14 hops. The same route via satellite may require only five low-latency hops. (The figure is drawn to scale).

Think about the possibility of a WiFi network with a low-latency, 50 mbps back-haul link to the Internet in every school or rural clinic in the world.

Wyler showed a prototype of one of his ground-stations and also showed how easy it is to set up. The operator just spreads the solar panels and turns it on -- five seconds install time. Here we see one on the corrugated roof of a building:

This ease of deployment would be terrific for establishing ad hoc communication in the wake of disasters that had disrupted terrestrial communication.

While I have focused on OneWeb's primary goal of providing Internet connectivity in developing nations and rural areas, Wyler also spoke of providing connectivity in aircraft (and ships at sea).

Of course, all of this is speculation for now. Some conference attendees and presenters were skeptical about Wyler's project, pointing out that his low-cost satellites would have to be replaced every five years or so -- a recurring expense. Critics also pointed out that much of the time, the low-earth orbit satellites will be over oceans, polar regions and other sparsely populated areas.

That being said, Wyler has been able to attract backers and partners, each of which brings money and expertise to the table:
Like a modern Internet company that follows the dictum "do what you do best and link to the rest," OneWeb will focus on the backbone and market through local retail Internet service and cell phone providers.

One can also imagine OneWeb providing competition for conventional terrestrial ISPs in developed nations. I can dream of going over to Best Buy, picking up a OneWeb ground station, installing it on my roof and escaping the clutches of my ISP monopolist Time Warner Cable. I am not holding my breath till that happens, but I will be keeping my eye on OneWeb's ambitious project.

For some background on Wyler's previous satellite company, O3B Networks, and more on his plans for OneWeb, check out this video:

Update 3/20/2015

FierceWirelessTech interview of Greg Wyler.

Wyler says "We've got a pretty clear path. It's not just a technology problem. It is a technology, regulatory, implementation, education problem. It's kind of a little bit of everything." In the interview, he talks about terminal design, their business model and spectrum.

As mentioned above, he stresses ease of installation and low cost for the terminals. OneWeb has the rights to the Ku and Ka spectrum they will use and patent-pending technology to assure non-interference with geo-stationary satellites in those bands. Scale is critical to their business model -- once the constellation is operating, they marginal cost of a new customer is very low.

Friday, March 13, 2015

5G mobile update from Ericsson and Samsung at Mobile World Congress (MWC)

5G mobile communication is coming and prototypes are being developed along with demonstrations. (Remember the saying that projects had to "demo or die")?

Here are a couple of 5G prototype demos:

Samsung transmission speed demo -- 7.5 gbps standing still and a 1.2 gbps in a car going 112 kph:

Erricson demonstrates seamless hand-off between LTE and 5G:

and they brought their virtual reality, remote control excavator with them to MWC:

The final products will probably not be as fast as these prototypes, but they will eventually cost about the same as today's mobile radios.

Both Samsung and Ericsson are talking about initial deployment around 2020, but general rollout and ubiquitous adoption will take many years after that. (This is one technology in which developing nations, which are generally more mobile reliant than developed nations, may somewhat narrow the digital divide). Furthermore, there are no 5G standards, and you can bet there will be more than one.

Wouldn't it be nice if there were a global 5G standard -- everyone using the same license free spectrum and protocols -- your phone moving seamlessly between nations and carriers -- cars that were compatible with instrumented roads everywhere ... like WiFi ...?

Awake again -- Maybe I will get a Verizon 5G phone for use in the US around 2022.

By that time, out mobile devices will be 10-20 times as powerful and there will be a lot of "things" connected to the Internet. What new applications will we find for this high-speed, low-latency wireless connectivity?

Thursday, March 12, 2015

Leosat -- a third satellite Internet company

I've been tracking Greg Wyler and Elon Musk's plans to launch low-earth orbit satellites to provide Internet connectivity. Musk's SpaceX and Wyler's OneWeb have now been joined by a third would-be low-earth connectivity provider, Leosat.

I've not heard about this effort until now, but former Schlumberger executives Cliff Anders and Phil Marlar have been developing the network architecture, spectrum plan and satellite payload since 2013, and they just hired satellite industry veteran Vern Fotheringham as CEO.

Leosat will not be marketing to individual end users, but will target government and business -- maritime applications, oil and gas exploration and productions, telecom back-haul and trunking, enterprise VSAT, etc. Their market seems closer to Wyler's former company O3b, but Leosat plans to cover the entire Earth, while O3b is restricted to locations near the equator.

They plan to offer encrypted connectivity at up to 1.2 gbps with latency under 50 ms using a constellation of 80 to 120 small satellites, with launches beginning in 2019 or 2020.

While SpaceX and OneWeb have focused their publicity on end users and developing nations, they will also have the ability to deliver low latency service over long distances. As shown below, a terrestrial link from my home in Los Angeles to La Universidad de Magallanes in Punta Arenas, Chile required 14 hops whereas a satellite route could be achieved with five hops. (The following illustration is drawn to approximate scale assuming a satellite altitude of 700 miles).

The Ping time for the terrestrial link averages around 224 ms, considerably slower than the sub 50 ms latency Leosat hopes to achieve.

Like many Americans, I am served by a monopoly Internet service provider. Might these folks actually be able to provide competition -- at least in the developing world -- some day?

Monday, March 09, 2015

Google and Facebook report on developing world connectivity at Mobile World Congress

I've been studying and working on the Internet in developing nations since 1991 when only a few nations had any sort of Internet connection, as shown in Larry Landweber's 1991 connectivity map:

Every nation is connected today, but the digital divide remains as deep as it was in 1991. Both Facebook and Google are working to bring the 3-4 billion people who do not have Internet connectivity online and they described their efforts at the recent Mobile World Congress in Barcelona.


Sundar Pichai, senior vice president of Android, Chrome and Apps at Google Inc., updated the audience on two projects -- Project Loon and Project Link.

Project Loon seeks to deploy a constellation of balloons at an altitude of around 20 kilometers -- above the mountains, air traffic and weather.

The balloons will be airborne routers able to communicate with end users, each other and Internet back-haul locations.

Pinchai said the balloons now average more than six months in the air and keep nearby smartphones operating at 4G or LTE speeds, around 10 megabits per second. “We are well on our way to a platform that, by the end of the decade, will touch 4 to 5 billion people.”

He also gave a progress report on Project Link in Kampala, Uganda where they have installed over 800km of fiber, creating an urban backbone.

As is often the case with municipal networks (as in Stockholm), Google is not a retail Internet service provider, but provides wholesale connectivity to retailers. Pichai said they would be expanding Project Link -- installing fiber backbones "many more" African cities this year.

For more on the Kampala deployment and a thoughtful analysis of the reason for its success, see this post by Steve Song.


Facebook CEO Mark Zuckerberg spoke about, which hopes to make basic internet services affordable, so everyone with a phone can join the knowledge economy.

While Google is working on long range projects (including an investment in Elon Musk's SpaceX project to provide Internet service using low-earth orbit satellites), is already up and running in Ghana, Columia, Kenya, Tanzania, Indonesia and India.

Facebook and their partners are focusing on improving traditional terrestrial cell phone technology by improving mobile infrastructure, mass producing cheap, powerful cell phones and caching and compressing data. Their partners reflect this orientation – phone manufacturers, Opera, a Web software company, and Mediatek, a fabless semiconductor company.

Note that they want to provide only “basic Internet services,” not access to the open Internet. For example, in India they offer access to Facebook and 37 other web sites.

Facebook also has a Connectivity Lab lab working on more exotic, long-range solutions.

Short videos on Project Loon and

Project Loon:

Friday, March 06, 2015

The Internet routes around censorship

The Indian courts failed in their attempt to stop the showing of "India's Daughter," a BBC documentary exposing the New Delhi bus gang rape of a medical student and its aftermath.

The Indian government banned the showing of the film and the BBC blocked it on YouTube for copyright reasons. (Perhaps it is visible in Britain).

Banning the video gave it notoriety, increasing its popularity. (This is an example of the so called "Streisand effect," referring to the rush to view an aerial view of Barbara Streisand's house when she objected to it being posted online).

I am not certain when it was banned on YouTube, but it became available on Vimeo on March 5 and by the afternoon of the 6th had been viewed 60,000 times, but it was subsequently taken down.

As of this writing, it is available on the Daily Motion site. By the time you read this, it may be gone from there, but you will probably be able to find it using Google search. (If you are reading this from England or using a VPN -- is it still available on the BBC Web site)?

At nearly the same time, the Chinese government blocked access to "Under the Dome," a scathing documentary on pollution, which had hundreds of millions of view on Chinese Web sites within days of its release.

It may have been banned in China, but it is readily accessible in other nations (with English subtitles) and to any Chinese person willing to use a VPN to view it on YouTube.

Information wants to be free.

Friday, February 13, 2015

Innovations in vocational education and certification

Certification options from Coursera and the California Community Colleges

MOOCs are often used for vocational training rather than a traditional college degree and Coursera has launched six "Coursera Specializations" for vocational training.

A Coursera Specialization requires completion of a group of related courses followed by a capstone project. A Specializations consists of several online courses, developed at universities, leading up to a real capstone project/case study developed by a company in the relevant industry.

For example, the University of California at San Diego (UCSD) and Instagram have collaborated on an interaction design Specialization. UCSD will provide six MOOC-format courses and the capstone project will come from Instagram.

Students can still take the courses for free, but they will not receive a certification of completion and will not be allowed to do the capstone project. Students wishing to complete the capstone and receive a certificate of completion will pay a tuition of $343.

A sample Specialization completion certificate

Another approach to vocational education is being taken by the California Community Colleges, which are proposing fifteen vocationally-oriented bachelors degrees. For comparison with Coursera, consider this proposal for an interaction design degree from Santa Monica College (SMC).

The SMC interaction design curriculum

The SMC program would take longer to complete and would cost more in terms of tuition and opportunity cost, but it covers more ground and is taught face-to-face.

These are interesting, innovative times for vocational education. Hiring practices and societal values will eventually determine the winners, but for now, how would you advise a young person who wanted to become an interaction designer?

Saturday, February 07, 2015

Tomorrow's SpaceX launch: a reusable rocket, science and Earth's next selfie

Sunday February 8 at 6:10 EST (two minutes after sunset), a SpaceX rocket is scheduled to launch. Previous SpaceX satellites delivered payloads into low-Earth orbit, but this one is destined for the Lagrangian Point nearly 1 million miles from Earth.

At the Lagrangian point 1 (or L1), approximately one
million miles from Earth, the gravitational forces between
the sun and Earth are balanced, which provides a stable
orbit that requires fewer orbital corrections for the
spacecraft to remain in itsoperational location for a
longer period of time.
Source: NOAA

There are several reasons I will be watching the livestream of the launch.

SpaceX will attempt, for the second time, to recover the rocket. The first time they tried to recover a rocket they failed, but they understand the reason for the failure and hopefully will succeed this time.

The satellite, called "DSCOVR," has scientific and symbolic goals. At the Lagrangian Point, DSCOVR will remain stationary with respect to the Earth and the Sun, enabling it observe the Sun and serve as an early warning system for potentially disruptive solar flares.

Being stationary relative to the Earth will also enable DSCOVR to serve as a distant "Web cam" providing us with a feed of the entire, fully-lit Earth -- an ever changing version of the famous "Blue Marble" picture taken from Appolo 17. (Al Gore called for this space cam while Vice President and, after a long political struggle, his vision is about to be realized).

Earth's first selfie -- from Appolo 17

If SpaceX succeeds in recovering the their X9 rocket, they will refurbish and reuse it in a subsequent launch, cutting cost significantly -- and moving us a step closer to Internet access using a constellation of low-Earth orbiting satellites.

Update 2/8/2015

With a bit more than two minutes to go, the Falcon 9 launch was scrubbed -- there was an apparent problem with part of the telemetry system as well as at an Air Force radar tracking station.

They may try again tomorrow about two minutes earlier than today.

The picture below is from the launch live stream just after it was scrubbed.

Update 2/9/2015

Weather conditions are not favorable for a Monday launch and so NASA, NOAA, the U.S. Air Force and SpaceX have made the decision to postpone the launch until Tuesday, February 10 at 6:05pm ET with a backup date of Wednesday, February 11.

Thursday, January 29, 2015

Regulation of global satellite Internet service providers

Would global Internet service providers require unique regulation and, if so, what should it be and who has the power to do it?

SpaceX CEO Elon Musk, who hopes to orbit a constellation of Internet-access satellites, recently gave an invitation-only talk announcing the opening of a satellite-design office in Seattle. (An attendee recorded the talk and posted it on YouTube).

Many invitees were engineers and Musk was recruiting, saying "it's a difficult problem so we need the smartest engineers in the world." Then, after a pause, he joked "and at the same time to make sure we don't create SkyNet."

The audience laughed, but he was, perhaps inadvertently, alluding to a serious issue. Issac Asimov wrote of Gaia, a sentient planet, and, while the Internet may be the embryonic nervous system of our planet, I am less worried about Musk creating SkyNet than creating Comcast on Steroids.

Two companies, Musk's SpaceX and Greg Wyler's OneWeb, are competing to provide Internet connectivity in locations that are now unconnected -- as Wyler puts it, to connect "the other three billion." If one or both succeed, we might have have a monopoly or oligopoly ISP serving half the Earth's population.

As a Time-Warner Cable Internet customer, that worries me. They would be able to charge monopoly-level prices and offer the same last-place customer satisfaction as American ISPs. They would be global companies with political power and the ability to control half the world's information -- a combination of the Koch brothers, Fox News and Comcast.

Do these potentially global service providers require unique regulation and, if so, what should it be and who has the power to do it?

What might the regulation be? I can ask the question, but neither I nor anyone else knows The Answer; however, one suggestion is to keep both SpaceX and OneWeb out of the retail Internet service market -- restrict them to providing wholesale transport service on an equal basis to any would-be retail ISP. Even if only one of the two companies succeed, that would allow for retail competition and would help out with the monopoly price and crummy service issues.

A possible approach to avoiding political abuse would be to prohibit them from refusing service to any retail ISP in any nation.

Regardless of what we wish to do, who has the authority to create and enforce such regulations? Musk said SpaceX has the ITU's permission to launch the satellites and recognized that he will have to negotiate for the right to provide service on a country by country basis. SpaceX and OneWeb are both US corporations and therefore subject to US law, but is it right for global infrastructure to be regulated by a single nation?

Lest this sound too negative, I hope SpaceX and OneWeb both succeed in connecting the other three billion people on the planet -- the benefit to mankind will outweigh the difficulty of defining acceptable, effective policy.

Update 1/31/2015

Jason Koebler compares Elon Musk to the 19th century railroad barons, saying that being the first to develop technology to soft-land and reuse rockets will give him an unassailable first-mover advantage in space -- for imaging, communication and other applications.

Saturday, January 03, 2015

I'm on vacation

You will not see new posts on this blog before January 21.

CIS 471: Why I have not been posting on this blog lately: too busy with events in Cuba

Cuba has been in the news since President Obama announced changes in our Cuba policy and agreed to the prisoner exchange that freed Alan Gross, who was serving a 15 year sentence for bringing tech equipment into Cuba.

I've not posted anything on this blog for several weeks because I have been busy with recent events on another blog I maintain on the Internet in Cuba. The following are my recent posts concerning Alan Gross and the future of the Internet in Cuba. They are in chronological order, beginning with a November 11 post asking whether Gross was about to be freed:

(for background on the case -- what Gross brought into Cuba, its technical and propaganda importance, his incarceration, court cases, and negotiations for his release, click here.)

Alan Gross brought three of these kits into Cuba.

Alan Gross and his wife Judy just after his release from prison

Apple store vs. Microsoft store

Why is the Apple store jammed and the Microsoft store nearly empty?

On December 27, I went to the Microsoft store in the Century City mall in Los Angeles to take a look at low cost laptops I had heard reviewed favorably on a podcast, thinking I might get one to take on an upcoming trip.

The sales people were friendly and left me alone while I played with a variety of computers for around half an hour. This is what the store looked like:

It turned out the cheap laptops were too cheesy so I left and walked around the corner to the Apple store:

The store was noisy and jammed and there was a roped-off line of people waiting to be allowed in when others left.

The Microsoft store had ultrabooks with great keyboards, trackpads and touch screens and a variety of all-in-one computers. The product quality ranged from those cheesy laptops to very nice machines that were better deals than comparable Apple computers. You could walk right up and talk to a tech support or sales person in the Microsoft store, but needed an appointment to talk with one of the tech support "geniuses" in the Apple store.

How do you explain the difference? Is it all due to the iPhone?

Saturday, December 20, 2014

A digression -- a joke and some cool images on US-Cuba relations

This post is off topic -- not about the Internet per se -- but the joke cracked me up and the image gallery accompanying this NY Times article are a terrific recapitulation of US v Cuba since 1959.

Click here for the image gallery

Friday, December 19, 2014

Welcome home Alan Gross!

By now you must have heard about the release of Alan Gross after five years in prison for bringing several satellite receivers, computers and WiFi access points into Cuba.

I've been quite busy since his release, so have not had the time to comment on it on this blog. I've another blog on the Internet in Cuba and have been following Alan Gross's story -- both the technology and the politics -- for several years.

I am very happy to end that thread!

I'm also happy that his release has removed an obstacle to the normalization of relations between the US and Cuba -- that will benefit the Cuban people and the Cuban Internet.

This may or may not be the end of Alan Gross's involvement with the Internet in Cuba -- he is clearly an advocate of Internet freedom and a friend of the Cuban people.

I hope he adjusts quickly to his freedom and am looking forward to hearing more from him if he cares to share his experience.

Monday, November 24, 2014

RAND Corporation's contributions to computer science

What comes to mind when you hear the word Rand? Ayn Rand? Rand Paul? For me, it is the RAND Corporation. Project RAND (research and development) was housed at Douglas Aircraft in Santa Monica, California immediately after World War II, and became an independent, nonprofit organization in 1948. Perhaps the first "think tank," they spun off their development work, creating the the System Development Corporation (SDC) in 1957.

I don't know how one ranks research institutions, but, for me, RAND ranks right up there with Bell Labs, IBM Research and newcomers Microsoft and Google research. The following are summaries of a some of the computer science advances made by RAND researchers and consultants.

Communication satellites: Science fiction writer Arthur C. Clark outlined the vision of geostationary communication satellites in a short article published in October, 1945. Five months later, Frank Collbohm and James Lipp published a comprehensive engineering study on a "Preliminary Design of an Experimental World-Circling Spaceship."

Arthur C. Clarke's vision (left) and RAND's design 

Artificial intelligence: Herbert Simon, Allen Newell and Cliff Shaw did early work on artificial intelligence at RAND and Carnegie Tech. They asked people to talk out loud while proving theorems and noted that their strategy was to apply operations that reduced the differences between the current state of the proof and the theorem they were trying to prove. Their programs, Logic Theorist and General Problem Solver, did the same -- and so does this pigeon:

A pigeon solves Wolfgang Kohler's box-and-banana
problem by applying the Box Move operator.
Operations research: George Dantzig and Richard Bellman invented mathematical techniques for finding optimal or near-optimal solutions to complex, but well defined problems. This work has applications in network design and you use Dantzig's simplex algorithm whenever you build an Excel spreadsheet to solve a linear programming problem.

SIMSCRIPT: Harry Markowitz and Bernard Hausner invented the SIMSCRIPT programming language for simulating systems like customers moving through checkout stands at a market. SIMSCRIPT was an early object-oriented language in that it modeled the world as sets of entities and their attributes. Entities could be created and destroyed and their attribute values and set memberships changed when events in simulated time occurred. (SIMSCRIPT is close to my heart because it was the subject of the first class I ever taught. Unfortunately, my wife threw out my SIMSCRIPT t-shirt years ago).

T-shirt -- Entities, Attributes and Sets

An early research computer: Early computers were built as research projects at universities. You can recognize them by their names ending in "AC" for automatic calculator. RAND's JOHNNIAC (named in honor of mathematician and computer architect John von Neumann) was a stored program computer. It was used for applications including the early artificial intelligence research and operations research mentioned above.


Timesharing: Terminals and drum storage were added to the JOHNNIAC, enabling Cliff Shaw to create JOSS (JOHNNIAC Open Shop System), one of the first interactive time sharing systems. JOSS was "open shop" in that users interacted directly with the computer rather than dropping off jobs to be run at a later time by a computer operator. I was fortunate to see demonstrations of both JOSS and QUICKTRAN, an interactive FORTRAN interpreter built by John Morrisey of IBM. The advantage of these systems over batch processing was immediately and strikingly apparent. This led me to SDC (mentioned above), which by then had a more advanced time sharing system that I used for my dissertation research on man-machine data analysis.

Programmer at a JOSS terminal
The RAND tablet: The RAND Tablet, the great grandfather of the iPad, was built by Tom Ellis and his colleagues. Their GRAIL (graphical input language) software featured object-oriented drawing and character recognition. Their publications are some of the earliest work on human-computer interaction -- GRAIL was the great grandfather of Macdraw. (The following video clip is narrated by Alan Kay of Dynabook fame).

Object-oriented drawing and character recognition on the RAND tablet

Large, packet-switching networks: I've saved the best for last. Paul Baran published a series of eleven reports on Distributed Communications Networks in 1964. In Volume 2, he described the network architecture:

Traffic to be transmitted is first chopped into small blocks, called Message Blocks or simply messages. These messages are then relayed from station to station through the network with each station acting as a small "post office" connected to adjacent "post offices."
After simulating this system and considering the technology of the day, Baran concluded in Volume 11 that
It appears theoretically possible to build large networks able to withstand heavy damage whether caused by unreliability of components or by enemy attack.
He was right!
Paul Baran's distributed network arhitecture

In the following video, Baran reviews RAND's work on distributed networks and packet switching -- from early concern over the possibility of a nuclear attack disrupting military communications through skepticism about packet switching and the creation of the ARPANet. (Since the ARPANet was just a research project, they did not need to bother with security and encryption).

Paul Baran on distributed networks and packet switching (38 minutes plus Q&A)

Two of the people cited here went on to win Nobel prizes. Harry Markowitz received the Nobel Prize in economics for his work on portfolio theory -- perhaps not tied to his work on digital simulation.

Herbert Simon also received the Nobel Prize in economics for his work on decision making. He noted that we do not make optimal decisions when choosing among alternatives because information about outcomes is incomplete, gathering more information has a cost and outcomes are multidimensional. In real life we make satisfatory decisions. This realization no doubt guided his studies of the thought processes of chess players and theorem provers and therefore his work on artificial intelligence.

(A personal note: I took a class from professor Simon as an undergrad. All I recall was that I liked him a lot and he told us about the chess game his "computer" -- whatever that was -- was playing with a computer in Arizona. I also met him much later, and he was modest and helpful -- told me he stored most of what he knew in his friend's heads).

You can learn more about any of this work on Wikipedia or using Google, but -- better yet -- download the historic reports by these researchers from the RAND Web site.

Update 11/29/2014

As noted above, RAND spun off its development work in 1957 when SDC was set up to build the SAGE (Semi-Automatic Ground Environment) air defence system, designed to defend the U.S. against nuclear attack.

SAGE was the first computer network and the project trained most of the computer programmers in the US at the time. The project also produced many innovations in programming and programming project management.

After SAGE, SDC built an advanced general purpose time-sharing and software development system on an AN/FSQ-32 (Army Navy Fixed Special eQuipment) computer built by IBM. The Q-32 was used for ARPA-sponsored research projects in man-machine interaction -- including my dissertation project. More on SDC in a forthcoming post.

The AN/FSQ-32 supported research on man-machine systems.

Monday, November 17, 2014

18F is doing e-government and gaining traction

18F: Open source and transparent processes -- who says government has to be old fashioned, slow and inefficient?

In an earlier post, I described USDS and 18F, new government agencies that are intended to improve US e-government in the wake of the HealthCare.Gov debacle. USDS is a management consulting firm for federal agencies that favors lean startup methods, open source and agile development by small teams. 18F complements USDS -- they build tools and implement government systems.

You can check 18F's open source projects at the "alpha" version of their project dashboard. As shown here, they currently have twelve projects in various stages of development.

Scrolling down, one sees the entries for each of the 12 current projects. For example, they are building a portal for submitting and searching for Freedom of Information Act requests for the Justice Department. (Note that the department is a partner not a client).

The project descriptions have links to pages where you can see and contribute to the code, discuss the project with the developers and the public, and read a news release describing the project.

18F is not unique. The UK Government Digital Service has the goal of "transforming government services to make them more efficient and effective for users." They were formed several years ago in response to dissatisfaction with the British Health System Web site. You can learn more in this NPR story.

18F and the UK Government Digital Service have something very important in common -- they are staffed by skilled experts who could be making more money in the private sector but have elected (perhaps temporary) government service. I saw the same thing in a study of the Internet in Singapore where the "best and the brightest," went to government service.

How great would it be if all of government were staffed by the same sort of people?

Thursday, November 13, 2014

Elon Musk and Greg Wyler's plans for global satellite connectivty

Routers in orbit

In the early 1990s, cellular pioneer Craig McCaw, Microsoft co-founder Bill Gates and Saudi prince Alwaleed bin Talal founded Teledesic, with the intention of providing global Internet connectivity using low-earth orbit satellites. The satellite and launch technology were not good enough and the company failed.

Teledesic animation showing a satellite
constellation that would cover the planet.

But satellite and launch technology have come a long way since that time. In an earlier post, I asked whether Google could connect the "other three billion" in developing nations and rural areas. The post surveyed Google projects involving high altitude platforms like blimps, drones or balloons that hover or circulate in the stratosphere, low-earth orbit satellites used for imaging and telephony and medium-earth orbit satellites used for communications and navigation.

One of those projects was a collaboration with O3b (other three billion), a company founded by ex-Google executive Greg Wyler. O3b began with four satellites in 8,000 kilometer equatorial orbits and planned to serve all parts of the Earth within 45 degrees of the Equator. Wyler has left the company and they now have 12 satellites in orbit.

In describing the O3b project, I wondered "whether they are considering a low-earth orbit constellation" and it seems they were. Mr. Wyler subsequently left O3b to found WorldVu, which planned a constellation of 300 satellites at between 800 and 950 kilometers in altitude and has acquired Ku-band spectrum. Service will be marketed under the OneWeb brand.

That takes care of the improved satellite technology, but how about launch technology?

The Wall Street Journal recently published an article saying that Mr. Wyler would be teaming up with Elon Musk, founder of SpaceX to provide global Internet access using a constellation of 700 satellites, each weighing less than 250 pounds. Musk confirmed the plan in a couple of Twitter posts, but also criticized the Wall Street Journal reporting.

I hope they will be able to realize Teledesic's 1990 vision using 2020 technology.

I concluded my earlier post on this topic by "wondering whether Jeff Bezos, founder of Blue Origin, Elon Musk, founder of SpaceX, and Richard Brnason, founder of Virgin Galactic are eyeing those other three billion people." I still wonder about Bezos and Branson.

Update 11/13/2014

After writing this post, I attended a session at Rand Corporation's Politics Aside conference and had a chance to ask Simonetta Di Pippo, the Director of the UN Office for Outer Space Affairs, about her take on this proposal. She did not give a direct answer, but said that Elon Musk is a very smart man and he has never failed to succeed at anything he committed to do.

A SpaceX executive overheard my question and said he could not comment, but he reiterated Elon Musk's tweeted statement that the Wall Street Journal article had errors and we would have to wait a couple of months for the full announcement of their plans.

I guess we will have to wait to see, but this could be a Big Deal.

Update 12/15/2014

SpaceX will carry micro-satellites made by Planet Labs to the International Space Station (ISS) for launch into orbit.

Two Plane Labs satellites just after launch from the ISS

The satellites shown being launched are Planet Labs earth-imaging satellites. They are smaller and orbit at lower altitude than those discussed above, but might a constellation of more, smaller satellites in lower orbits and carrying routers rather than cameras be suitable for Internet communication? (That is not a rhetorical question -- I do not know).

For more on Planet Labs, check this terrific Ted Talk by Planet Labs co-founder Will Marshall:

Update 1/14/2015

The competition heats up.

Greg Wyler’s OneWeb satellite-Internet company has received funding from the Virgin Group and Qualcomm and Richard Brnason of Virgin Group and Paul Jacobs from Qualcomm will have seats on the board. (If you are a Wall Street Journal subscriber, there is a longer article here).

Elon Musk also announced the opening of a Seattle office for the design of satellites. SpaceX declined to comment on Wyler's announcement.

Whoever builds the rockets, satellites and markets the service, it sounds like Teledesic is being reborn using modern technology and, if successful, it would be a major extension of the nervous system of the Earth and a significant enabler of Bill Gates' work in developing nations.

Greg Wyler wants to bring the Internet to the entire world.

Update 1/17/2015

Elon Musk announced that he plans to deploy a constellation of router equipped-satellites -- evidently in competition with the Greg Wyler's OneWeb project. Musk announced his plan at a closed meeting for potential employees of his new satellite office and state, local and federal government officials.

No details were released and those that were leaked appear to be inconsistent. For example, the Seattle Times reported that they planned to deploy 4,000 geosynchronous satellites and Business Week reported that they planned 700 low-earth orbit satellites.

This effort is not an end in itself, but part of a larger plan to reach Mars -- Musk says he wants to die on Mars.

It is terrific to see two powerful groups competing to fulfill Bill Gates' original vision of global satellite connectivity -- Teledesic. Teledesic failed, but with modern launch capability, micro-satellites and communication equipment one or both of these efforts may very well succeed. If they do, it will be an historic achievement and a significant complement to Gates' current work in developing nations.

Elon Musk will compete with OneWeb

Update 1/19/2015

When this thread began last November, it seemed like Elon Musk and Greg Wyler would collaborate on an Internet satellite venture, but now it looks more like competition.

A post on Ars Techica quotes Musk as saying “Greg and I have a fundamental disagreement about the architecture -- we want a satellite that is an order of magnitude more sophisticated than what Greg wants. I think there should be two competing systems.”

They quote Richard Branson as saying that Musk doesn't have a chance because Wyler has spectrum rights and there is not enough space for two satellite constellations. He thinks the logical thing for Musk to do is work together rather than separately.

Wyler's former employer Google, which has also been working on satellite connectivity, is said to be close to investing in Musk's Space X.

I can't wait to see where all this ends up in, say, five years.

Update 1/20/2015

SpaceX has confirmed an investment of $1 billion from Google and Fidelity for a reported 10% of the company. That leads to an evaluation of around $10 billion. (I may be old fashioned, but I don't understand markets that evaluate WhatsApp at nearly double the value of SpaceX).

It also seems that SpaceX is considering the use of modulated laser beams to cope with OneWeb's advantage in spectrum holdings.

Regardless of the technology, OneWeb and SpaceX will have to deal with regulators in each nation they serve, which seems inefficient -- would it make more sense to establish some international regulatory rules?

Looking forward -- what if one of these companies pulls their plans off and ends up serving a billion or two billion customers -- should we worry about their power? It sounds like Comcast on steroids. Even if they both succeed and establish a duopoly, they will have immense power.

Update 1/26/2015

Business week has published a background piece on Greg Wyler -- his biography and personality. It is interesting to read for general background, but has a few details that are new to me. He says he plans to orbit 648 satellites at an altitude of 750 miles and hopes to sell the user terminals for around $200. Since there will be several satellites within range of any point on Earth, he says their antennas will not need to be professionally installed or move to track satellites, as is the case with O3B.

The article is accompanied by a 4:33 video in which Wyler describes O3B and his plans for OneWeb -- here are a couple of stills from the video:

Wyler with a mock-up of a user terminal

Wyler illustrates the latency differences between
low, middle and geostationary orbits

Update 1/26/2015

Cell-phone video (25:53 min) of Elon Musk's talk at the closed-door announcement of the establishment of a satellite design office in Seattle. Many high-level details on the project.

----- Update 2/12/2015

Third time is closer, but still no cigar :-(.

Update 3/10/2015

Excerpts from a Via Satellite interview of Greg Wyler.

VIA SATELLITE: With industry verticals being served, and O3b connecting the other 3 billion people, where does OneWeb fit in to the communications landscape?

Wyler: O3b Networks does link around 150 Mbps and up, and this is about links that are much lower speeds than that. Our primary core competency will be sub 50 Mbps to small, inexpensive terminals.

VIA SATELLITE: How difficult was it to get investors like Virgin and QualComm to buy into this vision?

Wyler: Qualcomm knows more about communications chips, handover protocols and LTE then any other company. They also have a long background in satellite having built Globalstar and many other satellite communications systems. Virgin has Richard as the leader with a strong understanding in things that you can’t imagine he would have a sense about, and then this deep bench of players.

VIA SATELLITE: I understand an RFP is already out regarding the manufacturer of these satellites? When do you hope to finalize this?

Wyler: We are building satellites at high volume. They need to be done in a production line, rather than a one-off manufacturing process. We are going into a partnership where we will own a portion of the factory and the manufacturer the other portion.

VIA SATELLITE: Is 2017 a realistic timeframe to launch the first satellites?

Wyler: I am an optimist. I think 2017 is a realistic time to have our test satellites up. I am not saying the constellation will be working then.


Update 3/18/2015

A third would-be satellite ISP, Leosat, has revealed plans for a constellation of Internet satellites. They will not be marketing to individual end users, but will target government and business -- maritime applications, oil and gas exploration and productions, telecom back-haul and trunking, enterprise VSAT, etc. They (and the others) hope to be able to provide low latency links over long distances. As shown here, a route from Los Angeles to southern Chile requires only 5 satellite hops as opposed to 14 terrestrial hops.

Wednesday, November 12, 2014

Roku has made my Google Chromecast superfluous

Do you still need a Chromecast device?

I am a cord-cutter -- I use a Roku streaming device on my TV set and do not receive regular cable channels. (I have a trusty rabbit ears antenna for local over-the-air broadcasts, but that will not work for many people).

I also have a Google Chromecast, which I said I loved in a review a little over a year ago. But, a year later, it turns out there is nothing in the Chromecast app library that I want to see that is not also available on my Roku. In fact, I watch several Roku channels -- like PBS -- that are not currently available for the Chromecast. I guess Bill Clinton would say "it's the content, stupid."

But, I kept my Chromecast around for screencasting -- mirroring my computer or phone on the TV set -- until now.

It is now superfluous because Roku has released the beta version of Miracast screen mirroring for Windows 8.1 and selected Android devices.

As you see here, when I open the Screencast setting on my Android phone, I have two target destinations -- the Chromecast and the Roku streaming stick. (Both are connected to the same TV set).

I did an informal test of the two devices using the CBS All Access video streaming service. I watched episodes of "Big Bang Theory" using both devices, and did not notice a significant difference in quality. The video did occasional half-second stutters a few times and the audio would also drift out of synch from time to time, but the program was watchable on both the Roku and the Chromecast. I have no doubt that next generation hardware and improved video and compression algorithms will take care of those small glitches (as long as my ISP keeps the bits flowing smoothly).

Miracasting is only available on two Roku models -- the Roku 3 and the Streaming Stick -- and selected Windows 8.1 and Android devices, but no doubt wider support is coming. If you have a miracast-compatible device, you might as well unplug your Chromecast.

Update 11/20/2014

A couple days ago, I got the Android Lollipop update for my Nexus 5 phone and CBS updated their All Access application, so I decided to retest the streaming video quality. I used this as an excuse for watching another episode of Big Bang Theory and (subjectively) noted that the video had smoothed out -- there were no stutters -- but the audio had deteriorated -- it was out of synch the entire time. I don't know whether this is attributable to Lollipop or the app or a combination of the two.

I ran this test twice -- once streaming to a Roku Streaming Stick and the other to a Chromecast -- and the subjective experience was the same.

I'm disappointed by this step in the wrong direction, but I faster hardware and better algorithms will smooth out video glitches and synch the audio.

One other cosmetic change -- the Lollipop screencast screen has changed to black on white:

Monday, November 10, 2014

Google testing high-speed wireless -- the last kilometer for Google Fiber?

Google could even take the Android approach -- make the technology available to municipal governments and others and watch their advertising business grow as it is deployed.

In 2012, Goldman Sachs analyst Jason Armstrong looked at Google Fiber and estimated that it would cost them $70 billion to connect less than half of all US homes. He also estimated that it had cost Verizon $15 billion to bring FIOS fiber to 17 million homes. Armstrong concluded that he was "still bullish on cable, although not blind to the risks." (Armstrong has since left Goldman Sachs and works at Comcast and Verizon has cut back on FIOS).

That sounds grim, but what if wireless technology could significantly reduce the cost of connecting homes and offices?

Google has asked the FCC for permission to conduct tests of millimeter wave-length wireless communication for 180 days.

As shown below, short wavelength, high frequency (E-band) signals travel relatively short distances and can not pass through walls or other obstructions, but they enable gigabit and faster data transmission rates:

E-band wireless in context: The current market is dominated by a few companies selling equipment for cell phone backhaul and other point-to-point applications, but what if the smart guys at Google could figure a way to use it for neighborhood links? (Image: E-band communications.)

How much of Armstrong's $70 billion estimate would Google (or anyone else) save if they could run fiber to the block or neighborhood and reach individual homes using this radio technology?

Google Fiber started in Kansas City and today it is available in two other cities (and some surrounding areas). They are currently evaluating 34 additional cities and those cities would look a lot more attractive if they were able to use wireless links to reach homes from neighborhood poles. Google fiber could also provide backhaul for mobile communication.

If this dream materialized, Google would provide stiff competition to the incumbent phone and cable companies and drive connectivity prices down, but would that be the best solution for the public?

In the US, most of us have only one or perhaps two competing Internet service providers. Google would be a second or third, but we would still have an oligopoly and, while Google may not "do evil" today, who knows about the future?

Google, Comcast or any other ISP must deal with local government for things like access to tunnels, phone poles and utility boxes. Might we not be better off in the long run if local government owned the infrastructure regardless of the technology? This solution has worked well in Stockholm, Sweden, where the municpality owns the infrastructure and sells wholesale access to ISPs who service customers.

What will Google do if this technology works out? They could become nationwide wholesale or retail ISPs or even take the Android approach -- make the technology available to municipal governments and others and watch their advertising business grow as it is deployed.

All of this is highly speculative, but if the technology and business model work out, we may be able to get low-cost gigabit connectivity without moving to Kansas City.

Saturday, November 08, 2014

Harvard study of variance in lecture attendance

Attendance varies between courses, with the day of the week and with special events like exams and guest speakers.

Samuel Moulton, director of educational research and assessment for the Harvard Initiative for Learning and Teaching gave a presentation on their preliminary research on lecture attendance. He reported on attendance from 10 classes, and found that attendance varied depending upon day of the week:

Attendance soared on exam days and dropped the Friday before Spring break:

(Attendance can be over 100% since student drop classes after enrolling and attending the first few lectures).

He also saw that special events like an optional movie or a guest speaker had an impact on attendance:

Moulten further analyzed the data by dropping events like exam days and fitting fitting a line to the data:

He concluded by showing the plots for all of the courses in his sample:

As you see, there is significant variation in attendance.

How do these results compare to your experience? What factors contribute to attendance variation?

Watch Moulton's presentation: