The Fixed-Satellite Service (FSS) industry is entering its fourth year of a downturn. It is well understood in the market that there is an over-supply of capacity and that prices are falling. Due to the rapid expansion of terrestrial networks and applications, our customers are reconsidering their future commitment to geostationary (GEO) satellites. We have seen customers such as DirecTV announce that they will shift all their future traffic to fiber and they will no longer be purchasing satellites. The “Big Four” (Intelsat, Telesat, SES, and Eutelsat) and regional operators are continuing to report a reduction of revenue.
Aside from our own oversupply of Geostationary Orbit (GEO) capacity, we see competition from the rapid expansion of Fiber, 4G and soon, 5G networks. This incredible explosion of bandwidth has led to the growth of Over-the-Top (OTT) and online video systems lead by YouTube to dominate the global internet traffic and viewership hours taking advertising dollars away from traditional broadcast media. Meanwhile, we are experiencing a wave of upcoming Low Earth Orbit (LEO) projects that are squarely aiming for our industry. While most of these systems have dubious business plans and face enormous financing, technical and regulatory challenges they nonetheless add to the dissonance on the future viability of our industry.
While things look bleak for GEO FSS operators, there are some signs of hope — a new application will emerge that could potentially ignite the growth of our industry. Thus, I say it is not time give up but to double down on our sector because there is light at the end of the tunnel.
The rapid growth of fiber and 4G systems have indeed occurred around the world, but investments of these systems require one key aspect — a minimum level of population or subscriber density. Fiber will only be rolled out in dense environments. For broadband 4G and 5G wireless to work, sufficient subscribers must exist to justify the cost of the equipment, monthly rental charges, and the large electricity bills run up by these power hungry networks.
If there are thousands, if not tens of thousands of users in a city or suburban area, then the CAPEX and OPEX of running broadband wireless system can be justified. However, if the areas to be covered are very sparsely populated, then the economics of deploying cellular networks cannot be justified, meaning vast areas will go unserved.
It is estimated that more than 10 million homes in the United States do not have access to any form of broadband connectivity. Addressing this unserved market, both Hughes Network Systems (HNS) and Viasat both reported double-digit growth in top-line revenues and EBITDA. During the crisis in the FSS GEO market, these companies are the shining beacons able to generate significant growth because they are serving a market niche that is uneconomical to be served via wireless or fiber.
Today, there are hundreds of thousands of homes in the USA being served by High-Throughput Satellite (HTS) systems and this is definitely one of the key applications which will drive the next growth phase for GEO systems. To point to the success of Viasat and HNS would be too simplistic because the market opportunities outside of the United States aren’t compatible with the same business model. Any company today purchasing a $300 million to $600 million HTS satellite to indiscriminately cover Africa and Asia with spot beams will find it extremely difficult to achieve success without carefully considering the local conditions in these emerging markets.
According to the International Telecommunications Union (ITU), there are about 3.2 billion people still unconnected to the Internet at the end of 2018. Most of these people live in rural areas of developing economies. There are three recurring patterns of rural communities in emerging markets: the income levels are very low; the population density is also low; and the availability of electricity is very low, non-existent, or very expensive.
Pacific Island nations and African countries have the highest electricity tariffs in the world. Aside from the lack of fiber backbone, the cost of sourcing electricity is the main cost driver of wireless network deployment. Most mobile operators in emerging market countries supply diesel to generators that power their wireless systems. Since diesel fuel is very expensive and logistically cumbersome, I don’t see mobile operators easily addressing the rural digital divide any time soon.
As an example, let’s consider an LTE base station equipped with 2 by 2 Multiple-Input and Multiple-Outputs (MIMO) and three sector antenna systems. This type of station requires about 4 Kilowatt Hour (kWh) to 5 kWh of electricity. Supplying diesel fuel to power to rural cell sites could cost well above $1,000 per month. The range of mobile systems are relatively short range, so unless there are thousands of subscribers in a given area, even if they could receive low-cost backbone via satellite, mobile operators will not deploy LTE stations to rural areas. In essence, unless we do something about it, the lack of power in emerging markets is what will slow the growth of adoption of satellites. Saving power in all forms will be the key driver of growth for our industry.
If the GEO FSS operators are able to provide a power-efficient, satellite-delivered solution to these 3 billion people living in rural and sparsely populated areas, then the global demand for HTS satellites of 100 Gigabits Per Second (Gbps) or more can be over 30,000 satellites (3 billion people x 1 Mbps/person divided by 100 Gbps/satellite equals 30,000 satellites: Assumes each user needs 1 Mbps dedicated). Obviously, most of this demand will be taken up by advancements in other wireless technologies but there will be plenty of room for satellites. Even if we are able to service just 10 percent of that demand, then it will mean that we need to deploy 3,000 GEO HTS satellites.
Since there are less than a handful of HTS satellite deployed currently, there will be plenty of growth to be had in our industry. The main catch is that we cannot use the current HTS satellites that are being used by HNS and ViaSat today to address this market because they are too costly and do not allow power savings on the ground. Yet if replicating the Viasat and HNS model will not work, what will?
First, the income levels of the rural poor will not justify an Average Revenue Per User (ARPU) of $50 to $100 per month. The rural poor cannot purchase $500 broadband Very Small Aperture Terminals (VSATs) requiring electricity to run because they don’t have access to electricity. This means that the monthly tariff has to come down one order of magnitude, the user terminal costs also have to come down one order of magnitude, and the user terminals need to run without commercial electrical power. How is this possible when the satellite terminals still cost $300 to purchase and by the time it’s imported and installed at a cost more than $500?
This seems to be a daunting task because HTS satellites offered by the large vendors cost between $300 million and $600 million to get the cost of the capacity below $1 million per Gbps. Therefore, a complete rethink of the GEO satellite-delivered broadband has to be done in order for satellites to fulfill its potential.
If the consumption of electrical power is what is limiting the wireless networks — and also current satellite networks — then we need to address the power issue. Currently, VSAT terminals consume 50 to 100 Watts of electricity. To reduce the power required by the satellite terminals, we need to improve the downlink power and the uplink Gain-to-Noise-Temperature (G/T) of the beams on the Earth. Achieving plus 60 Equivalent Isotropically Radiated Power (EIRP) in downlink power and plus 25 dB/Ks G/T will result in VSAT terminals needing only milliwatts of uplink power to close the link with a GEO satellite.
This will allow a very low-cost solar panel and battery system to provide reliable power to a VSAT terminal that would perhaps use only 10 to 20 Watt-hours of electrical power. In order to achieve such a high-powered EIRP and G/T, the spot beams on the earth have to be about 0.1 to 0.2 degrees in size. Such beam sizes can be achieved using large antennas in space and a beamforming system. In addition, since rural poor people in emerging markets cannot afford $500 VSAT systems, we have to find a way for the bandwidth to be shared by a community rather than used by a single user as it is today in the USA.
The VSAT system can be integrated with a Wi-Fi access point to deliver the service to a 50 to 100 meters with an Omni antenna or if bundled with a long-range Wi-Fi system like Curvalux (a broadband multi-beam phased array system) it can extend the range of a VSAT system to over 10 kilometers (Km). Curvalux requires less than 3 Watt-hours of energy to deliver 2 to 3Gbps to a 10-Km area. This will enable hundreds and thousands of rural communities to share the same VSAT installation. User devices such as Wi-Fi enabled and low-cost smartphones, tablets, and Wi-Fi Customer Premises Equipment (CPE) can all be charged by the sun using affordable solar cells.
By using these techniques, we can solve expensive terminal and power problems on the ground. Now, how do we solve expensive satellite and power problems in space?
Providing 0.1- to 0.2-degree beam sizes to solve the power issues on the ground means that to cover the entire earth from GEO, a massive satellite with tens of thousands of spot beams would be required. Is this currently feasible, practical and most importantly is it affordable? The answer is obviously no. So, why do it that way? Perhaps one of the solutions is that someone designs a satellite that addresses a smaller area (e.g. coverage of a smaller country) with a sufficient number of spot beams and investment size to make it economical (i.e. $1 million per Gigabit) to offer affordable broadband services to the emerging markets.
Why is $1 million per Gigabit a magical number? For a 15-year satellite, it means that each Megabit costs $5 per month and each Gigabyte delivered to the user costs less than $0.02 per Gigabyte. With mobile tariffs now running well higher than $1 per Gigabyte, certainly, there is room for GEO satellites to play an important role in serving the world’s 3 billion unconnected people.
Up to now, the only publicly available source of a $1 million-per-Gigabit satellite was ViaSat-3 — a ginormous $600 million-plus satellite with more than 1 Terabit of capacity that needed to cover all the Earth. Be that as it may, Viasat-3 does not provide 0.1- or 0.2-degree beams that will enable plus 25 G/T that allow 100 Milliwatt user terminals. Viasat CEO Mark Dankberg is a visionary, and not too many of us in the FSS industry has his courage to take on so much investment, nor do we have access to his level of funding. Thus, there is a need for HTS satellite capacity that can be acquired for $1 million-per-Gigabit that can be acquired in smaller incremental investment sizes for the rest of the industry to participate in the next evolution of our industry.
The key to being able to have an affordable HTS satellite in orbit that doesn’t require a half-billion dollar investment to achieve $1 million per Gigabit is a dynamic beam-forming and channelization system enabled by a digital processor or an Application-Specific Integrated Circuit (ASIC). An HTS satellite payload with a digital beam forming ASIC will use significantly less electrical power and will be much lighter than systems that require Field-Programmable Gate Arrays (FPGAs) or analog filters and waveguides.
For example, a 100 Gbps HTS satellite system, if built with analog or FPGA channelization techniques, will cost between $180 million and $200 million. That satellite would require 10+kW of power, weigh between 5 and 6 tons, and require a $60 million launch, with insurance costs of more than $300 million. However, in the near future, all digital satellites using beam-forming ASIC chips will be able to process a similar amount of capacity, requiring only 2 kW to 3 kW of power and weighing less than 1 ton with electric propulsion. Reducing power requirements will also reduce the size and mass of the spacecraft and thereby lower its costs dramatically.
A few companies are already working on all-digital beam-forming systems. Boeing’s digital processor for SES’s mPower (O3B 2.0) and the European Space Agency-funded Quantum platform purchased by Eutelsat are examples of such systems. There are also non-publicly announced private initiatives that are on-going for beamforming ASIC developments. In the near future, I predict these 100Gbp all-digital satellites would cost less than $100 million to be delivered into orbit (including launch and insurance). This may be a bold prediction, but I am certain that all-digital beam-forming HTS satellites will be made publicly available in 2019 for deliveries in 2021.
Combining low-cost satellite capacity and satellite terminals with long-range Wi-Fi is the key to unlocking the rural digital divide. This would result in an explosive growth of the FSS industry that would secure our future and, most importantly, bring knowledge, information communications and wealth to all the emerging markets of the world. The light at the end of the tunnel will be upon us sooner than we think