Jost: Software radios must be affordable

Technology suppliers are out of sync with the Defense Department’s priorities for developing and maintaining software-defined radios (SDRs), a top DOD official has said.

“If [an] SDR is not cost-effective over time, it will not be used — the department will likely turn away from it,” Ronald Jost, deputy assistant secretary of defense for C3, space and spectrum, told vendors at the Institute for Defense and Government Advancement’s Software Radio Summit Feb. 25. “Folks are always worried about the performance of radios, but it has to be cost-effective, it has to be upgradable at low cost.”

“If I took a survey of people in this room, you’d probably say SDR is there, so you can add new waveforms,” he said. “That's not what SDR was about [to DOD]. SDR was about the migration [of radios] to get a unified network so we could access enterprise services appropriately — a unified method for transport. If I want to provide that in a more rapid manner, SDR allows me to do that, but only if the radio is designed right.”

Not every radio needs to be based on the Software Communications Architecture, the basis of DOD’s Joint Tactical Radio System (JTRS) program, Jost said. “If I have a radio that was $50, why would I make it SDR/SCA?” he asked. “I'd just throw out the old one and make a new one. How about $500 or $1,000? At $10,000, there's probably a good case for an SCA radio.”

Jost pointed out several flaws in current SDR development that represent potential threats to the long-term viability of the technology. Multiple types of processors and components from multiple vendors have made it difficult to move waveforms — the communications profiles created for specific radio applications — from one radio to another. “We’ve used every processor and every [field-programmable gate array] possible,” he said. “We can't do this — you're killing us.”

Portability of waveforms was one of the key reasons for developing SCA. “We can't move waveforms from one radio to another fast enough — not in a month with two or three people but in a year with 100 or 200.”

Low-cost upgradability was essential because of the length of time it takes DOD to deploy systems, Jost said. “When you start a program, it takes four years to get it out of requirements, and to get it through acquisition, it takes seven to nine more years. So it takes at least 11 years to get [the program] out there.”

JTRS’ Ground Mobile Radio program has an even longer projected life cycle, he said. “How long will it take the Army to deploy JTRS? It'll be 30 years from when the program was started until the last radio will be installed.”

As a result, DOD's ability to upgrade its network is much slower than the cycle for changing the applications that run on it. Developers must be aware of the realities of the DOD network when they’re building applications, Jost said.

“You'd never see someone put an application on the shelf at Best Buy that required…a 10 gigabit connection to run,” he said. “They'd go bankrupt. But we do that in the DOD all the time.”