c-span2. >> here's a look at what's coming up. .. >> host: to spur economic growth. two members of that council of advisers on science and technology are joining us this week on "the communicators." on the left, mark gorenberg, and on the right, craig mundie of microsoft. they join us from the microsoft

studios in redmond, washington. mr. gorenberg, if you would, start by telling us what is pcast or the presidential council of science and technology? >> guest: it started back in the fdr days. this pcast has 21 members, it's chaired by john hultgren and vice chaired by bill press and maxine savitz. pcast works on number of reports, reports that were done for s.t.e.m. education, manufacturing, nanotechnology, health i.t., a number of other issues, and we took on this study in the fall as you talked about to look at the full potential of federal spectrum. >> host: craig mundie, if you

would, summarize this report that you all gave to the president. >> guest: well, in the report we make a series of recommendations that span quite a range of issues that begin with the question of what should the strategy be long term for the country in order to avail itself of the benefits that come from all the new technologies, the new electronic and software technologies that are available today. but for a whole variety of reasons, haven't been able to be broadly deployed. the report speaks to methods of creating incentives and measurement systems within the broad use of spectrum by the federal government in order to accelerate the movement from the classical model of spectrum allocation to the new model. and the report basically identifies specific strategies. so how this process could begin early even though the complete

transformation if it's to be done will probably taking take g as 20 years or more. >> host: also joining us iswall pair key, senior editor with -- paul kirby. >> host: thank you. the report calls for the immediate identification of a thousand megahertz of sharing and says such sharing should be the norm rather than reallocation for exclusive use. the wireless industry has reacted somewhat cooley to this, they say they prefer the option of using for exclusive use one of the reports said this is providing coffer for the department of defense. can you give us a sense how realistic it is that the report listen implemented as a result of this opposition? >> guest: well, that certainly is up to the white house to decide how they want to move forward on the report that we've

put together. but in terms of the situation itself, we're looking at dizzying demand. we're following on the broadband plan as you know, paul, that came out two years ago that called for the accessing 500 megahertz of spectrum for commercial use, making that available. um, the growth even since the broadband plan has been greater than expected, and if you look at some of the studies coming out there now, those studies look at an economy of perhaps $4.5 trillion by 2020, 50 billion devices, huge growth not only in spectrum as we use it today, but also in new be ideas like the internet of things and machine to machine. so if you look at that, you can say, well, that may call for 50 times more effective spectrum. if you look at current technology, if you look at just increasing spectrum, that might

give you two times. if you look at technologies being used like lt, that might give you five times. but to get 50 times, you have to go to a new idea for working with this spectrum. and what we found is that over the long haul the clearing and reallocation of federal spectrum is not sustainable as you saw in the recent excellent ntia report that just came out regarding the 1755 ban that called for over ten years $18 billion and a lot of disruption. what we looked at, and it's interesting that people consider us to be forward thinking, we think we're actually being very pragmatic and very conservative. and in the report we're looking at using technologies that are available today and deploying those technologies really around new policy to start to create these spectrum superhighways. we actually want to get spectrum

to these commercial users sooner. we think perhaps within three years by implementing this. in terms -- rather than ten years as talked about in the ntia study and, again, we're only looking at federal spectrum. >> guest: i'd jutte like to add one -- just like to add one thing which is, you know, there's a lot of member tear, as you mentioned, from ctia and some of the traditional cellular interests. but when we sat as the science advisers and looked at the country's aggregate need for spectrum over the long term, we conclude that's just going to be one application. there's going to be many, many applications both in the private sector and in the government environment and the ultimate sustainability question is not really whether you could just clear enough spectrum to make the traditional cell phone industry happy. we don't think that's even possible. but when you look at the broader demand over time, there's no possibility to do this without a

fairly radical transformation of how we think about allocating and managing spectrum. >> host: let's scroll down a little. randall stevenson, who's the ceo of at&t, said technical solutions are, quote, nowhere near prime time. now, the report talks a lot about using various technologies to try to enable this spectrum sharing. is there -- i'm curious if you have any more reaction to the criticism of the technology, if you will. >> guest: yes. and by the way, um, randall stephenson, who is a great ceo, we absolutely agree with the comments that he's made both in his op-ed that came out beginning of june, and then he did a speech for about an hour or so right after that. but if you look at what he said, he said, one, that it's going to be six to eight years if you cleared and repack spectrum for them whether it's commercial or federal spectrum and that sharing technologies, device

technologies that he talked about -- smart antennas, dsa, etc. -- would be five years or more. and they're looking for spectrum more immediately. and that was his call. we actually believe in what we put forward, that we're espog although we mention those technologies in the report, if you see, paul, in our recommendations, we don't talk about the deployment of those technologies at all. we talk about technologies that are here today already being used. we talk about database technologies and the idea of building a spectrum access system which is really an evolution of the white space be be -- white space technology that craig has been very involved with now for probably over ten years and could talk about. we talk about the deployment of small cell. and by the way, small cell technologies, as you know, we'll be shipping more units of small cell next year than macrocells. and we're talking about simple

technologies like dfs and the five gigahertz band and the sharing of military systems. so we're actually talking about using the technologies today. but we think that if we can create that backbone, that will be the spurring litmus to move these other technologies forward, and we'll see those in the five to ten years that will happen, more likely to see them, by the way, if we can get these superhighways started and get a framework going where they'll be successful. >> host: craig mundie -- go ahead. >> guest: i was just going to add a thought that many of these technologies don't have to be advanced substantially to even begin. i mean, today the whole world uses a lot of blue tooth and wi-fi which are very, very small cells. and, in fact, even the carriers today are offloading in high

density metro areas a lot of the data traffic from their cell phones onto these essentially wi-fi cells which are very small. but they create -- [inaudible] you could take the existing cellular system and just start to shrink the cell size dramatically, and you'd essentially get increased reuse even within the classical ark architectures. so as mark mentioned, you know, we're not betting the farm on the idea that there has to be an immediate introduction of radical new technologies. however, we do think over time that each of these additional technical changes sort of creates a multiplying effect in terms of spectral efficiency. and that's why we think that over time you can, in fact, accrue a much, much greater capacity utilization of the total spectrum than we're getting out of the current model of allocation. >> host: craig mundie is the chief research and strategy officer for the microsoft

corporation. mr. mundie, are there previous concerns when it comes to sharing spectrum space with the federal government? >> guest: no. basically, you know, sharing is a word that as your question connotes implies a lot of things with a lot of people. but in this case the fact that radios may coexist in a common set of frequencies in no way implies that there's access from one radio system to the next radio system. whether they were two private systems or a private system and a government system. those architectural questions are completely resolved, and, you know, i don't think there's any privacy issue there at all. >> host: in your report, gentlemen, you write that shared access to federal spectrum should be governed according to a three-tier hire around ri. federal primary systems would receive the highest protection from harmful interference, secondary licensees must use a

database and may receive some quality of service protections, possibly in exchange for fees, and general authorized access users would be allowed op o por tunistic access to unoccupied spectrum. could you explain that a little further, please? >> guest: oh, sure. so you look at the idea of, first of all, deploying federal spectrum, excuse me, and saying, okay, so you would have in different areas, you could have a federal user be the primary incumbent. you could then have folks that could get exclusive secondary access and that, by the way, could be current lte systems, it could be small cells that are looking for quality of service, almost like a carpool lane on a highway. or you can -- and then in general access you could have the idea of people who don't have exclusive access to that spectrum sharing that. in some ways similar to unlicensed but not unlicensed

spectrum, a way that it would be either registered or could be automatically registered so that there could be track of all of this in a database. >> guest: i'll just offer a thought. i mean, i think sometimes people get very animated about this question of federal preemption or priority, but there's two things that are important to realize. i mean, we have these situations, if you will, even in the wire line telephone network. you know, the government has for many, many years had the ability to preempt traffic on the classical phone lines or even the cellular networks in emergency situations. so there's, i think, always a need to reserve some kind of capability for government in specific situations. here we think that instead of that only being in an emergency environment, that when you look at the aggregate use of spectrum measured over time and over geography, there's lots and lots of places and times where people can have very sustained use of

this, even if there comes a moment where the government says, hey, i've got a plane or an exercise or a boat that happens to need to operate in that harbor, and you may have to move out of the way for some period of time. but when you look at the aggregate gains, we believe they would dramatically outweigh, you know, the risks associated with it or the disruptive effects of that kind of tiered model of priority. >> guest: the assumption is, the assumption is that the government is not using the spectrum that often, but when they do need to use it, it's very important. it's in small areas, it's infrequent times, it's in such a way where it's not very disruptive, as craig is saying, to commercial users, particularly ones who opt to get a specific secondary use of that spectrum. and, in fact, one of the arguments that the commercial world has used is that the federal agencies are not using the spectrum as often and,

therefore, not efficiently, and that's a reason why we should be, um, giving that spectrum for use to the commercial world. actually, if that's true, in this system it's even a bigger benefit to them this terms of how often they're going to get used to that spectrum. and over time we believe through the mechanisms we've argued for in terms of incentives, putting other pieces of funding in the white house, that those systems, those federal systems will become more and more efficient and, therefore, less obtrusive to the secondary user, commercial user and, in fact, make it easier even over time to clear some of those systems is and move them around to other locations. so we think that this is a way to kind of ease into the issue of getting much better shared shoes. and by the way -- views. and the by the way, over 50 president of the spectrum up to the 3.7 gig band is shared today as well, but it's shared in a very static way.

so this is not a radical departure. it's just a much more efficient use of how spectrum can be shared. >> host: you're watching c-span's "the communicators" program. our guests are two members of the president's council on advisers on science and technology. mark gorenberg is a venture capitalist from silicon valley, and craig mundie is the chief research officer and strategy officer with microsoft. paul kirby is our guest reporter. >> host: we're talking about priority access and kind of the hierarchy of sharing for the federal spectrum. a number of parties have made the case the federal agencies, for this to work, have to be assured that their spectrum, their operations won't experience interference and even in the report it mentions there could be, basically, a shutoff of the commercial operations if there's harmful interference. mr. mundie mentioned the tv white spaces databases. there's a couple of them right now, but it's not a nationwide

operation. can you give us, either of you, a sense of what type of assurance you think the federal agencies will need, and has the technology developed quick he enough such as the databases to give them the assurance that they will be able to to continue to operate in a sharing regime without interference? >> guest: well, in a sense, you know, immomenting all these things -- implementing all these things is a little bit of of technology and a lot of sort of operational insubstantiationuation of these capabilities. and you could say at the tail end a little bit of regulatory control to insure that the right deterrents are in place to keep people from doing the wrong thing. part of the reason in our proposal for this -- i even think of of it as sort of a transitional model of these stacked priorities -- is it allows us to get started immediately before the time where a government radio system might be modified to be more inherently able to operate in a

similar, generally-shared environment. and so a lot of the quest we had was to say how could you make use of technologies that are currently demonstrable and in the process of getting rolled out both in the united states and outside the united states and piggyback on the progress that's been made there in both the regulatory and a developmental sense over the last ten years and use that to accelerate the transition into this? each of these choices can sort of be made band by band. and so there's not, doesn't have to be sort of one giant database in the sky that controls everything all the time. it may be practical or even desirable to do that, but it's not a prerequisite to start to move in this direction. so, for example, we identified, you know, an area in the spectrum where the government has a lot of uses, for example, like maritime radar cans. and -- radars.

and while that's an important use when a ship is operating in a particular area, that's typically only going to be near the coast or maybe the great lakes. the entire rest of the country is currently reserved for those boats that'll never be to operate their radars. so our view is, you know, if the government's willing to know -- this is the idea of registration -- you know, what those devices are, then you have the ability to come back on anybody who violates it. these kind of violations happen today. i mean, you get people whether their amateur radio operators or others who east intentionally or accidentally, you know, start intruding somewhere they're not supposed to be, and the fcc has enforcement powers against those kind of issues. so to create this environment, yes, there'll be some technological activity, but you'll have to take these conventional mechanisms that are well established for enforcement on the fcc side, and then you'll have to have that contract made with the federal user. these have been done on an ad hoc basis already in specific

bands, for example, sharing radars with other applications. and so it's not like this is without precedent. what we're trying to do is say there's now a general architecture that would allow this to become the norm as opposed to things that are done on an ad hoc basis, you know, one after another. >> guest: and as craig is saying, we -- although we believe a system like this can be applied to federal spectrum, you know, regardless of frequency, we look more at the higher frequencies, and the higher frequencies, let's say 2700 and above, are particularly well suited to small cell. and everybody's sort of converging to this idea of, for example, in the, in the 3550 band which was of the first band that was analyzed by the ntia, they looked at it initially against large macro towers and created 200-mile exclusion zones around the coast which is not that interesting. but if you put in small cell technology into these area by

commercial users whether they have a priority access or whether they have a general access, you can shrink those exclusion zones significantly, and at the same time give much better solace to the federal government that they're going to have sort of a free and clear to operate. >> host: now, the report contends that the new federal spectrum regime could more than make up from revenues lost there not auctioning spectrum, but members of congress love auctions as you probably know because it brings in money. how can you then convince lawmakers? now, some democrats on the hill have been very favorable about the report, one republican commissioner and some republican aides on the hill less favorable. how can you convince them, or will you try to convince them it's okay if we don't get the auction revenues now because down the road it will create more than enough for the economy? >> guest: well, again, we're looking only at federal spectrum, and ironic -- if you look at that, auctions started in 1994, there have been $53

billion created by auctions, but other 90% has come from repurposing commercial spectrum for commercial reuse. a very small part has been from federal spectrum, and there's really only been one significant, successful auction of federal spectrum, and that was this 2006. which netted a little more than $5 billion. but, again, that is also being used, um, the omb will score that over ten years. realistically, it should be considered over many more years than that. so that's only a few hundred million dollars a year in total to date from federal auctions. and if you look at the ntia report that just came out, again, that talked about $18 billion, and they're not even planning right now, i believe, to auction that because the costs of clearing would be greater than what they suspect will come in. so when you look at federal spectrum, the amount of revenue they were going to get anyway would be relatively small. now, in our proposal we look --

once you've created a spectrum access system and you have the idea of secondary users, you have a system already set up where you could have a number of current types of revenue -- different can types of revenue models. you could have recurring revenue models from using that spectrum, you could have auction models and, in fact, we could be looking at a system where we perfected far more auctions to be used, but these may be smaller geographic auctions, may be secondary, smaller time frame auctions. so we think you could open this all up to a large number of auctions in secondary use, and that could bring in quite a bit of revenue over time in these federal bands which today it's unclear they will even bring in future revenue. >> guest: but i think it's also important in the spirit of answering your questions of what assurances could people have. i guess the question i would pose back to them is have they

ever used blue tooth or wi-fi? these are on a completely unlicensed, totally shared environment with no database that administers them, and you can go all around the planet today, and there's a huge economy built around the use of wi-fi, the sale of the equipment, the integration into phones, the offloading of data traffic, you know, all new equipment that people are buying in their homes, new cars. every one of these things is essentially aiding the economy, if you will. and that's why the title of the report talks about the impact of making a strategy choice like this that does move you away. in fact, our belief is over time implementing this really moves us from an artificial shortage that we have today in the apparent availability of spectrum to one where spectrum appears to be abun adaptly -- abundantly available, that people who want to do an application in any domain should be able to with no capital

expenditure get immediate access to spectrum to do their problem. and in that environment, you know, we think it's going to, basically, boot the economy up further. and so the indirect rave -- revenue that comes from economic growth we think will dwarf that. one of the things in the report, there's a citation to a recent economic study done in europe about making a fairly specific allocation of spectrum into a new, sort of larger unlicensed band, and they estimated that these few tens of megahertz put into that model in europe would be the rough equivalent of an 800 million euro stimulus program. >> guest: actually, billion. >> guest: i'm sorry, billion euro stimulus program. whether you look at it as an economist or just practically and what has happened with the stimulus created by the arrival of wi-fi and blue tooth in, basically, less than a decade, you should have some considerate that these things -- have some

comfort that these things could happen. >> host: when the report was released, officials were very complementty of it. have you gotten the sense that they plan for the president to issue the memo you call for anytime soon, and if so, when? >> guest: well, i'll just say, i mean, as mark said earlier, it's ultimately the president's choice about, you know, what he does in terms of the report. but, you know, i think it's safe to assume that, you know, we wouldn't issue the report in a complete vacuum. there's been a lot of discussion with all the stakeholders, both in the white house, throughout the federal agencies and, frankly, even many of the commercial companies. even those that are sort of in some ways complaining about different aspects of the report. we're all party to the discussion, you know, before we even completed writing it. so, you know, it's not like we went into a closet somewhere, produced this thing ask just dropped it on the president's desk. there's a lot of thoughtful process that goes into the development of it, and as such,

you know, we are pretty hopeful that the president will, in fact, you know, move to act on the recommendations. exactly how, you know, that's up to him. >> host: mark gorenberg, why a thousand megahertz? where did that number come from? >> guest: well, first of all, we look at the demand, and we see the demand is going to be huge. the other is why a thousand megahertz, the ntia looked at the idea of when the broadband plan came out, they started to identify spectrum, and initially they looked at up to 2.2 gigahertz of spectrum and then keyed in on 1.5 gigahertz or 1500 megahertz that should be analyzed. from that we started to look through that and said there is a good opportunity how it will evolve in this area of, say, 2700-3700, and there are a lot of other bands as well. so 1,000 megahertz is both a combination for the need of demand for sharing with the federal government, but also the idea of setting a big goal that

sets a marker out for us to go after. >> guest: i'll add, though, that there is sort of a technical basis for why we want the big goal. our model in the long term is that we shouldn't be giving people tiny little slivers of spectrum to operate in. that, in fact, the new architectures that are emerge anything software-defined radios and other related technologies really operate better if they get to range across a much broader chunk of frequencies. and that's, in fact, where we think that you'll get much higher overall utilization. the military, of course, moved to some of these kind of architectures years ago, spread spectrum communication, for example, not for spectral efficiency, but to basically make it harder for people to jam them. but, in fact, the inverse of that will be a benefit. if you can, basically, get a lot of people operating across a much broader band, you're likely to get much better spectral

efficiency over the long them. to get everybody into that model will, in fact, require that all radio systems are rearchitected from the classal 900-year-old model -- 100-year-old model, and that's why we claim the entire process will take, you know, two to three decades. we can't go up and get the satellites and service them and change their radios. some of these systems just have a long lifetime, and you won't be able to get them all rearchitected. but i believe over that period of time, and with a loot of it happening much sooner, that these things will actually happen. >> host: final question from paul kirby. >> host: as you say, the report says that implementing all this could take decades, but, mr. gorenberg, you mentioned some things could occur in a few years. realistically, three years from today if the president were to issue the memo, what do you think could be realistic three years from today? fi guest: yes, paul, thank you.