Competition for global voice

The contenders

Inmarsat Mini-M and GAN

As an international treaty organisation which countries joined, Inmarsat has held a lot of clout (though it's now transitioned to a normal privatised company and lost the .int). Inmarsat has begun deploying Global Area Network - 64kbps to a laptop-sized terminal, such as this terminal from NERA. The blueberry iMac of satellite communications? Thrane & Thrane produce similar hardware.
64kbps is comparable to the older Inmarsat A/B/C services that require larger terminals; not quite broadband. Related equipment from Livewire Digital.
ICO seems to be off to a relatively late start, so Inmarsat has launched new GEO satellites with stronger spot-beams to provide a new Inmarsat Mini-M service with smaller, lighter terminals, shrinking the briefcase-sized Inmarsat-M terminal down to the size of a laptop computer.
This launch completes the set. (Florida Today Space Online, January 1998)
The time delay to GEO orbit and back of a quarter second means that the perceived voice quality isn't necessarily as good as LEO could offer, but right now these are the best you can get, and since GEO is well-established Mini-M doesn't have the regulatory problems that plague the other voice telephony schemes, or big LEOs.
Nera's own WorldPhone; the Inmarsat 2000. Skanti's maritime Mini-M phone. Thrane & Thrane's maritime mini-M/C phone.
Voice codecs from Digital Voice Systems.
A recent Inmarsat satellite launch (Florida Today Space Online, 18 December 1996)
Planet One also uses Inmarsat-3 satellites.
Inmarsat has a wide range of other services, including their new messaging D+. SkyWave Mobile will sell you a terminal for it. See also Inmarsat Ventures.

The super-GEOs

A number of 'super-GEO' schemes consisting of a few large and powerful GEO satellites to provide regional coverage to hand-held phones are under construction, mostly in the Far East. These take the Inmarsat-M concept to its logical limit.

A leader here is ACeS (alternate DNS entrypoint), built by Lockheed Martin. The ACeS Garuda satellite (with very large antennas) was due to launch in September 1998, but the launch was delayed a number of times when the Asian markets crashed. It finally launched on 12 February 2000 and is now being tested. The Analysys database has a profile of ACeS.

Other regional GEO systems include Satphone, ASC (Agrani), Thuraya (Analysys profile and overview of Thuraya from Boeing, now owners of Hughes Space. Thuraya is based on GMR-1. Thuraya accessories from Geonix.), APMT (also a Hughes contract) and Matra Marconi Space's EAST (which was claimed due for service in 2002).
A consortium covering EAST and ACeS announced the joint Geostationary Mobile Satellite Standard.
Some regional GEOs are already operating - such as AMSC over the US. And then there's Inmarsat...

The late starters

Ellipso

The official Ellipso website. They're now investing in an Internet startup called Ineva along with Virtual GEO, and retargeting some of their capacity towards the burgeoning satellite radio market with Penguin Radio.
As of March 2001, they're also now partnering with ICO; ICO is a lot bigger than Ellipso, and got a lot further along. Here's a summary of the partnership announcement from the Seattle Times. I've no idea what the partnership involves, but Ellipso did gain an FCC frequency license, and frequency real estate is considered a valuable commodity -- unless you believe in opn spectrum.
Ellipso planned to give up global coverage to concentrate on the parts of the world likely to provide a market. Using bent-pipe transponders, it's just more big, expensive amplifiers in space. Nothing radical here either - use of larger (Molnya, Tundra) elliptical orbits is well-proven for transmission of Soviet television.
Boeing invests in Ellipso; launching Iridium and Globalstar and building Teledesic weren't enough, obviously, but oddly Ellipso plan to launch on Ariane. Apparently, Boeing was considering taking a controlling stake in Ellipso (Wired News, 30 April 1999) although plans have been shelved.
Ellipso passed the major hurdle of getting FCC approval quite late, as detailed in this FCC authorization of July 1st 1997.
They have some patents on their orbit choices: US Patent 5788187, US Patent 5669585 and US Patent 5582367. These were based upon John Draim's patents on a four-satellite elliptical constellation for near-global visibility (4,854,527 and 4,809,935).
There was a graduate study investigating the impact of orbital debris on the Ellipso constellation; here's a presentation.
Satellite construction by Orbital Sciences.

Constellation (Aries)

The now-offline official Constellation website had some information about the expanded LEO system that they filed for with the FCC. The equatorial component of Constellation did not need FCC approval, as, being LEO and equatorial, it doesn't cover the US. Constellation finally received an FCC authorization on July 1st 1997. According to that, Raytheon would provide financial support. The FCC licence allows an additional seven planes of six satellites (five active plus one spare), in addition to the twelve equatorial satellites (eleven active plus one spare).

ECO-8

Here's the official ECO-8 website. This is very similar to Constellation's equatorial plane - 11 active satellites in an equatorial ring, although their coverage map shows nine... In fact, at one point ECO-8 was called ECCO and was associated with Constellation; I've no idea what, if anything, their current relationship is. You could think of it as Brazil's own satellite constellation proposal.

Courier

The official Courier website - this is a 72-satellite proposal from Satcon GmbH, and not to be confused with the Russian Courier-1/Konvert messaging proposal. Their website claimed that their system is voice-oriented, that they plan to have more active satellites than the famously unsuccessful Iridium, and that they're due to start launching in 2000 when even ICO has trouble getting funding.
Courier is probably best described in a chapter in Das Handbuch der Satellitenkommunikation (Amazon.de), partly written by someone associated with the project. I will be extremely surprised if this system gets built, given the current financial and market climate.

Discontinued

Odyssey

Odyssey is no more, as of 17 December 1997. TRW will be contributing to ICO after being unable to get funding. Odyssey as envisioned is described by the NASA Mike's Spacecraft Library.
It was part-owned by Teleglobe, investors in Orbcomm, and TRW.
Technical summary from the Big LEO tables.
Odyssey is most famous for the legal actions it has been involved in. TRW filed suit against ICO, since ICO uses very similar MEO orbits and the US Patent Office let Odyssey patent, amongst several other things, the orbit, in US patent 5,551,624, US patent 5,439,190 and the earlier US patent 5,433,726. There are also US patent 5,415,367 and 5,415,368. (This is the same US Patent Office that let Comptons patent multimedia and then took some time to reject the patent. They're not awfully bright. More details in this Wired article)
Patent an orbit? Heck, patent gravity while you're about it, since no-one else has. On 10 May 1996, TRW finally filed a lawsuit against ICO. TRW claimed that agreements between TRW and Inmarsat prevent Inmarsat from making its own use of information supplied by TRW, including orbital radii.
Here's a reasoned view of the implications of that patent. The ICO/TRW furore was resolved with the announcement that TRW was to become a major ICO shareholder (this was prior to the Teledesic takeover of ICO). TRW cross-licenced all those pesky patents and gave up the frequency spectrum allocated to it by the FCC.
TRW was awarded a European patent for the Odyssey handset (EP 648 027, December 1997)
simulate Odyssey with SaVi.

Competition for global broadband data

Anything that promises broadband networking worldwide.

Most of these satellite systems want to use Ka-band for ground-to-satellite communication, and their details are available from the list of accepted FCC applications. Here's a summary of the regulatory status of these Ka-band applicants.

This Network Computing Special is a detailed snapshot of the broadband market. (Network Computing, 15 March 1998)

Newer filings are going beyond Ka-band to V-band and up; with the recent close of the FCC window for applications for use of the 40GHz range, a rash of new systems have been announced and filed for, including Hughes' Expressway, StarLynx and SpaceCast, TRW's GESN, Loral's CyberPath, Inmarsat's Horizons, GE Americom's GE*StarPlus, Orbital Science's Orblink, extensions to Ellipso, PanAmSat's V-Stream, and a system from Lockheed-Martin.

If these all get built, I'm a banana. High frequencies like 40GHz would seem more suitable for wireless local loop, rather than for high-power satellites.

Non-GEO broadband schemes

Operational

O3b

MEO system. I wrote a paper on O3b.

Not operational

Laser Light Communications

Less likely to become operational is the MEO Laser Light Communications. See their introductory presentation.

SkyBridge (Sativod)

Here is the archived official SkyBridge website.
Sativod (satellite video on demand), later renamed SkyBridge, proposed by Alcatel, was a 64-satellite Ku-band LEO constellation concept. They've now announced that they're planning on eighty satellites in a redesign. (Alcatel press release, 1 June 1998)
In March 2001 it was reported they'd be looking at services over geostationary satellite in the interim, while in January 2002 reports surfaced that SkyBridge was on hold (ComputerWire, 4 January 2002).
Here's an introduction to Skybridge from Alcatel.
Links to archived MPEG videos (1, 2, 3) of their latest 80-satellite design.
Alcatel have filed with the FCC as the first step towards being allocated frequency spectrum (company press release, 1997).
You can simulate Skybridge with SaVi.

Sterling Satellite Communications (s2net)

Their official homepage will tell you that they want to do IP switching onboard satellite. Very ambitious claims. Apparently they're developing coiled tethers to gain power from sweeping through the Earth's magnetic field.

Orblink

Orblink, from Orbital Sciences, is the first commercially-planned MEO broadband system I'm aware of, although it was later followed by the additional Spaceway filing and others; a ring of satellites with ISLs at 9000km altitude, which is an odd choice given the location of the inner Van Allen belt. (press release, 25 September 1997)
Here's a description of the proposal. (Wired News, 25 September 1997)

Pentriad

The official Pentriad website would appear to be no more. Pentriad is a proposed primarily V-band constellation using thirteen satellites in highly elliptical (HEO) Molnya orbits over the Northern Hemisphere. They also want to provide wireless infrastructure for HALO. Right now, the proposer, Denali Telecom, is one Dennis Burnett of Pierson and Burnett.

Virtual GEO/VIRGO

Here's the official Virtual GEO website. They're now investing in an Internet startup called Ineva along with Ellipso, apparently. Elliptical orbits, wanting to share Ku-band with GEO - which is Skybridge's argument. They have two patents on the orbital choice: US patent 5845206 and US patent 05957409, continuing Draim's orbit work as used in Ellipso.

contactMEO

Here's the official @contact website (alternate DNS #1, alternate DNS #2). It's proposed by the people behind WildBlue/iSky/Ka-Star/whatever.
This has now morphed into the contactMEO proposal.

Subsumed into Teledesic

M-Star

M-Star was the Motorola followup to Iridium, at least until Celestri was announced. They asked for FCC approval, and they thought it would take four years from getting approval and six billion dollars to build this 72-active-satellite constellation with laser intersatellite links and millimetre wave, rather than Ka-band, to ground. Cutting-edge stuff.
This wasn't a global scheme - it's between 57 degrees of latitude only. It was aimed at providing wireless connectivity to mobile base stations, or a wireless infrastructure to the wireless infrastructure, as well as services to fixed terminals. M-Star was revamped and remarketed as the LEO component of Celestri.

Celestri

Motorola's M-Star was itself superseded by the less ambitious later-announced Celestri. (Wired News, 19 June 1997)
On 21 May 1998, it was announced that efforts on Celestri and Teledesic would be merged. Celestri was intended to use Ka-band rather than the millimetre-wave of M-Star, but both apparently were to use laser intersatellite links. Celestri was said to have a GEO component; I have no idea if that's a reinvention of Motorola's own GEO Ka-band Millennium GEO scheme (as discussed in the FCC Millennium filing), and the GEO component is not mentioned in Motorola's FCC filing. That's marketing for you.
The domain name of the official Celestri website has been recycled.
You can simulate Celestri with SaVi.

Proposals

GIPSE

was a study to design a satellite constellation capable of providing broadband multimedia personal communications. It's one of a number of satellite constellation studies undertaken within Surrey.

Rostelesat

A Russian constellation proposal. CommunicationsWeek International described it as a combination of LEO and MEO constellations - 7 planes of 13 satellites apiece at 700km altitude (the N-system) and 4 additional planes of 6 satellites apiece at MEO (the V-system). (CommunicationsWeek International, 20 July 1998)
Further information appreciated.

GEO broadband schemes

There are a number of schemes that, though not strictly constellations, and in some cases not strictly broadband, do provide or plan to provide Internet over satellite from GEO. Such schemes include DirecWay, DirecPC from HOT Telecommunications, Gilat-to-Home (or Starband as it's now called), Web-Sat, Internet via (formerly in) the Sky from Europe Online and Tachyon. Yes, the Internet does work over GEO satellites.

Spaceway(Galaxy)

Hughes have used the name 'Galaxy' for many things over the years, including most of their satellites, so Spaceway was known as Galaxy or 'that Galaxy project' by third parties like the FCC.
It's a geostationary constellation which has all but one satellite linked with inter-satellite links (ISLs). It's a very big ring. The geostationary ring is well-established; Hughes later filed for an additional medium-earth-orbit constellation of twenty active satellites called Spaceway NGSO.
Launches were planned on Sea Launch (Hughes press release, March 2001).
Some extremely outdated third-party technical information.
The details of this project have changed slightly since the paper describing Spaceway mentioned on my References page was published.
Hughes later filed for a scheme called Expressway with the FCC - 14 GEO satellites in ten locations, linked with optical ISLs, with an estimated cost of $4billion. This does sound rather like Spaceway reinvented - but it's apparently Ku-band and the very high-frequency V-band rather than Ka-band, so this looks like a bid to gain more frequencies if more capacity is needed.

Astrolink

Here's the official Astrolink website.
Astrolink, from Lockheed Martin, was planned as a nine-satellite GEO constellation with intersatellite links. Apparently whether Lockheed should build or bid on a constellation was hotly debated internally. They apparently went down to four satellites and now appear to have funding problems. Rather like everyone else.

Cyberstar

Cyberstar, from Loral, also planned ISLs.
Here's the official Cyberstar website.
Apparently Cyberstar will now be marketed alongside Skybridge via a partnership with Alcatel; the Globalstar/Cyberstar/Skybridge partnership stacks up against Iridium/Celestri/Teledesic and the two broadband systems are expected to be combined into a single company. (Via Satellite, August 1997)
Cyberstar appears to have been repurposed to the short-term as services via existing satellite transponders, although they did say they were still planning to build Ka-band satellites.

Euroskyway

Euroskyway is a five-GEO satellite constellation from Alenia Aerospazio. There's an official Euroskyway website. VIP-TEN was an project with local participation looking at voice over IP telephony interworking with Euroskyway.

GESN

According to a press release of 4 September 1997, TRW has filed with the FCC for a high-frequency data system - four GEO satellites and five MEO satellites, apparently, due to start in 2005. This looks like a response to Hughes' Expressway filing, requesting similarly high frequencies at WRC '97 - but TRW didn't get sufficient funding to build Odyssey. How will they fund this?
The working name is the TRW Global EHF Satellite Network, or GESN for short - catchier than TRWGEHFSN, certainly. Fifteen MEO satellites at the same altitude as ICO and Odyssey, and four geostationary satellites.
They've since filed again and said it's really a Ka-band system.

WildBlue/iSky/KaStar

WildBlue used to call themselves iSky and before that they were KaStar They appear to have forgotten about www.kastarcom.com.
Last I looked (which was a name or two ago), they were planning a two-satellite GEO system with ISLs, but wanted the financial requirement waived. They're concentrating only on the US.
The people behind KaStar or whatever-it's-called-now also filed for @contact, a MEO constellation with planned coverage between 65 degrees of latitude.

Aster

Aster, from Spectrum Astro Inc, planned a five-satellite GEO system with inter-satellite links. (company press release)

SWANsat

SWANsat claim to be planning a number of high-capacity geostationary satellites, based on a frequency license from Nauru. They previously had a frequency licence from Tuvalu but gave it up. They want to use the high-frequency W-band, whose use from orbit has yet to be proven viable.
SWANsat is operated by Charles and William Welty. Back in 1999 Koinonia House enthused about the idea. William Welty claims to have been awarded a PhD in 2005 from the unaccredited Louisiana Baptist University; the SWANsat business plan formed his "dissertation". Charles Missler, the founder of Koinona House who enthused about the idea in 1999, signed off on that document.

V-Stream

V-Stream, from PanAmSat, was another GEO broadband system planned to operate in V-band, with twelve satellites (company press release).

WEST

WEST was a combined MEO-GEO Ka-band proposal from Matra Marconi Space (now Astrium). Since Matra were building Celestri and were then helping with Teledesic, WEST looks unlikely.

Cancelled constellations

Voicespan

Voicespan was from AT&T. Here's a report of their initial announcement. (Edupage)
Voicespan is a trademark of AT&T modems.
AT&T's FCC application has been withdrawn. Here's their own announcement of withdrawal. (AT&T press release)

The not-a-satellite-in-sight schemes

Global fibre

Undersea fibre nets

Colossus from Alcatel may be a worldwide undersea fibre net rather than a satellite constellation, but it could take a bite out of the same market, especially when you think of the narrow network links between countries at present.
Project Oxygen is a similar undersea fibre net, with Lucent as preferred supplier.
Africa One is an effort to ring the coastline of Africa with fibre for cheap connectivity; it's a cooperative effort between AT&T and Alcatel, and Kenya is enthusiastic about the idea. The Africa One website wasn't very informative. There was also Southern Cross.
It's interesting to compare these to Buckminster Fuller's idea of a worldwide network for distributing electricity, since information and the electricity to store it go hand-in-hand.

Concert

Concert is a rather more simple plan to upgrade the internet and improve international hubs. Unambitious, but likely to have more real effect than many of the other schemes on this page. Here's the official Concert website, the ideal home for you to vent all that pent-up hatred left over from the untimely demise of mother Death Star, AT&T.
Concert's demise as an entity was announced on 16 October 2001. But at least it had customers, which is more than you can say of a number of constellation projects...

High-altitude communication

Stratospheric platforms

Sky Station

Here's the official Sky Station website.
Sky Station has nothing to do with satellites or orbits. It's based on balloons (or aerostats, for the technically inclined), it has filed with the FCC (here's their FCC frequency petition and interference predictions), and prototypes are underway.
These schemes are geostationary without being as far away as GEO or LEO - think of the blimp floating above the city that bounced signals down to Dick Tracey's radio watch, and you've got the idea. They're planned to be at least 10km up.
There are problems, though. You need many more balloons than satellites -- although balloons can be deployed incrementally as coverage is extended, while LEO/MEO constellations must be completely operational to give full coverage all the time. You're in atmosphere, which is a much tougher environment than space (for a start, it's colder - now you have conduction and convection as well as radiation) with a whole host of how-do-you-solve-these? engineering problems.
The people behind Sky Station talk about it. (Wired, September 1996)
Sky Station has already filed with the FCC. Here's an FCC ruling resulting from the application. (May/July 1997)
A case for allocating frequency to Sky Station at WRC '97
An introduction from TELE.COM, mentioning Alfred Wong.
And to think that Teledesic sounded crazy, until this.

In Europe, the Helinet project is looking at stratospheric communication, and the University of York has now spun off SkyLARC.
Skylinc is pursuing this commercially (BBC News, 26 May 2003)
ATG is pursuing this commercially. (BBC News, 27 February 2002).

Take a look at the description of RotoStar, another proposed balloon-mounted telecommunications platform.

Space Data is proposing flying cellular transceivers on weather balloons. Article from TechTV. Introduction to HAPS from Surrey.

Sky Station? Angel? This brings back memories of Captain Scarlet and the hovering Cloudbase - spectrum is green!

High-flying planes

Charles Platt's Wired article provides an introduction to some of the systems.

HALO

Here's the official HALO website from Angel Technologies (alternate DNS entrypoint. Apparently their use of broadband.com is no more.)
Their Proteus aircraft, which is also the first stage of an X-Prize entry, has made its first public flight. Here's another report of the flight and a summary at the time of the flight from Wired News.
They're High Altitude Long Operation (HALO) aircraft - but they have pilots working shifts. (Remember that in Arthur C. Clarke's 1945 paper he suggested manned geostationary satellites - there could be some room for design improvements here, and unmanned aeroplanes aren't unheard of.)
I took a lot of flack from people who thought that Sky Station was a lot of hot air; I'll merely point out that the Board of Directors of Angel includes one Edward F. Tuck, who came up with Teledesic, the satellite constellation joke that predated the balloon joke that predated this joke - which was also based on US Dept of Defense technology that has been around for decades. (news.com, 12 September 1997)
And I can't wait to see what jest is in store next.
HALO is summarised by TBTF.
An alternative to HALO is the solar-powered unmanned Helios, based on the Centurion design from Aerovironment. Read their brochure; this has fed into SkyTower.
There are possible television applications. High-altitude television has been tried before - I'm told that "Stratovision" was an idea proposed in the late 1940s in the United States to put TV transmitters on aeroplanes and fly them nonstop when VHF capacity was limited - but in 1952 the UHF band was opened up to terrestrial television instead. (Broadcasting Magazine, August 1945)
In the sixties, MPATI, the Midwest Program on Airborne TV Instruction, flew a DC-6 from Purdue University's airfield, beaming educational programmes to schools in the region. MPATI ceased transmission in 1968.
And in South Vietnam, the US Army flew "Operation Blue Eagle" from 1966 to 1972. Aeroplane-based television isn't a new idea.
I'm indebted to Jim Schwoch for the above historical information.

Platforms International

There's an official website; here's a brief overview of Platforms International. Lower profile than Angel Technologies, but aiming at unmanned communications aircraft as well as military applications.
I gather this has risen from the ashes of Skysat, a similar, now defunct, company (further information). Same CEO, anyway. Information on initial funding (March 1998).

ESA is conducting a study into similar High Altitude Long Endurance (HALE) platforms. Work on high-altitude platforms for communication has been carried out at the University of York. There is also Heliplat in Italy.

Radio constellations

Yes, satellite radio. These days, it's not just a transponder feed to your local radio station so that it doesn't have to broadcast its own news or through the night itself. These players wish to broadcast direct to a satellite radio you own, even one installed in your car. This seems to be generating enthusiasm.

Sirius and XM Radio

Recent satellite-radio efforts include the equatorial XM Satellite (alternate DNS entrypoint, stock tracking) and the three-satellite Tundra/Molnya-ish constellation of Sirius Satellite Radio (formerly CD Radio, stock tracking) - here's information on Sirius's first launch, second launch, and third launch. Here's XM's first successful launch and its second from the Odyssey sea platform. Don't confuse Sirius Radio with the other Sirius satellite.
They're both aiming at the car industry - XM has signed with GM, and Sirius with Ford. Both have shipped receivers for their terrestrial networks for use in cities (where satellites are shadowed by street canyons), but satellite receiver modules are slow in coming. Here's a comparison from Forbes. And you wonder why cars cost so much... These constellations are both aimed squarely at the US, where cars drive long distances and can only pick up the two types of music. Country. And Western.
Interesting interference problems with a new type of lightbulb are raised in the Wall Street Journal. These lightbulbs could be used in cars too... (August 2001)

Ondas Media is trying to replicate the US satellite radio model in Europe.

Worldspace

Worldspace began as a way of getting digital satellite radio to developing countries to make foreign propaganda from the Voice of America sound better, but now the Internet is doing a much better job of promoting American cultural hegemony they're going to provide an internet service instead. Worldspace won't be licensed in the United States, according to the regulatory update for satellite digital audio. There's not much point in preaching to the converted.
1997 Forbes overview and 2002 followup.

Messaging constellations

Many people think of these little LEO schemes as dull, dull, dull - if they think of them at all. After all, they're not interested in how their meters might be read or in other mundane real-world applications. But if you're interested in the least glamorous, most immediately useful satellite constellation schemes, this is as good a starting point as any.

Here's a summary of the regulatory status of little LEOs and a summary of the players.

The FCC recently announced licensing rules for this category, and five schemes were seeking licences. Here's a summary of the applicants. (Florida Today Space Online, 10 October 1997)
Further local information on these systems is available.

Allocation of radio frequencies to little LEO satellites is considered a threat to amateur radio.

Orbcomm is the most advanced in deployment, so it gets its own webpage. A number of proposals that I know very little about are documented on Alex's small satellites webpage.

LLMS/Iris was interesting because it's a store-and-forward design initially utilising secondary payloads attached to other satellites, rather than purpose-built satellites. Here's a local overview.

Final Analysis

The official Final Analysis website. Final Analysis is also known as FAISAT.
A case for allocating frequency to Final Analysis at WRC '97
VITA is partnering with Final Analysis on a number of things - here's a VITA paper on little LEOs.

Leo One

Here was the official Leo One website. They're planning 48 active satellites providing store-and-forward messaging, and according to this press release they'd be launching on converted SS-19 ballistic missiles. (October 1999)
Status of Leo One from Analysys.
They hold US patent 5678175 on the design. Although the Leo One patent may appear to refer to a patent on urine specimen collection (U.S. Patent 5009235), I've finally learned that that's the result of a typo in the original patent; it should really be referring to this more relevant patent with a slightly different number (U.S. Patent 5099235) I find this slightly more amusing than Iridium accidentally registering the box around their original constellation logo as an essential part of the trademark as a result of translating the graphic between Mac and PC file formats, which caused the box to appear in the first place...

E-Sat

E-Sat was a proposed six-active-satellite constellation aimed for launch in 2000 that is now cancelled. DBS Industries, the company behind it, is focusing on remote meter reading applications. Surrey Satellite Technology (SSTL) was to build the platforms.

AprizeStar

AprizeStar is a proposed 48-satellite UHF constellation from SpaceQuest that is aimed at asset tracking and meter reading.

Proposals lacking funding

KITComm

KITComm is being launched by AeroAstro. KITComm is described. (Wired, September 1996)

GONETS

The official GONETS website. The constellation appear to have similar geometry to Globalstar and LEO One, judging from their system parameters.

LEqO

LEqO is a local proposal from Surrey Satellite Technology Ltd (SSTL), owned by the University, to launch a single LEO equatorial ring of eight microsatellites to provide email services. Again, funding welcomed.
simulate LEqO with SaVi.

Picosat

Picosat was a packet-radio pipe-dream constellation that wanted your funding. Not impossible, but very unlikely. Given the number of people who have even heard of packet radio, my money's on the schemes at the top of this page.
The site hosted by inuviktv appears to have vanished.

TropNet

TropNet, similar in intent to LEqO, is a proposal that comes to you from the same people responsible for putting the cross in space.

Cancelled constellations

GE Starsys

Starsys was originally planned by NACLS, but in June 1996 GE Americom completed buying 80% interest, meeting the requirements detailed in this FCC notice.
Alcatel obtained European distribution rights and was to build the satellites (press releases, July 1996).
In August 1997 GE Americom withdrew from the constellation race and handed back its FCC licence. I'm told GE Astro Space has since filed for a little LEO system; I've no idea if this is in any way related to Starsys.

Navigation constellations

Also used for surveying and anything where knowing your exact position is important.

Already active

GPS

Introduction to GPS from Trimble.
Good overview of GPS from Technology Review (July 1999)
Better overview of GPS from Peter Dana.
GPS is described by the NASA Mike's Spacecraft Library. This is a constellation and it's functional, but it's for positioning, not communication, it's high-altitude, not LEO, and it's been done to death in the media. Competitor? No. But an interesting complement, certainly. Think of the combination possibilities...
The GPS satellites are built by Boeing.
Operation information from the United States Naval Observatory. Lots of official information here.
The US Coast Guard has information on GPS and differential GPS.
GPS component information from ELIRIS.
GPS information and companies indexed in Yahoo.
Institute of Navigation GPS proceedings give an overview of the world of GPS research and development.
US Interagency GPS Executive Board.
Casio announce first GPS watch (BBC News Online, 7 January 1999)
GPS gadgets from GPS Warehouse.

Glonass

Glonass is described by the NASA Mike's Spacecraft Library. This is the Soviet equivalent of GPS, although people seem less than keen to rely on it in these changing times.
Index to Glonass information from the Onsala Space Observatory.

EGNOS is a geostationary navigation system using Inmarsat-3 satellites. Description of EGNOS from Racal Avionics. Information on esa navigation activities.

China is launching its own geostationary navigation system, called Beidou. Here's a report on the third launch.

Planned

GNSS/Galileo

Galileo, formerly known as GNSS-2, is the European Community's effort to create a navigation system that is in international hands and independent of foreign military powers. When you're thinking of autonavigation of planes, you want a system you can trust.
Eurospace's recommendations on Galileo.
Is Galileo necessary? Some people are quite happy to rely on the US military...
Surrey Satellite Technology has been selected to build the first test satellite, to a tight schedule dictated by the terms of use of the frequency license (Surrey press release, 11 July 2003). More information from the Surrey Advertiser, covering local news (18 July 2003).
US puts pressure on against Galileo (Wired News, 17 January 2002)
Europe resists pressure (Wired News, 18 March 2002)
It's expensive - will the EU pay for it? (BBC news, 7 December 2001)

Remote sensing

The remote sensing market has traditionally been the domain of single specialist dedicated satellites, be they individual remote-sensing LEO satellites from SSTL or the Triana political football dreamt up by Al Gore. Below are the first proposed remote-sensing constellations that I'm aware of. Orbital Imaging also controls two remote-sensing satellites and are to launch another two.

FUEGO

Here's the official FUEGO home page. A proposed constellation of twelve satellites devoted to fire detection; ESA calls it Forest Fire Earth Watch.

RapidEye

Plans a constellation of four polar satellites. Their official website might tell you that SSTL was contracted to build the satellites. Here's a summary of RapidEye.

Skymed-Cosmo

Here's the official home page, such as it is. Further information was available from HCSA, the Greek partner. Skymed-Cosmo is a preliminary study to examine building a constellation to provide remote sensing of the Mediterranean basis; odd, until you consider that Alenia Aerospazio build satellites and has capacity standing idle once the 64 Globalstar satellites are finished, and that they have remote-sensing experience.

Disaster Monitoring Constellation

A local proposal for disaster monitoring from Surrey Satellite Technology Ltd has been fleshed out to reality; the TMSat design forms the basis of the international disaster monitoring constellation (DMC). Images are available from DMC International Imaging.

Formosat-3

The Taiwanese government, through its National Space Organization, has a constellation of six satellites for remote sensing; this is known as COSMIC (Constellation Observing System for Meteorology, Ionosphere and Climate) in the US, and formerly ROCSAT-3. All six satellites were launched together on 14 April 2006.

Satellogic

Sixteen-active-satellite imaging constellation. Official website. Description from Satellite Today.

Use of remote sensing for both commercial and military intelligence purposes raises interesting questions.

Other

NASA's Tracking and Data Relay Satellite System (old site) can be thought of as a private constellation; it's used for communication by the Space Shuttle and the International Space Station.

Astrium's Grace is a two-satellite constellation (with an intersatellite link!) to do ocean and gravity monitoring. It was launched on 17 March 2002. Further information and a launch diary are available.

Suicidal satellite sponsorships

ICO's Everest website and their Branson balloon challenge are, alas, no longer online. But Iridium's all at sea in 2002.

On suicidal missions:
There used to be a long list of insane exploration/extreme sport/slogging over Arctic wilderness/nutty balloon things here, but, oddly enough, all the webservers telling you about these vainglorious schemes have committed suicide as the rent on vanity DNS names wasn't paid.
Phillipe Kahn races yachts with Iridium to hand (CNN report)

Iridium sponsored suicidal sportsmen before eventually killing itself in bankruptcy.
Falling back on paging at north pole because the Iridium phone got wet.

Crossing the Empty Quarter using Inmarsat phones.

And finally...