Eight years from rulemaking to license: Peter Rysavy on how CBRS works
I asked Peter Rysavy a pointed question: why is spectrum sharing a poor choice for 5G? He took it apart. It depends what kind of sharing you mean.
The kind the Department of Defense has been contemplating does not exist yet. The idea is that military systems use a band for radar and other operations, and when they are not using it, commercial 5G networks pick up the spare capacity. There is no way to do that efficiently with today's commercial 5G networks. But there is a working form of sharing already, in the band just above the one the Pentagon is looking at, and it is called the Citizens Broadband Radio Service, CBRS.
CBRS was originally contemplated in a notice of proposed rulemaking by the Federal Communications Commission in 2012. And it's only now in 2020 that the license portion of that band is being made available. The whole system is quite complex and really was a very ambitious spectrum experiment.
Then he explained how it works, which I had never heard laid out so plainly. CBRS covers 3.55 to 3.7 gigahertz, 150 megahertz in all, and up to 70 megahertz of it can be licensed in each area. At its center is a database called the Spectrum Access System. Every base station that uses the band has to ask the database whether it may transmit at that moment. The Navy is always the primary user, but it does not tell the database when it is using the band. Instead, a network of sensors along the coasts listens for Navy radar. When the sensors hear it, they tell the database, and the database takes those channels away from commercial users.
Below the Navy come two tiers. Licensed users have priority, and general users can transmit, in something like an unlicensed mode, whenever neither the government nor a licensee needs the spectrum, as long as the database says yes. The licenses themselves cover areas as small as a county, much smaller than traditional cellular licenses. That puts spectrum within reach of organizations that could never buy it before. People imagine private networks at shipyards or aircraft plants. Cable companies like Comcast and Cox won licenses. Verizon bought a great deal of it to add capacity to its 5G network. How it all shakes out, he said, is too early to tell.
The timeline is the part policymakers should keep in mind. A sharing system far simpler than what the Pentagon wants took eight years to go from proposal to licenses. He thinks any more advanced scheme needs the same patience: perhaps three to five years of design, then specifications written into 6G for deployment in the 2030s. Trying to do it now, on the theory that it helps the country compete with other nations, does not make sense to him.
I suggested we let CBRS play out before anyone calls it a proven model. He agreed, and expects that anything the Defense Department builds would be far more ambitious than CBRS, which makes it another grand experiment that will take a long time. I told him I would put away the CB radio I had been planning to repurpose.
What I admire about CBRS is that it is honest engineering: a database, a sensor network and a priority scheme, each simple, combined into something that works for one band. The lesson is not that sharing fails. It is that sharing done properly takes years, and the Pentagon's proposal asks for something harder, faster.
From Lincoln Shorts, the interview series I hosted for Lincoln Network, recorded 18 October 2020.
