PARIS — Satellite ground system provider Kratos and fleet operators Intelsat and Hispasat plotted a route to fuller satellite integration into the terrestrial IP network. Satellites will lose some of their uniqueness but offer lower costs and higher flexibility to users.

Everyone wins, with the possible exception of some satellite manufacturers.

“Satellites in the past have been special, and that will go away,” said Stuart Daughtridge, vice president for advanced technology at Kratos Defense & Security Solutions. “They will be just another node on a network.”

The key to the transition is frequency reuse and the coming software-defined satellites, whose first models are now being ordered. These GEO-orbit satellites are generally smaller than their predecessors, with less capacity. But they can be modified in orbit, making them more fungible.

They are also capable of modifying coverage and frequency, giving their owners a hedge against market changes, important for a 15-year-long asset, and the ability to adapt to changes in customer demand and location — and even to the weather — with only a few minutes’ notice.

“Software-defined satellites are more analogous to a network switch in the sky,” Daughtridge said during a Sept. 2 webcast organized by Kratos and the Paris-based Euroconsult consultancy. “It’s a network device, and we have to start treating it that way. How do you operate that network switch with the ground infrastructure? That’s the mental shift you have to make.”

Bruno Fromont, Intelsat’s chief technology officer, said software-defined satellites can be built in a standardized way that should reduce production time and lower overall capex. He said mobility markets will benefit the most, but even media-broadcast-focused businesses will benefit from the lower satellite cost.

Intelsat sees software-defined satellites as so important to its future, post-Chapter 11 business that it asked its bankruptcy court for special permission to purchase two Airbus Defence and Space OneSat software-defined spacecraft to serve Intelsat’s acquisition of Gogo Inc.’s commercial aircraft in-flight-connectivity business. (Read: Intelsat investing $2B in software-defined network; 2 satellites ordered, RFP out for another https://bit.ly/3oh2ObP on SpaceIntelReport.com)

Fromont said the two satellites are part of a $2 billion investment by Intelsat into a space and ground segment redesigned into software-defined networks. The company recently issued a request for proposals to satellite manufacturers for another software-defined satellite.

“This will translate into new services for end users, from long-term commitments to more pay-as-you-go models, which can be efficient to run. Before, it was difficult to do this.”

The change will force satellite service providers, whose long-term bandwidth commitments have been costly during the Covid pandemic, to adapt. They now are changing the line of separation between what they do and what satellite operators do.

“In Covid we saw how fragile the service providers have been,” Fromont said. “Look at all the restructuring that happened. We acquired Gogo for a reason: They were fragile. Others may be impacted by redrawing the line between what is a service and what is providing network capacity. That is going to have a lasting impact and I think it’s healthy.”

Hispasat Chief Technology Officer Antonio Abad said the long debate in the satellite industry over how to accommodate itself to the dominant terrestrial networks is now being answered with software-defined satellites.

“This will accelerate integration into this IP world,” Abad said. “We are a very small portion of the terrestrial infrastructure. The role of the satellite is to provide connectivity in a seamless manner with the other 99% of connectivity providers, who are terrestrial.”

One temporary headwind: the current satellite ground networks. They will need to be redesigned to match the agility of the software-defined satellites and this will be costly.

“In the past, the ground segment was a small percentage of the total capex,” Abad said. “Now, in connectivity programs, it’s the third expense after the satellite and the launcher.”

Daughtridge said a software-defined satellite’s payload can be updated every five or 10 minutes to move with changing service-delivery circumstances, and the ground network needs to move that quickly too.

“Having a satellite that can update every five minutes is great, and having a ground system that can deploy services on the fly is great, but if you’re planning process takes three hours, it limits the value.

“If your planning cycle is every 10 minutes, you can optimize on weather and stop optimizing around interference and things like that,” Daughtridge said. “It’s pretty easy to know what the weather is going to be every 10 minutes. You can apply power and bandwidth where needed.”

Fromont said moving to software-defined satellites will not necessarily increase or decrease demand for GEO-orbit satellites. He said he could imagine a time when satellite fleet operators share these satellites’ resources and focus their competition on differentiated services.

Fromont acknowledge that there will be some costs to satellites becoming standardized equipment. But the capex saving and flexibility and future-proofing is worth it.

Fromont and Abad agreed that this type of GEO-orbit satellite is every bit as efficient, when measured in cost per bits, as any of the low-orbit broadband constellations entering the market now.

The panelists did not address whether the future satellite industry, as they described it, would not need far fewer satellite manufacturers than are in the market today.

Abad said satellite builders could store standard-design software-defined satellites in inventory, allowing an operator to make an order and see it in orbit in months, not years.

“At the end of the day they are going to keep building the same satellite,” Abad said. He said the challenge of making different manufacturers’ spacecraft fully compatible in a network could be handled by third-party ground segment providers that will control the payloads.

You’re going to need to abstract the satellite uniqueness from the different manufacturers away to dive operational efficiency, just as has happened in C2 and other systems in the past,” Daughtridge said.

Wouldn’t it be simpler and less costly for satellite operators to settle on two or three manufacturers?

Read more from Space Intel Report.