Monday, June 25, 2018

Evolving sharing

Now that it’s clear the Trump administration supports spectrum sharing (FierceWireless), this option will become even more salient in policy debates. I think the hardest question is how to allow a legacy incumbent to adjust their operation over time, potentially encroaching on the rights of new entrants.

Karl Nebbia, former NTIA Associate Administrator of the Office of Spectrum Management (OSM), framed it well at a recent NTIA spectrum policy symposium: How can services that remain continue to grow and change and evolve in bands that the wireless industry seeks to enter? Tom Power of the CTIA hedged his answer. The federal representative, Col. Fred Williams, was clear that the DoD and NOAA needed to flex their mission but are currently locked into a legacy configuration. Paige Atkins, Nebbia’s successor and the current OSM Associate Administrator, stated that reciprocally flexible access is one of the most significant challenges we face. (I’ve reproduced their exchange from the transcript at the end of this post.)

Such negotiations are difficult even where both entrant and incumbent are commercial players (e.g. cellular and satellite): When one industry generates more $/user or $/MHz than the other, it will always have an advantage in both market and regulatory transactions, like auctions and lobbying, respectively. (See also my post “Bringing a gun to a knife fight (spectrum edition)”.) 

When one of the services isn’t measured by money, e.g. if the incumbent is a government service, it’s even harder. There’s little or no upside for federal users to share spectrum, and a large downside: they will be blamed if there’s a mission failure. (This is just one of many examples why aversion to risk and innovation is rational behavior for bureaucrats.)

An approach that allows change as long it doesn’t create “harmful interference” requires on-going government participation, at least as a backstop, since the term’s definition requires case-by-case interpretation. (“Harmful Interference. Interference which endangers the functioning of a radionavigation service or of other safety services or seriously degrades, obstructs, or repeatedly interrupts a radiocommunication service operating in accordance with [the ITU] Radio Regulations.”) It’s possible to negotiate harmful interference boundaries in the market (e.g. between wireless operators at cell edges), but it’s rare – and typically between operators of similar services that interact repeatedly.

One can imagine getting the regulator out of the game if there are sufficiently well-defined interference protection rights that allow operators with very different technologies and missions to adjust boundaries, and if necessary make side payments. Since current rights definitions aren’t clear enough – and as a practical matter may never be clear enough – to allow such diverse parties to negotiate successfully (cf. the WCS/SDARS argument, which continues), this seems to lock the regulator into indefinitely and continually making adjustments to the rights of the sharing parties.

I don’t think overlays such as the PCS licenses auctioned in 1995 in the 1.9 GHz band will work when incumbents and entrants continue to coexist in a band. (There were thousands of point-to-point microwave uses in the band targeted for PCS. The FCC allowed the microwave users to stay for a while, but allowed PCS entrants to pay them to move.) The model in the 1695-1710 MHz sub-band of AWS-3 clearly won’t work, since NOAA is locked into operating at specific locations, and the calculations of coordination zone radii didn’t (couldn’t) take into account interferers that aren’t shielded by terrain and clutter, like drones.

Harm Claim Thresholds (HCTs aka Interference Limits, see e.g. the 2013 FCC TAC white paper, and my 2017 paper with Janne Riihijärvi  and Petri Mähönen) may be of some use since they are a measurable but reasonably technology neutral way to define interference protection rights. However, a limitation is that new services may require revisions in its HCT, e.g. protecting drones will requires HCTs to be defined in three dimensions, and an initial HCT may have been defined only at ground level.

Another possibility is to use interference risk profiles for each of the respective services. For example, a rights framework could that require no adjustment by one service can increase the risk beyond the pre-defined boundary without compensation. A risk profile charts the probability that a severity metric such as aggregate interference or percentage degradation in data throughput is exceeded. (For an example of risk-informed interference assessment, see my 2017 paper with Uri Livnat and Susan Tonkin that analyzes the 1695–1720 MHz AWS-3 band.) Just as with HCTs, problem will arise if a services extension requires the definition of a new severity metric which wasn’t included in the original contract.

TRANSCRIPT EXCERPT (my highlights)

Event page:
Webcast page:

KARL NEBBIA: ... But my question was in past reallocations, the approach even for those reallocations that have been willing to leave behind some systems to stay there long-term, whether it is satellite systems and 1755, 1780 or the two DoD locations in 1710 to 1755. Those reallocations have always been based on the idea that those staying behind were fixed in place. And the wireless industry was essentially granted maximum flexibility to use the rest of the space in terms of technology, various choices, aggregating areas, and so on. As we look to the future, is there a regulatory construct that the wireless industry can accept, that accepts changing needs of the satellite industry and the government in remaining in these bands? Is there a process or structure that we can actually foresee where those groups can continue to grow and change and evolve even staying in the bands that the wireless industry seeks to enter?

TOM POWER: I'm sure the answer is yes. (Laughter.) But it is almost 11 [o'clock, the scheduled end time]. I mean, that is ... that is a huge question. I mean, I think the one thing we keep saying up here is that there is a lot we still don't know. There is a lot -- the one thing we do know, is that constructs are changing. Yeah, something like that, yes. But I mean we haven't defined it. We are nowhere close to defining it yet.

PAIGE ATKINS: Fred, did you want to say something?

COL. FREDERICK WILLIAMS: I am not going to answer it, but I think it's important to have a couple examples. I mean, look at what we are learning in 1695 to 1710 [MHz] right now. We started that with exclusion zones. That turned into coordination zones because the service providers brought a technology in. So we accepted that. But to his point, we are locked in there because of lifecycles of satellites, but it doesn't -- the system doesn't allow for me to flex my mission. For example, I have had to go under BRAC or whatever. Maybe I do have a location change where I need to move those assets. That is like a nonstarter in the negotiations with our satellite providers or with the mobile broadband. The other thing is what if the Europeans launch a satellite. We may not get an assignment in this country, but they are launching L band satellites in Europe and we would like to catch that data. This is where DoD and NOAA are going to be in lock step to say hey, just you know, don't lock us down. We need to flex mission. At least allow for the negotiation to happen. And this is kind of green fields and frontiers, right? Hey, we shouldn't do an all rush to Oklahoma and do a land grab. If you do and you got there first and you got something important, great, but just don't lock everybody else out. That is not leading the world in 5G. I would argue if we could show the world a model where industry and the federal government share in a bidirectional manner and sell that abroad. That works.

PAIGE ATKINS: I would say that I think this is one of the most significant challenges we have, is how do we address, I'll say flexible access for new entrants and incumbents so they can evolve and grow over time and do it effectively with the certainty that both sides need to make it happen. So I do think that's one of the most significant challenges, that we need to come together and address.

No comments: