How Neutral is Neutral?
I keep talking about technology neutral licensing of spectrum but is it actually possible? Is it possible to devise a set of rules that allow an entity to use spectrum, as that entity sees fit, without causing pain and harm to all those around and WITHOUT reference to the details of the technology that will actually be deployed?
As aside the word neutral is used far too recklessly in my opinion. Ireland is a neutral country (oh yeah?). And in my opinion Net Neutrality is just one kind of business model that does favour certain entities rather than being ‘neutral’ to all involved in the network. By saying this I am not saying I am against it, just saying it is not as neutral as it sounds.
Back to spectrum … and technology neutral licensing …I took the diagram below from a document on spectrum licensing that I found. The document is written by William Webb. I always seems to be quoting William Webb. The document is all about how we license spectrum in a more neutral manner.
This diagram shows the options as he see them starting with transmitter power limits (defined by things such as Block Edge Mask) and moving to more complex interference based restrictions (for which he suggests that Ofcom Spectrum Usage Rights make sense). Webb views the approaches as an evolution of each other and I think this is a very good way to look at the options.
Let’s look at each end of the spectrum of options here. Let’s look at the Block Edge Mask (BEM). The BEM on the face of it seems like something that could be technology neutral.
It merely details the in-band and out-of-band power levels that should be used by a transmitter in a given band. The problem arises when it comes to setting the values of these levels. To do this, you have to begin to imagine what kind of technology might exist in a band and in neighbouring bands to put the numbers in. So in essence you put the technology back in the picture.
Reza Karimi and co’s paper on calculating a BEM that might be useful for neighbouring cellular and DTT services is an example of what I mean. Though the specific technical details of each network are not needed the topography and deployment configurations (e.g. antenna heights and details) are. Actually I suppose you could say a certain amount of the technical details of the network are implied. So the cellular network considered in the example in the paper is a multi-frequency cellular network (which could be a GSM or LTE network) rather than a single frequency cellular network.
The Spectrum Usage Rights (SUR) approach is considered to be more technology neutral. SURs involve setting parameters for power flux densities at the receiver that control the amount of emissions that can leak into neighbouring geographical areas and that control emissions inside and outside of the licensee’s bands. To set these limits the regulators need to consider likely usage scenarios for the bands and deployment details as well. Hence again while the specifics of the technology do not need to be considered some broad considerations relating to the types of networks that will operate in the bands are needed (that indirectly relate to the technologies in use).
So does all this matter? I feel there is a way that licenses can be defined that do not require us to envisage usages and think about deployment scenarios (and thus indirectly link with a broad class of technology). But would this be needed? To a certain extent maybe we can always see ahead enough to know the usages we need for the ‘next while’. And the SUR approach for example does include mechanisms for changing the SUR values- maybe these mechanisms are as dynamic as they need to be?But maybe I am falling into the trap and thinking too statically too.