I was Director of Marketing with Teledesic, Craig McCaw and Bill Gates's Internet-in-the-Sky from 1995 to 2000. I'm a huge fan of what Musk is doing, with order of magnitude cheaper launch costs and satellite hardware. AMA.
How far did you get? What went wrong? What sort of latency/bandwidth did you expect to achieve? Was it relaying via geosynch/higher orbit 'hubs' or peer to peer sideways routing?
We raised $1.6 B, received the largest spectrum allocation in the history of the world (1000 Megahertz, in the Ka band, available globally), and launched one test satellite.
But, the biggest problem with these sorts of systems is that there is no MVP. Essentially everything has to work before anything can.
Our design was for a network of Low Earth Orbit satellites, which would mean cellular-like latency. We were planning to offer tens of megabytes per second, but it would have been really expensive. Our design used lasers for inter-satellite links, so there was no need for lots of local gateways.
In retrospect, the biggest mistake was to try to raise money from the incumbent telecom operators around the world, whose business we should be have been planning to disrupt.
It seems there are still phases you could do. Instead of trying to bring up an entire constellation at once doing two-way communications covering the entire globe, you could start with just a few broadcast-only satellites. Not competitive with at my house in urban North America where wifi and LTE is ubiquitous, but that's never going to be your market anyway.
The receiver on the ground is going to have to be specialized hardware anyway, so build in a battery and a real time clock so it can wake itself up when one of the satellites is in range, and then record everything the satelite is transmitting. Send the entire contents of Wikipedia once a week, send technical manuals, send the week's news, Windows updates, Linux distros; the Internet is so ubiquitous these days but fast access to it is still elusive in remote parts of the world.
This wouldn't have been useful in 2000 but 15 years later, a box with an wifi access point and a 1 TB hard drive chock full of information that's relatively up to date?
Not as cool as "internet anywhere in the world", but a start. Phase in delayed 2-way communication next, reusing the physical antenna from version 1; adding more satellites as you go.
Of course, 15 years later, the cost of putting 1 kg in orbit has also shrunk, and having a WiFi-capable computer in my pocket to look up information on is "normal" - which wasn't so, 15 years ago.
1. As an American-based company, SpaceX would be subject to CALEA [0], meaning that they would have to enable law enforcement intercept of communications with a warrant. This would likely be done by replicating the packet streams at the gateway. However, it's quite possible that you could use the system in a repressive country, without them knowing about it or having a way to stop you.
2) The purpose of the system (and the spectrum) will likely be broadband data, not hand-portable voice. So, it should be disrupting Time Warner and Verizon's cable and fiber offerings, not cellphones. The real value is internationally to have reliable broadband available anywhere.
"The purpose of the system (and the spectrum) will likely be broadband data, not hand-portable voice."
You're unlikely to get one without the other.. if down the road you can get an affordable internet connection on your phone without having to deal with cellular networks, I'd bet that you would see a huge shift to VOIP.
Edit: Whoops, looks like phones probably won't be able to receive this signal practically.
"I believe Musk is, like Teledesic, planning to use the Ka-band. That means the smallest antenna would be the size of a large laptop" [0]
How will devices on the ground communicate with the satellites? i.e. do we still need fixed dishes, or are we looking at something new so an iPhone can get signal right from the satellite?
I believe Musk is, like Teledesic, planning to use the Ka-band. That means the smallest antenna would be the size of a large laptop (and that's only if phased arrays become cheap enough). So, it would work for a car, plane, boat, or house, but would be too big to use while walking.
There are satellite (I think Ku band) dishes for larger (~50ft) boats nowadays. They are using gyroscope assisted motors that wouldn't suffice for compensating rolling motion of a smaller boat.
The laptop-sized antenna he mentioned doesn't need a gyro because it uses phased-array tech. Essentially you resolve signals in any direction (up to a resolution given by number of elements and array size) by carefully combining them with certain phase offsets such that signals from all but one direction cancel out. This adjustment is made on the fly to track a signal with changing direction.
Directionality and sectioning has been used since the beginning of cell technology: first with cell splitting and then with cell sectoring, and recently more aggressive splitting (smaller cells) -- it's an ongoing evolution. The kind of tracking this has required is very easy to do: once your signal is stronger in a competing sector for a significant amount of time, switch sectors. The same principle can be applied to phased arrays but the small number of sectors makes it much more tractable computationally and in terms of hardware.
A phased array in a city environment needs to track in real time the position of users, with enough feedback from them to make sure your prediction matches with their position well enough, and use this information to continuously adjust sub-nanosecond (depends on the freq/bandwidth) phase offsets to array elements. In principle it`s even possible to do multi-tower coherent communication. Needless to say, this tech is expensive right now, and I imagine hard to get to work reliably.
MIMO is also closely related, the difference being traditional MIMO actually adjusts amplitudes only (non-coherent) of multiple data streams to create channels (after a linear transformation at the receiver). This means the signals are still ominidirectional for your wi-fi MIMO. Being non-coherent means it's much cheaper.
1. Among the companies which have announced plans for space-based internet like SpaceX, OneWeb, LeoSat, Yalini etc which do you think can deliver good quality internet at the lowest costs.
2. What according to you is the most likely terrestrial delivery model to succeed. Partnering with local mobile operators/ISP's (like OneWeb) or creating an altogether new infrastructure/gateways/ground stations (like Yalini).
3. I couldn't find details of SpaceX's plans (like throughput, speed, latency, costs, delivery model etc.). Can you refer me to any publicly available information on this.
If Musk achieves reusable launch vehicles, and that had been available to you guys back then, how much would that have changed things for you? Would the system pretty much be up and running today?
We were looking at about $10 K per kg to orbit, and we could barely make the economics work. It looks like SpaceX is offering $4 K. [0] Presumably, reusable launch vehicles will drop his costs another order of magnitude, but they shouldn't actually be necessary to make this sort of system viable.
From the article: "SpaceX would need permission from countries to operate the service, a process that could be difficult, if not impossible."
Yep. To operate in many countries an ISP has to block access to a lot of content. It differs greatly from country-to-country. Would they really get involved in that? That would mean taking orders from governments all over the world and implementing blocks per those orders. Or would they forget about providing access in those countries? That's a large number of countries.
Added: Forgot to mention one of the biggest obstacles to getting permission in many countries: surveillance. Beaming directly from a user to a satellite that relays to a ground station in another country is going to bypass local surveillance.
It's easy if consumer devices can talk directly to your satellites. AFAIK, this still needs a ground based relay to be of any use to the average person - it's a replacement for transoceanic cables or ISP backbones, not for WiFi.
Depends how expensive the receiver equipment is. Seeing as how this is an LEO constellation and not GEO one, and that internet requires bidirectional communication instead of a pure receiver they'll be more complicated and expensive than satellite TV for sure. I just don't know enough about the current hardware to know if it's within reach of the average consumer.
(BTW, I find it likely that even if the ground stations ate cheap enough to be privately owned, they'll be more like a cable modem than an 802.11 chip - a relatively expensive stationary piece of equipment that you hook up to an 802.{3/11} network in the home.
I imagine it would be easy for them to ignore the individual requirements of countries. But individual users would have to be able to broadcast in order to communicate here, and would therefore be more at risk of being found by authorities - if they can get their hands on the equipment to do so.
Questions about "crime" are a bit academic when it comes to international issues like this - it's less confusing to stick to more concrete questions like "would anyone shut down his satellite?"
But to answer the question, I believe a satellite is like a ship in international waters: you're obliged to follow the laws of the country whose flag you carry (and if you don't fly under a national flag you're... not really a criminal so much as an outlaw).
There are international agreements about radio transmission but AFAIK none have the force of international law. There is a long and glorious tradition of "pirate" radio stations transmitting from international waters (at least around Britain), which were often perceived as shaky (and technically illegal to listen to), but never prosecuted.
Note that countries have not succeeded in stopping satellite phone users within their borders... for example, the Taliban in Afghanistan use satellite phones extensively.
Right. That's an alternative, smuggle your satellite internet equipment into a country where it is prohibited. Could be risky during the smuggling. Also very likely easy to detect since it's not exactly a LPI (low probability of intercept) signal.
Beats me - but I do bet that these companies don't have any business presence in Afghanistan. And despite the Taliban being involved, which most countries in the world (and the UN) don't like, it hasn't been stopped. Which gives you an idea of how difficult it would be to stop ordinary people from using such a service.
Musk claims to want to launch 4,000 or so satellites. The Goddard Space Flight Center indicates that there are 2,271 satellites in orbit now.
I'm wondering to what degree this would make it more likely to have a negative feedback loop of satellite destruction in the case of a satellite being hit by space junk.
The problem has to do with cross-section -- these satellites will each be small, a few hundred kilograms each. In comparison, cubesats are 1-3 kg, and big satellites, which are most of the existing ones, are a few tons of mass each.
ah interesting. My main concern when reading this was latency, which has traditionally been crap with satellite ISP. In todays world i also feel that once you have a certain bandwidth (say 10-20Mbps) the benefits of a latency of <50ms vs 200ms outweigh those of more bandwidth because literally everything, from loading web content, chatting, making calls, realtime anything, is more responsive.
So the latency being so much better with this (hopefully) excites me :)
I think they're mostly in geosynch/geostationary at 36,000-42,000km or so, with latencies of 250-350ms (pure travel, not including queuing or processing delays)
There are a lot more large pieces of space junk in LEO than that. We track about an order of magnitude more than that above 5cm diameter. Exploded upper stages, debris from collisions, discarded separation bits and pieces, no-longer-active satellites.
4000 is still a large number, especially at a higher LEO altitude where junk lasts forever. There's a reasonable case to be made for shifting to a regime where we require that satellite providers remove more mass and more pieces of debris from orbit than they put into orbit. But we don't want to split the baby and force them into bankruptcy either. The weak interim guidance might be that built-in deorbiting devices are required for anything with more than a ~10 year orbital lifetime. This doesn't end the Kessler Syndrome cascade, but it at least doesn't contribute very much. Presently there is only a ~25 year voluntary initiative, and adherence is not universal.
The most advanced initiative in this area is the geosynchronous graveyard belt, which the FCC enshrined into law for US telecom providers, but even that fails a large fraction of the time.
I believe they would be low enough that if they failed they would pretty quickly fall back into the atmosphere, and be replaced on the fly. I'm pretty sure I heard this in his talk about it in Washington on youtube.
At the very least, it would bring more ISP competition to underserved /sarcasm extremely remote areas /sarcasm like where I live - that would be Portola Valley, CA - where the only internet provider available to most of us is Comcast. Which all of us hate.
I live 16 light-µsec from Stanford (OK, free-space velocity for C), and my neighbors and I have one ISP to choose from.
Go, Elon!! I'd light the fuses on the launch vehicles to see this done.
My first thought, here, was "Well what about the speed of light? Wouldn't that cause satellite internet service to suck, just like it does now? And isn't that why news correspondents are always delayed?"
But then I did the math:
Speed of light = 186 miles per millisecond. Satellites orbit at about 380 miles above earth (Source: Hubble).
So it only takes ~4 ms for a round trip to a satellite directly above?
I did not know that. Weird, wild stuff.
If someone knows specifics about why satellite communication is currently so slow (both internet and video), I'd be interested in the EL5 version.
Good question. The vast majority of satellite communications today are conducted via geostationary orbit, 35,786 km above the earth. This is the distance at which a satellite orbiting around the equator moves with the same velocity as the surface of the earth, keeping the satellite in a fixed position. This is good for communicating via the ground, since you can just set up your dish pointing in the direction of the satellite. Additionally, such a satellite can cover a wide swath of the earth since it is so far away; see, for example, the coverage of AsiaSat-8 (http://www.spaceflightinsider.com/wp-content/uploads/2014/08...). The disadvantage (especially for internet) is that the round-trip time of a signal at the speed of light is on the order of 250 ms.
As described in this article, the SpaceX constellation would be in a much lower orbit. The advantage is a much faster ping time (say, around 10 ms). The disadvantage is you need a large number of satellites to provide continuous coverage for the entire earth...but that's less of a disadvantage if your company also launches rockets.
Hmm, interesting. Well, I have Exede Satellite service as my ISP (best I can get) and my best latency, to any server, is 800ms. The way I figure it, 35,786 km / 300 miles per ms would be 120ms up * 4 for the round trip. So that's 480ms, plus another 20ms-100ms to a google or another large server. So that's theoretical latency of 600ms, so where are the other 200ms of latency coming from! :(
I am wondering because I maintain some linux servers for some of my projects, and any time I need to do something serious, I have to drive to a coffee shop and borrow their connection. SSH over 800ms latency is frustrating, to say the least.
And the other disadvantage is more expensive and complicated ground equipment that can handle changing angle to satellites and handoff on three order of every several minutes. But I imagine that's only getting cheaper...
The Iridium constellation is 485 miles above. I've used it for both voice and data at sea via a motorola handset. Voice was very responsive with no delay, though the compression was very apparent. Data is painfully slow at 2.4 kbps. We needed to use a specialized email client that compressed email, handled the duplex connection, and recovered from errors very gracefully.
I'm curious the reason for such slow data too. Looks like the design was optimized to maximize the number of voice connections for a minimal number of satellites. Combined with the age of the design (first launched in '97) gives pretty poor data connections.
Speed of light in vacuum > speed of light in optical fiber. Same reason high frequency traders are using microwave between Chicago and New York instead of a fiber connection.
Is it possible to build a multi-mode "fiber" with an air core so that a significant part of the light's path is through air? The outer sheath made of optically dense material would only be used for internal reflection to "guide" the beam...
No. The core has to be denser than the outside for the guiding to work; if you made a "fiber" like you describe the signal would just travel through the sheath - if it was heading towards the core it would internally reflect to stay within the sheath.
My understanding of it is that when light passes through a medium, it does not do so in a straight line. So, while the actual speed of light remains constant, it takes longer to get from one end to the other.
To be precise, light can travel through a medium in a straight line, but it bends when it passes from one material to another.
The light bounces off the inner walls of the fiber in a zig-zag fashion as it propagates, so it cannot possibly travel at the speed of light. It depends on the index of refraction as mentioned.
Unless the content you're accessing is hosted on the satellite, it will have to jump several more times over a much longer distance for a complete round trip. Can you factor that in to your estimate?
> A key employee leading Google Inc.'s efforts to beam Internet access from satellites has left the company and is now working closely with Space Exploration Technologies Corp. and its founder Elon Musk, according to people familiar with the matter.
> When he left, Mr. Wyler took with him the rights to certain radio spectrum that could be used to provide Internet access
I don't have access to the article so I don't know if it addresses my question but... How in hell can a single employee "Take with him" the radio-electric spectrum allocation???
If you google "greg wyler" and then click the wsj link you can visit the full article.
But as to your question, it appears that he joined google from a company he founded that owned the rights, then left and took the rights with him and formed a new company to control the rights again.
> When he left, Mr. Wyler took with him the rights to certain radio spectrum that could be used to provide Internet access, according to a person familiar with the matter. The person said Mr. Wyler had formed a new venture, WorldVu Satellites Ltd., that designs satellite systems and controls the rights to that spectrum.
> Brian Holz and David Bettinger, who joined Google with Mr. Wyler from O3b Networks Ltd., also have left Google, according to the person familiar with the matter. O3b is a private satellite company founded by Mr. Wyler.
Commodity hardware goes obsolete quickly and fails frequently. Satellite-grade hardware is much older and slower than what we use for today's websites. Putting anything into orbit is expensive. Energy is much more expensive in orbit than on the ground. And it's not like the majority of your customers are going to be satellite internet users.
If you use regular hardware you'd have to worry about bits getting flipped by radiation, which could cause some big problems depending on what your server is doing. Unfortunately radiation-resistant hardware is super expensive (maybe not a problem) and super slow (probably a problem)
Google is interested in deploying a global data network, hitting all the places that have never seen a fiber installer. They're pursuing every avenue available for that simultaneously - long-endurance balloons, solar drones, and LEO satellites; They are involved in SpaceX's project.
Satellite broadband is likely to always be an order of magnitude more expensive than terrestrial. It also has a very limited capacity in any given area. I don't know if they're pursuing this, but at Teledesic, we looked at hybrid models where when demand got large enough in any given geographic area, you could use a blimp to serve customers in the existing spectrum with the same user terminals. So, it could prove very complimentary to Loon (assuming both actually work).
As mentioned up thread, the size of the ground-side antenna is likely to be the size of a full laptop, so not something handheld. So this project would provide the back bone, and the Project Loon balloons would provide the "last mile", as it were.
It looks like the approach was different: instead of having several thousand of small and 'cheap' satellites, they were aiming for an expensive and big one.
Presumably the one big satellite was intended to be backhaul for the solar planes. They can probably buy that backhaul from someone else with less risk and effort.
The ability to go off-grid is becoming more of a reality for me now. I'd love to be able to get a cabin in the middle of nowhere, slap some solar panels on top still have access to highspeed internet.
Even if you didn't want to go off-grid, this could open up a lot of developing countries to being more viable remote-working locations. I recently spent some time in Bali and considered moving there for awhile but ultimately couldn't solely because the internet was too slow and unreliable.
The internet is a grid of computers to which you connect.
More than that though, conceptually "off-the-grid" means not accepting a utility serivce from a provider but rather catering for that yourself. An internet connection is arguably a utility, and you'd be accepting it from an ISP. That goes against the concept, which would suggest you should maintain your own isolated network of computers and not connect with the outside world.
Replacing terrestrial broadband is great, but what happens when your internet stops working because it is raining outside? How exactly do they plan to deal with rain fade while operating in the Ka band?
I wonder how it would compare to Exede Satellite service. I have Exede, it's the only thing available here in High Desert :( The speed is very good for what it is, I am getting 16-20 mbps down and 3mbps up. But, the latency is a huge issue for me. Most servers, even once in US, ping at 800ms, this makes it such a pain to do any kind of SSH server admin!
With low orbit devices I would imagine the latency should be better, I am guessing, but how much better.
> With low orbit devices I would imagine the latency should be better, I am guessing, but how much better.
Your satellite's 22,000 miles above the earth. If you were directly under it, round trip latency would be 236ms.
If these satellites are 684 miles up, and you were directly under one, round trip latency would be just 7ms.
The distance between satellites, and between the satellites and whatever ground stations actually connect them to the 'net, will all be smaller because of the lower orbits. Latency should be much better.
Besides enjoying Internet globally, I see hope freeing humanity from government censorship with Internet from satellites. This will also mean we will have cheaper global roaming prices for cellular access as if its more expensive with cell phone data people will just use the satellite based system instead.
How much power can such micro satellites transmit with?
How much power does the earth based transmitter need to communicate back to the satellite in space?
It worries me that they are launching satellites for the NSA whilst also build something the NSA would certainly love to tap into – I mean, realistically, the NSA are going to get their hands on the data if they want to, but giving them extra leverage over the company is a shame.
It also saddens me they are launching satellites for the NSA full-stop, at least until the NSA has been reined in a bit.
Quite exciting. At the same time, did anyone watch Kingsman?
I worry about a big play like this in a market with ridiculously high barrier to entry. We might find ourselves in a vulnerable position if we invite monopoly.
To me this is just 1 spoke of the wheel he will be providing this internet service to google loon balloons that need base stations on the ground currently.
Just because something is being built with other people's money does not make something a success. None of the companies Elon has been involved with have ever actually generated a profit -- they have largely just taken advantage of speculative investors and their willingness to buy into his ideas.
"Just because something is being built with other people's money does not make something a success."
It doesn't preclude it, either.
Pretty much every business larger than one or two person operations uses Other People's Money to grow to profitability. See Facebook, etc.
Use of OPM doesn't guarantee success, but for industries that have large capital entry and operating costs, it's necessary.
Musk is working in fields that manipulate matter, not just bits - so it will take longer to see if Tesla and SpaceX reach profitability ("success") than it would for a generic software startup.
Be patient. You may not know whether a company or a child is successful for a couple of decades, at least.
