Back in the late 1990s, Low Earth Orbit (LEO) satellites were thought to represent the future of the industry, and vast amounts of press attention were lavished on small LEOs, big LEOs and broadband LEOs. In some ways this was hardly surprising because of the huge amounts of money involved: $5 billion was spent on Iridium, almost as much on Globalstar, and Teledesic was expected to cost at least $10 billion. However, in the end these projects proved to be a sideshow from the key development in satellite technology during the latter part of the decade: the increases in satellite power and lifetime, which enabled traditional Ku-band transponders to be built and launched for roughly half the cost (in dollars per year of life) from just a few years earlier. As a result, when growth expectations changed after the bursting of the dotcom bubble, from 2001 to 2004 the satellite market experienced a significant downturn, not due to the failure of Iridium and Globalstar and the demise of Teledesic (which had almost no impact, apart from on investor sentiment), but because of overcapacity and dramatic price erosion within the core Fixed Satellite Services (FSS) transponder leasing business.

Now we are faced with much the same situation as OneWeb and SpaceX compete for attention with their proposals for massive LEO broadband constellations, while the industry (and press) tries hard to ignore the elephant in the room, namely the dramatic increases in capacity available from high throughput Geostationary (GEO) satellites. In just a few short years, we’ve gone from satellites with capacities measured in the Gigabits per second, to tens and now hundreds of Gigabits per second. More importantly, these higher capacity satellites are not just being devoted to consumer satellite broadband, like ViaSat 1 and Jupiter 1, but with Intelsat’s Epic program, and copycat spot beam satellites from SES, Telesat and others, High Throughput Satellites (HTS) are moving into more mainstream FSS markets, such as maritime and aeronautical services and business VSAT networks. We’re also seeing substantial price erosion, with Global Eagle recently declaring, after striking a deal to move its In-Flight Connectivity (IFC) services over to SES’s satellites, that it had obtained new, far more attractive lease rates with price cuts that were “not … in the 10 percent range or even the 25 percent range, significantly lower.”

Developing workable low cost terminals to track LEO satellites as they move across the sky — which was one of the factors that prevented Teledesic from moving forward — still remains a significant concern.

Meanwhile, a close examination of OneWeb and SpaceX’s proposals shows that many technical challenges remain to bring their plans to fruition. Developing workable low cost terminals to track LEO satellites as they move across the sky — which was one of the factors that prevented Teledesic from moving forward — still remains a significant concern. Moreover, extrapolating from the successful development of small imaging satellites to communications satellites is fraught with difficulties. The power required for a communications satellite to close the link with two-way terminals on the ground is much greater than for an imaging satellite, and implies that the satellite must have a bigger battery and solar arrays. Power requirements are driven up further by the need to use higher altitude orbits (800-1100km, rather than 350-500km), in order to achieve reasonable look angles to terminals on the ground.

The two biggest problems for Google Loon’s balloon-based connectivity system are firstly that balloons do not stay in the same place, so very large numbers of them must be launched, and secondly that the available power simply does not support a substantial throughput from each balloon. Small, low cost LEO HTS satellites will face the same issues, and even if substantial savings can be obtained by building satellites on a production line, predictions that these communications satellites can be built for a few hundred thousand dollars each are likely to prove far too optimistic. And that’s not to mention the challenges involved in simply finding enough launch capacity to get these new constellations up and running in the near future or the difficulties in building a credible business plan under which tens of millions of people will buy Internet access via satellite, instead of using terrestrial infrastructure. It’s hardly surprising that SpaceX characterizes its plans as a long term effort that may take a decade or more to come to fruition.

Above all, the biggest unanswered question is whether latency really matters in the end. It didn’t for Iridium and Globalstar, which have not gained a significant competitive edge from their lower latency over GEO systems like Inmarsat. And it didn’t for Teledesic, which faded away to be replaced by Ka-band GEO satellites. Existing broadband GEO satellites work well for providing Internet access to people in remote areas, as seen in North America and Europe, and for the foreseeable future, geostationary satellites will continue to offer the best price/performance ratio for satellite broadband. As a result, GEO HTS is likely to be deployed even more widely in emerging economies in the next few years.

However, the advances in GEO capabilities mean that the satellite industry faces the clear and present danger of a potential price war in two-way connectivity services. New entrants, especially those deploying Ka-band capacity, have set aggressively low prices in regions such as Africa, but Intelsat and others are fighting back. There is an ongoing land grab to lock up large anchor tenants for the new satellites that will be launched over the next few years, as seen in the intense competition (finally won by Eutelsat) to secure a contract with Panasonic to meet its IFC needs in the Pacific Ocean region.

It’s clear that service providers like Global Eagle can now commit to take a multiple of their existing contracted capacity on these new satellites, and receive a significant price cut on a per Megahertz basis, which means they only pay slightly more in absolute dollars. We saw a similar approach adopted by fiber providers like Global Crossing in the late 1990s, but that business unraveled when the backhaul market became saturated and early purchasers moved to resell their excess capacity to other service providers. It’s too early to tell if the same could happen in the satellite market, but if the growth in end user traffic, for example in IFC, does not live up to expectations, then problems will undoubtedly occur. Fortunately, most of the big operators are buffered by their large broadcast business segments, which are stable or growing. But for smaller operators, especially those reliant on two-way communications, the next few years could become very difficult indeed.

Overall, I would therefore caution everyone in the satellite sector not to get carried away with enthusiasm for new LEO projects, and especially not to take your eyes off the ball when GEO HTS systems are experiencing such rapid changes in both capabilities and pricing. These are certainly exciting times for the industry, but I hope that we can learn the lessons of the first Internet bubble, so we won’t suffer as badly when the next one pops. VS

Tim Farrar has worked as a technology consultant formore than 19 years, specializing in satellite communications and wireless spectrum. He leads his own consulting firm Telecom, Media and Finance Associates, Inc.