Today I filed a reply comment on the FCC’s CBRS NPRM. Here’s the summary.

The most important frequency band class for the 3GPP 5G standard is n77, an internationally harmonized band extending from 3.3 GHz to 4.2 GHz. CBRS is an FCC experiment taking place on 150 MHz ranging from 3.5 to 3.65 MHz. For purposes of experimentation with a relatively novel spectrum sharing discipline unique to the United States, CBRS could use a wide variety of frequency bands outside of n77. The status quo interferes with and complicates the deployment of 5G in the US.

If the FCC had set out to deliberately roadblock 5G, it could not have made a more a more harmful choice than 3.5 GHz for the CBRS experiment. To be clear, the FCC did not go down this road with malicious intent; CBRS is a victim of history and emerging standards outside the FCC’s control. At the time that the FCC initially addressed 5G, the millimeter wave (mmWave) band class appeared to be a comfortable home for 5G that was not encumbered by high value incumbents. Unfortunately, the mmWave band didn’t live up to expectations in terms of propagation.

While CBRS is still a domestic science project, 5G has emerged as a vital global advance in both mobile and fixed wireless access (FWA) broadband. 5G is so powerful that it represents a competitive entry in the US residential broadband markets that were minimally competitive prior to its inception. The success of 5G FWA has placed a target on 5G as a whole. This turn of events is undesirable.

The FCC has at least two options for continuing the CBRS experiment without hampering the 5G rollout: 1) As AT&T and other filers suggest, CBRS can be relocated to the lower portion of the 3 GHz band; or 2) CBRS can be enhanced to include a new 5G-friendly license class permitting something resembling normal 5G operation in the existing CBRS band in selected areas. The second option is consistent with the over-arching goal of the CBRS experiment, namely examining the value of control systems (the Spectrum Access System) to permit the use of different kinds of licensing disciplines. The essence of CBRS is the SAS, not any particular band class, power limit, or coverage area.

As it presently exists, CBRS has two very different and contradictory goals. On the one hand, it is an experiment in the management of dynamic spectrum licenses and on the other it is a means of protecting federal users, mostly military, who rely on relatively crude systems such as radar from interference with more modern data-oriented systems. The nature of the federal systems requiring protection hampers the goal of achieving dynamic management. CBRS places 100% of the onus of cooperation on the private sector systems and expects absolutely no change in behavior on the federal side.

This is the wrong way to approach the problem. Federal users need to take a more active role in signaling their spectrum needs to private sector users. Instead of relying on the severely dysfunctional Environmental Sensing Capability, federal users should explicitly make their needs known through Incumbent Informing Capability (IIC) or direct notification through the SAS. A successful system of dynamic spectrum rights assignment begins with basic agreement of all parties to cooperate. The federal sector needs to step up and shoulder at least some responsibility to cooperate.