Monday, July 21, 2014

Low Earth Orbit Broadband Satellite

Google has generated a bit of press in the last week with their Project Loon and their funding of O3b Networks.  I for one am optimistic, but equally amused with the apt naming of a looney project. But internet from space is radio, and radio is something I know about. The big question remains.. is there any efficacy in  a space-bound ISP that can provide global connectivity? In a word?  Maybe.

These low earth orbit broadband satellite ventures are promising nothing short of Satellite Internet access. But in a way, that has been the promise all along. The first commercial communications satellite was Telstar 1, built by Bell Labs and launched in July, 1962. The first satellite to successfully reach geostationary orbit was Syncom3 which was launched the following year. Telstar 1 had only one transponder to relay data. It operated in the C-band (4 to 8 GHz ) receiving 6 GHz microwave signals and responding on 4 GHz omnidirectionally. This was not going to move a lot of data. It did relay television signals, it's orbit only allowed it to do so in 18-minute segments as it whipped over continents 3,500 miles above the surface. Telsar 2 was identical, but the Syncom series that began launching in 1965 were at least geosynchronous.

But the real excitement came in the 1990s. In 1993 the Hughes Aircraft Co.asked the FCC for permission to launch a satellite that transmitted in the Ka-band (26.5–40 GHz.) This matters because there is a direct relationship between frequency and bit rate. Frequency is just the number waves per second measured in Hertz (Hz). So a 100MHz signal can transfer 100,000,000 wave cycles of current in one second, but that's only 0.1 GHz. Bit rate is the number of bits of data transferred in one second. We measure this in bits per second (bs.) If you imagine each wave as a bit you can see how increasing the frequency increases the bit rate. (Though clearly encoding schemes also increase speed.) Consequently moving from 6 GHz to 26 GHz was pretty significant for data transfer.

In 1995, the FCC opened up applications for other Ka-band satellites. Fifteen companies applied including: EchoStar, Lockheed Martin, GE-Americom, Motorola and KaStar Satellite.. one of the first that tried to offer ISP service was Teledesic. They launched in 1998 operating between 28.6 and 29.1 GHz. 9 billion dollars and 8 years later it went bankrupt. The project was totally abandoned by 2003. But that same year Eutelsat launched 31A which provided broadband services in Europe. First called e-bird and later Eurobird, it was renamed in 2012. it's still in service. Anik F2 went into service in 2004 delivering a similar service to Canada, but was capable of a very impressive 140 Gbit/s with 114 transponders.  Of these 50 operate in the Ka-band, 40 in the Ku-band (12–18 GHz), and 24 in the C-band. That was a decade ago. In more recent years, ViaSat-1 and HughesNet’s Jupiter have proven capable of data rates of up to 15Mbit/s. That's better than I get from my cable provider. (ahem) There are now a slew of these high throughput satellites.
  • Anik F2 (2004) 
  • Thaicom 4 (2005) 
  • Spaceway-3 (2007) 
  • WINDS (2008) 
  • KA-SAT (2010) 
  • Yahsat Y1A (2011) 
  • ViaSat-1 (2011) 
  • Yahsat Y1B (2012) 
  • EchoStar XVII (2012) 
  •  HYLAS 2 (2012) 
  • Astra 2E (2013) 
  • O3b Satellite Constellation (2013) 
  • Inmarsat Global Xpress constellation (2013) 
So back to Google and O3b.  Those satellite broadband ventures of the 1990s and 2000s didn't fail. They've created a cache industry in rural and third world internet service. While Microsoft's funding of Teledesic may have gone nowhere, so did Windows ME, Vista and 8. Google has a better track record for holding and folding. They also clearly have the bank balance to launch and support more high throughput satellites. This isn't something new, this is more of something we have already proven can be done. I, for one, welcome our new Internet overlords.