Computing Clouds in Orbit – A Possible Roadmap
Last week I predicted that a large chunk of the internet and most cloud data centers would launch into space within the next ten years. Today, the only part of the Internet in space is a very small amount of bent-pipe access: signals that go from a user to a satellite and bounce back to a ground station that feeds them. in the terrestrial internet where all processing is complete and all queries answered by servers connected to the internet, many in cloud data centers. Responses follow a reverse path through a ground station, back to a satellite, and then back to the user. Below is a possible roadmap to the orbital Internet; the reality will certainly vary.
1. Starlink and Iridium prove the practicality of Internet access service based on satellites in low earth orbit (LEOS) – Finished
See: Internet and cloud go to space
2. Access to elbow pipes via LEOS creates a huge market in orbit
On the supply side, Starlink will go from 1,500 to at least 44,000 satellites. OneWeb, a European competitor, has launched 54 satellites and will be marketed this fall. Amazon’s Kuiper will be live in the next few years. Starlink will operate laser-based satellite to satellite and will be able to provide service to most areas of the world where it does not have ground stations.
On the demand side, all users of obsolete, slow-responding geocentric satellites convert to LEOS as fast as they can. Although fiber currently offers faster access than satellites, fiber is slow to deploy and may never reach the end of the road. Fiber cannot provide mobile access, which is required by both people and the Internet of Things (IoT), which will soon include all cars. Iridium is already providing mobile access, as will Starlink soon. 5G is the main competitor in this market. For emergencies, no terrestrial solution is adequate for communication. Poles and towers are subject to the same disasters as the people who depend on them locally.
Starlink says it has contracts to backhaul remote cell towers not over the fiber optic network to the Internet backbone. It is the first but will not be the last example of a cell phone acting as a concentrator and a distributor of traffic passing through LEOS.
Technology will increase the capacity of the LEOS service and competition will lower prices.
3. Caching for and in orbit
Caching in terms of the Internet means storing replicas of frequently accessed information close to the consumer of that information to speed up response times and reduce overall communication costs. Each time you click on a URL, a request is directed to a domain name server (DNS) somewhere to find the physical Internet address of the website your query is directed to. For example, google.com converts to 188.8.131.52. At least part of the DNS directory will be quickly cached in all ground stations. Large ISPs often host their own domain name servers to increase responsiveness; Starlink will be no exception. I’ll be amazed if Starlink doesn’t start caching DNS directories in Access Satellites shortly. Users will benefit from high responsiveness, and Starlink will save an exchange with a ground station for each truncated request.
Companies like Akamai and Cloudflare operate content delivery networks (CDNs). On behalf of content owners, CDNs cache copies of fairly static content (movies, for example, but also many other types of web pages) to locations on the Internet. It is a form of hosting that saves content owners from owning huge data centers with huge pipes themselves and ensures that content is quickly accessible from anywhere in the world including every owner of content cares. Whether Starlink will operate its own CDN or partner remains to be seen; what is certain is that terabytes of content will move through space to be “close” to satellite access users. At this point, we’ll see the first dedicated cache satellites. The access satellites will interrogate them by laser.
4. Intelligent routing in orbit
Once satellites can communicate with each other, they become routers and can manage QoS and dynamically optimize routing to some extent, just as terrestrial routers do. If a request can be answered in space by a cache satellite, the request will go there and get a very fast response with no bouncing in the terrestrial internet. If a request is to be sent to earth, it can be routed with one or two satellite hops to a ground station co-located with a data center that can handle some or all of the requests.
5. LEOS are gaining speed
With intelligent routing, caching, and content delivery hubs in space and at ground stations, a request sent through LEOS will often get a faster response than the same request sent over the terrestrial internet. All packets pass through a network of routers to reach their final destination. Each router adds a delay to the packet path for processing and queuing time. Each satellite can carry traffic to any other satellite with a maximum of four hops, usually less. If the packet is then served from a spatial cache or a data center with a co-located download station, which will be most data centers in a few years, there are fewer hops and more alternatives. to deal with congestion. The speed of light is also 50% faster in space than in fiber, but this is not as important for response time as reducing the number of hops.
6. Peering in orbit
Once competitive networks are firmly established in space, satellites from Starlink, Kuiper, OneWeb and other operators will begin to exchange packets with each other by laser; this type of traffic exchange, called peering, is already a common practice among terrestrial ISPs, even fierce competitors. They don’t do it to be nice; they watch because of Metcalfe’s law: the value of a network is directly proportional to the square of the number of ends. ISPs earn more by combining their networks than by keeping them separate. The same is true in space.
7. Terrestrial aggregation
People like high frequency traders and very serious gamers, for whom every millisecond counts, will start using LEOS to access it even when living in areas with fiber. Fiber operators will begin adding routes directly into space from their networks.
Mobile apps (think automated cars) that require a quick response will mostly be connected through cellular networks, unless space technology has evolved fast enough that they can connect to LEOS (which is possible). Even cell towers located on the fiber backbone will begin to communicate directly with LEOS to better serve their traffic.
8. Data centers in space – the cloud in orbit
In five years (I usually underestimate the time) the major data centers will be in space for simple economic reasons. The location of the data center depends on where the traffic is, the local price of power to run the data center, and the air conditioning that comes with it. Within five years, a high percentage of requests will pass through space; solar power is free once you pay off the capital cost of the solar panels and start it up; air conditioning is not necessary in the space. The physical bulk of a datacenter without air conditioning and designed for weightlessness will be relatively easy to put into orbit.
Security concerns alone are enough to make governments and businesses want to replicate key commands, control and data beyond the reach of physical attacks on the ground. Amazon is the world’s largest data center operator; they will move quickly in orbit; cloud providers who do not offer an orbital position will be at a significant disadvantage.
9. Computers designed for orbit
Currently, computers are designed to work on earth. Their speed is limited by the speed at which electricity can flow through their circuits; the energy lost during transport becomes unwanted heat. A chip designed for use in space can operate at temperatures close to absolute zero. At these temperatures, many materials become superconducting; they provide almost no electrical resistance. Computers in an orbital data center will be faster than their terrestrial predecessors.
If there is still mining of cyber-currency, it will be done in orbit where no energy used is polluting, and the calculations can be done more quickly than elsewhere.
Especially with increasingly stringent environmental controls, it will be difficult to justify building another data center on earth!
10. The new backbone, if needed, is built in orbit, not under the ocean
With spatial caching, orbiting data centers, and traffic relayed into space from mobile sources and aggregators as well as individuals, there will be no increasing demand for the earth backbone. Just as the road network has disrupted railways due to greater routing and dispatch flexibility, orbital routing and processing will reduce the demand for long-haul fiber. Even when communication is between two land locations, the shortest and cheapest route will usually be through space. Of course, a New York-London flight is a good way to get traffic between these two cities; but, if you are going from Minneapolis to Birmingham, do you really want to log into the hubs twice? Your packages either. Space is the area of the most direct connections.
This roadmap is just to demonstrate that there is a credible way for the internet and cloud computing to become primarily orbital. Surely it won’t happen exactly that way and might not happen at all. Since each step above reduces the cost of IT and communications, each of these steps – at least the ones that actually happen – will present huge opportunities for innovation and entrepreneurs.