Dear MSUA Mobility News Readers,
Do you click Space Intel Report articles and get frustrated that the content is behind a firewall?
You’re not alone. Each week SIR articles, are among
the top clicked news stories.
Good news – access is now within reach.
In partnership with Space Intel Report, MSUA is pleased to offer a 50% discount on SIR subscriptions. To take advantage of this offer and get unlimited access to SIR articles on SpaceIntelReport.com, click the following link and use coupon code 495rate. https://www.spaceintelreport.com/subscribe-to-space-inte
**********This Week's Featured Article***********
Google/Alphabet Loon LLC, now a full-fledged company, seeks satellite partnerships
by Peter B. de Selding | Sep 24, 2018
LONDON — Six years after emerging as a Google X project, stratospheric-balloon-connectivity provider Loon this year became a full-fledged company with its own management and balance sheet.
Loon LLC has won its first commercial contract, with Telkom Kenya, a geographically friendly inauguration of long-duration telecom coverage from a fleet of balloons at between 15 and 20 kilometers in altitude.
Max Kamenetsky, senior project manager at Loon, outlined the Loon business case and the lessons learned over six years about stratospheric winds that enable Loon to sign service-level agreements (SLAs) with customers.
Kamenetsky said satellite service providers should consider Loon as a potential partner to offer low- latency, high-capacity connectivity in small geographic areas within a satellite’s broader footprint.
Kamenetsky described Loon and addressed questions Sept. 19 at the VSAT Global and Next- Generation Satellite Applications conference organized by Informa’s tmt knect 365. Here are excerpts from his remarks:
Loon Senior Project Manager Max Kamenetsky puts the Loon High-Altitude Platform Station (HAPS) program in a satellite context.
The first experiments, about six years ago, were to put a router on a balloon and see what happens. We used a $200 weather balloon. And the answer is: It flies. It bursts after about two hours.
Fast-forward a number of years. Google parent is called Alphabet. And under that parent corp. There are a number of LLCs. Google is the most well-known one. One was X, an incubator focused on looking at large problems that may have a technological solution.
Project Loon found a home in X. And the objective of X is actually not to develop projects, but to quickly figure out if there’s a way to kill projects.
Project Loon went through exactly the same phases as all the other projects go within X, and we didn’t get killed. Actually, the opposite happened – we graduated. A couple months ago we announced that Loon is now an independent company within Alphabet. We’re a sister company to Google. But we have our own CEO, we have our own CFO, we have our own business plan, we have our own P&L.
Signing with Telkom Kenya
The second milestone is we signed our first commercial agreement to provide telecommunication services in Kenya to a global network operator called Telecom Kenya.
Alphabet really believes in this idea, really believes in the technology, in the business plan, and they really expect us to succeed.
We had previously proposed connectivity, but the work we had done so far had all been focused on temporary solutions for disaster recovery. What we’re doing in Kenya is longer term.
‘We all love GEO satellites, but...’
We all love GEO satellites. They’re great at providing coverage, they are great if you want to also have a regional solution. You can target specific countries.
Of course, they have some challenges in terms of latency, in terms of capacity density. By this I mean the number of bits you can deliver per square kilometer. This is why people started looking at lower- altitude solutions. Constellations in jMEO, LEO, and in some cases even BLEO — very low-Earth-orbit constellations.
Those are great at improving latency and capacity density, but you have a tradeoff. The coverage decreases, and at some point it becomes very difficult to deploy a regional solution.
Each Loon balloon flies at 15-20 km and has a footprint of 5,000 square km, with a direct LTE link to user terminals.
We think about Loon as basically a high-altitude pseudo satellite. The ITU [International Telecommunication Union] calls it High-Altitude Platform Station. The concept is that by flying a lot closer to the ground, and at approximately 20 km as opposed to at least 200-300 km for satellites, we’re able to have low-latency, very good capacity density and we’re able to deploy a regional solution.
For us to have global coverage, obviously we would need to have a very large number of balloons. But if we start by looking at this one country at a time, this is the more realistic.
‘The launch vehicle’
To relate this to the satellite industry, I’m going to try to use satellite technology.
We have a launch vehicle. It has side doors that you can roll down, it has wheels at the bottom, you can orient it so that we’re always downwind.
We built the structure because otherwise, launching balloons is difficult. It requires a lot of people, waiting for a time when the winds are quiet. With this kind of structure we can launch basically at will, and we can launch a balloon in less than 30 minutes.
‘Satellite bus and payload’
We have orbit-raising, and we have a thruster — a balloon. The outer dimensions of the balloon are about 15 meters by 12 meters. It’s composed of two envelopes — a wet gas inside and there’s air. By changing the amount of air inside the balloon, we’re able to change the altitude. So we’re able to go up and down.
We have a satellite bus and a payload. Our satellite bus looks very similar to what you’d find on a typical satellite. We have solar panels, avionics systems, batteries. We have controls. The controls are a way to pump air into the balloon or release air out of the balloon. And a way to orient the solar
Loon’s Kamenetsky was kind enough before a satellite audience to use satellite terminology in describing Loon. Here’s “de-orbiting.”
panels and antennas such that the solar panels are tracking the sun and the antennas are pointing in the direction we’re interested in.
The payload is again similar to a satellite payload. We have backhaul links that we use to establish communications with other balloons as well as communications down to our ground stations. For user links we’re flying LTE.
Today on average a balloon can spend three to four months flying in the stratosphere, although we have had flights of approximately 200 days and we believe that average is going to keep on increasing. End-of-life for us means that once we detect the balloon is ready to come down, we navigate it to a recovery region. We have a number of such regions around the world. We send a command to pierce the envelope. The lift gas is released, a parachute is deployed, the balloon comes down, is recovered, it’s recycled and reused as soon as possible.
Navigating in the stratosphere
We navigate. This is where we are a bit different from satellites. A satellite has a very predictable plane of orbit.
We navigate only by using the wind. The principle here is that in the stratosphere at different altitudes – we fly approximately from 15-20km altitude – the winds blow in different directions. You can imagine these different wind vectors pointing in all sorts of directions.
If you knew where the wind vector was pointing at each altitude, right now, and if you had a prediction of what would be happening in the future, then you could utilize that to navigate where you want to go, and to try to stay in one place. And the way you do this is simply by going up and down in altitude.
This requires a knowledge of where the winds are blowing. That’s a fairly complex problem. And people have looked a lot at weather, but typically they look at weather where people live, where airplanes fly.
We’re twice as high as commercial aircraft. So the amount of data for that kind of altitude is somewhat sparse. So we leverage the best third-party data we can find, but we also integrate our own data from flying in the stratosphere to be able to predict wind patterns.
We have two launch locations right now, our primary commercial launch location is in Puerto Rico. And in this case the service region is in Peru. We were providing services in Peru after some pretty serious floods there last year. So in this case you can see the balloon in Puerto Rico. It takes about 12 days to transit to the surface region. Ultimately it stays 98 days. It doesn’t stay perfectly still. It moves around, but it navigates back. Ultimately it went south and it was recovered in a region we have in southern Peru.
Selling a service, not a fleet of balloons
We don’t sell individual balloons or a service from individual balloons. We use clusters of balloons to provide services. Each individual balloon doesn’t stay in place. It might navigate around. But if you have a cluster of balloons and you’re able to control them as a fleet, then you can provide continuity of service. And from the perspective of the user, it doesn’t matter which specific balloon is providing service to that user at any given time, as long as we have control over the fleet.
Another thing that’s extremely important to us is the concept of mesh networking — using millimeter- wave technology. The reason this is important is because we are trying extend as far beyond existing fiber infrastructure as we can. We want to minimize the amount of additional ground infrastructure we have to put in place to deploy our services.
We were able to demonstrate a link with several balloons. Basically having hop-to-hop transmissions in this network. The total span was 1,000 km. The distance between the individual links that we demonstrated was over 600 km.
You can imagine that the way we provide services is by having balloon-to-balloon, inter-HAPS or inter- satellite in terms of technology link, and by having each balloon that also light up coverage area underneath.
The services we have been providing to date are infrastructure- or network-as-a-service to our telco, or mobile network operator partners. The typical scenario is there is a telecommunications provider who is trying to extend the reach of his network. And trying to cover places that are otherwise a little bit too remote for them, maybe the terrain is a little too rugged, the population density is a little lower than they would prefer it to be.
Loom connectivity is provided directly to the user’s handset. We’re using LTE. There is nothing a user needs to do to connect to a Loon. In the same way that you don’t know which cell tower you’re connected to, users within Loon don’t know which balloon. In fact they don’t know they’re in a Loon- based network.
The signal from there traverses our mesh of balloon-to-balloon hops. It terminates at ground stations that we have, from which there are IP set tunnels connecting back. We look very much like a roving tower partner, which they’re very familiar with.
A combined satellite-Loon network
One thing we’re exploring is: Can balloons provide services together with satellite connectivity? So here’s a hypothetical scenario. Right now let’s say you’re deploying coverage using satellite. This
Loon LLC wants to work with satellite fleet operators on hybrid networks.
could be a satellite constellation, could be a GEO satellite but there are places where you may not have sufficient coverage, there are places where you may not have sufficient capacity. You’ve got to envision a scenario where Loon works together as a solution with satellites. We’re able to extend this extremely high-data-rate point-to-point mix — over a gigabit per second of links to date. And that’s going to be increasing.
We also have extremely low latency, which is difficult, and in GEO obviously it’s just not possible. But you don’t have the same availability as a satellite system, simply because the balloons move around. So you can envision a surface where the balloons provide very-high-data-rate, point-to-point links and there’s a satellite service backing it up, such as you always have some amount of capacity.
Well, we have the network control system as well. There’s another area where we are actually very ??? doing an aerospace network, and specifically a non-geostationary aerospace network. When you think about mobility typically people think about mobility at the access layer. That means a user is moving, a plane is moving, a ship is moving. Mobility in networks is much more difficult when it’s happening at the backhaul layer. The balloons are moving and the network typology is changing, but the same thing is happening in non-geostationary satellite constellations.
There were a couple other projects we had. One was a satellite constellation pocket that’s not with us anymore. We were also working on other high altitude platforms at the time. We decided to solve the problem in a very general way. How can we orchestrate networks not just of balloons but with all kinds of flying, moving things together with terrestrial nodes?