Flexible Management of Spectrum Rights and the SAS Design Challenge
2 Pages Posted: 31 Mar 2015 Last revised: 2 Jul 2017
Date Written: March 30, 2015
The recognition that spectrum management practices have not kept pace with wireless technology or markets is hardly new. Efforts have been underway for well over a decade to effect a paradigm shift from old-style command & control regulation to more market-based regulation. Much of the urgency for effecting spectrum reform in recent years has been motivated by the need to satisfy commercial demand for mobile broadband spectrum. This has driven efforts to expand access for both exclusively licensed and unlicensed spectrum, as well as the introduction of new shared access models (e.g., at 3.5GHz). A significant thread in the policy debates has been whether exclusive licensed or unlicensed is the preferred model, and much of the academic debate has been framed in the language of property rights, with exclusive licensed spectrum mapped to a private property regime and unlicensed to an open access regime. Neither of these pure rights regimes provide a good match to the reality of spectrum management, and neither adequately captures the many intermediate cases where multiple classes of users with differing rights share the spectrum. Furthermore, as wireless technologies and markets continue to evolve, there will be a persistent need to continuously redefine and reassign spectrum rights. In addition to expanding options for shared spectrum access, we need to enhance our ability to flexibly evolve spectrum management to accommodate new rights assignments and the enforcement regimes those assignments imply.
The Spectrum Access System (SAS) which was proposed by  and is being developed to support the new tripartite sharing model proposed for 3.5GHz is a key bit of technical, regulatory and market infrastructure that has the potential to play a central role in a new model of more flexible dynamic spectrum management. To understand and realize this vision, it proves useful to treat spectrum as a Common Pool Resource (CPR) ,  and interpret the SAS as a mechanism for the sort of polycentric governance that CPR scholars suggest can be effective , .
In this paper, we explain why viewing spectrum as a CPR offers a useful framework for thinking about the choice of management regimes and the attendant enforcement challenges that alternative rights assignments imply. Following Jessop, Ostrom, Agrawal, Williamson – and others, we adopt the notion that governance refers to the processes by which rights are distributed among stakeholders, the means for ensuring that those rights are enforced , and the resource is managed for sustainability. We have argued previously () that flexibility is important as we consider enforcement in spectrum sharing and we extend this notion more broadly here. This perspective elevates the discussion above the overly-simplistic choice between exclusive licensed and unlicensed and forces consideration of the more layered and nuanced reality of spectrum rights and the challenges of dynamic spectrum management.
A key point is that current conceptualizations of the SAS have been too limited in their scope, focusing too much on the technical components of the system. This is akin to the mistake of early AI expert system designers who neglected to consider the user as part of the expert system, or alternatively, akin to the failure to adequately consider security issues in the original Internet architecture. Viewing the SAS as a polycentric governance mechanism rather than just a technical means to provide spectrum access forces holistic consideration of the stakeholders, market and regulatory institutions that together with the technical components of the SAS comprise the larger system. Polycentric governance provides a robust way to customize sharing solutions along the continuum of possible rights regimes, allowing spectrum management to evolve gracefully as circumstances require , .
We explain how governance may be incorporated in the design of the SAS to support Federal-Commercial sharing in the 1755MHz and 3.5GHz bands, as well for more futuristic visions of virtualized spectrum access , heeding the lessons of our previous research (e.g., ), which has shown that the particulars of spectrum sharing are specific to each spectrum sharing case.
This work was supported in part by the U.S. National Science Foundation under Grants 1247546, 1443796, and 1345256.
Suggested Citation: Suggested Citation