The FCC recently released proposed rules for the 3.5 GHz proceeding, a proceeding that has received a lot of attention as it is the first attempt to apply the President’s Council of Advisors on Science and Technology (PCAST) report that promotes the sharing of federal spectrum with non-federal users. The Commission has identified this spectrum as an “innovation band” and we applaud them for their “outside-the-box” thinking on increasing spectrum efficiency. However, it is crucial that the Commission find the right balance between implementation of this new sharing model and incentives to invest in the technology necessary to make it work.
The FCC’s proposal is a complex model where federal incumbents would share their radio spectrum (which they use primarily for ship-borne radar) with two non-federal user categories; the first being a licensed category called “Priority Access Licensees” (“PAL”) that will obtain licenses to spectrum in geographic areas and the second called “General Authorized Access” (“GAA”) which is similar to unlicensed operations. The three groups will operate in a prioritized hierarchy where the incumbent federal users have the highest priority, the PAL the next highest and the GAA users after that. In order to have all three groups coexist without interference, a database, or what the FCC calls a “Spectrum Access System” or “SAS” will need to be developed and implemented. The SAS will include a registry of all nodes and/or devices in the band, and will know their geo-operational status at all times, and can therefore theoretically enforce the band’s use and prevent unintentional interference.
As this 3.5 GHz spectrum is spectrally higher than what mobile broadband optimally uses, the FCC has proposed using this band for “small cell” deployments. Small cells are low-powered, low-elevation micro cells that are typically installed in office complexes, campuses and residential areas to augment the large macro cell sites that are used for main coverage areas. Small cells are typically used to increase network capacity in areas such as stadiums or metro stations where demand for network services is concentrated in a small area. Today, this is accomplished in many areas by using WiFi networks where available. But a licensed spectrum band such as 3.5 GHz could offer an alternative and potentially even better user experience than WiFi as small cells would be connected directly to a mobile operator’s macro network.
However, for AT&T to deploy small cells in the 3.5 GHz band, new infrastructure will need to be developed and deployed and new devices will need to be developed and put in consumers hands. This will require a significant investment and some degree of certainty is going to be required to justify that investment. Under the current proposed rules, this three-tier framework requires a yet-to-be-developed SAS, and since there are three user groups that will want access to the spectrum, the SAS-managed approach has a high level of deployment risk – especially if attempted all at once. As history has recently shown with other complex systems, AT&T is concerned that if this PCAST model is rushed to market, it could have disastrous consequences and could actually set progress back.
As other companies have proposed, AT&T recommends a transitional, phased-in, interim approach to deployment of PAL and GAA operations. First, we recommend dividing the band into PAL-only, shared use and GAA-only sub-bands. This will allow PAL and GAA service providers to develop their products for initial deployment in a familiar environment, such as in our case, licensed geographic areas with a five-year license term and a renewal expectancy coupled to build-out requirements.
Having separate PAL and GAA sub-bands would provide immediate access for deployment without any fear of interference and 100% of the attention can be focused on preventing interference to/from the federal incumbent users. Second, we recommend using the shared-use sub-band as the development space for the SAS. Here, SAS providers, PAL and GAA users can collaborate, field test and ultimately deploy the three tier model without any concerns of interfering with the PAL or GAA-only sub-bands. Finally, when the SAS systems have been developed and fully tested and the transition period ends, the three sub-bands can be eliminated, and allow all three user groups to share the entire band as originally intended by the PCAST report.
We believe that the 3.5 GHz proceeding holds promise as a basis for future applications such as small cells. It would be a shame for it to be delayed or worse, unrealized as a result of a rushed and potentially troubled deployment, especially when alternative proposals to reduce risk and increase certainty readily exist. Spectrum sharing can work, but a “walk before you run” approach is warranted and appropriate.