[CWN-Summit] RE: [Openspectrum] Spectrum updates

[Harold Feld]

At 04:04 PM 6/29/2005, Patrick Leary wrote:
>Harold,
>
>I wish to correct your comments regarding the "WiMAX Posse" and 3650MHz. I
>am qualified to do so as I was principally involved in the WiMAX Forum
>response (I wrote the first draft), the Intel/Alvarion/Redline petition
>(obviously), and the WCA petition (splitting the band was an idea I tossed
>out).
>
>Where you are incorrect is in your assumption that our efforts are entirely
>WiMAX centric. While this is partially true, what we (at least Alvarion) is
>most interested in a set of rules that 1) is technology agnostic, 2)
>harmonized with other similar allocations globally, 3) enable true QoS.

Patrick:

I understand that the rules are technically "technology agnostic," but 
everyone I have spoken
to has interpreted the filings as directed at promotion of WiMAx.  This is 
based not merely on assertions
within the comments that making the band WiMax compatible would provide 
positive public interest benefits,
but on the nature of the parties filing.

I will further add that the efforts of parties to recapture 3650-3700 for 
licensed use is having a very negative impact
on the efforts to get unlicensed access elsewhere, such as the broadcast 
bands.  If Alvarion, WCAI, Intel, etc. are
concerned about the long-term future of non-exclusive shared spectrum 
access, particularly in the television bands,
I strongly urge you to rethink your position and start making some 
political inquiries.  Especially in the current FCC,
which is at best neutral toward unlicensed, this sort of active internicine 
warfare only helps the NAB build its case.

>The "contention-based protocol" requirement is counter productive to the
>intent of the rule and also imposed a narrow framework for industry to deal
>with interference issues.

Actually, the intent of the rule was to deal with the "Canopy 
Problem."  And it wasn't just us.  I had a meeting with then-Commissioner 
Martin in which he explained that lots of people wanted unlicensed for 
routine stuff like meter reading, but were afraid that if they weren't 
"always on" they would never be able to get back on because other people 
insisted on being "always on."

Contention based protocols deals with this problem in a neutral way that 
does not require any FCC intervention.

>In fact, the rule simply cannot work, as I will
>explain later. Your post also implies that our community is looking to
>prevent or otherwise be hostile to Wi-Fi. Nothing is further from the truth
>and we prove it in the language of our reconsiderations. In fact, our plans
>actually make the environment more conducive to immediate deployment of
>802.11-based systems, along with WiMAX and any others, including Motorola's
>(very spectrally unfriendly) Canopy.

I do not mean to imply that you are "hostile" to wifi.  Many of you 
manufacture wifi equipment, after all.  No, the problem is that the vision 
you describe for the band is one that is neither sensible for the band, 
given the exclusion zones for incumbents, nor compatible with the vision of 
my community wireless/public interest clients, which is a band that allows 
easy entry at the lowest possible price.  The vision you propose for the 
band is essentially a carrier-style vision in which 3650-3700 MHz becomes 
an auxiliary band for licensed WiMax or other licensed technologies.  The 
rules proposed on reconsideration do not (in the opinion of those more 
qualified to judge than I) further the ability of multiple small operators 
to share spectrum cheaply and efficiently.

>We (Alvarion) met at length with the staffers at the Wireless Bureau and the
>OET who wrote the rule and we impressed upon them that a better approach
>would be to impose a technology agnostic rule.

The current rule is "technology agnostic."  Any contention-based protocols 
capable of sharing the band may be used.
Also, could you please provide a link to the ex parte of the meeting?  I 
confess my workload of late has kept me from following the docket as 
closely as I would like.

>  For example, 802.16 has
>formed Task Group H that is working on unlicensed co-existence. The TG is
>trying to establish a joint task group for this with the 802.11 crowd,
>because the goal of the TG is to come up with a TECHNOLOGY NEUTRAL scheme
>that could be software-applied to any 802.11, 802.16, 802.20, etc. device
>and would allow all the devices to recognize each other and cooperate via
>time slicing. But the 802.11 crowd is rejecting any involvement. In other
>words, it is the 802.11 crowd that is openly hostile (and terrified) of the
>"WiMAX Posse," not the other way around. The 802.11 crowd does not want
>neutral rules, they want protections.

Well, I cannot speak directly to IEEE politics from any personal 
experience.  But from what I have heard second hand, the 802.11 crowd are 
saying "we could slap together something that would satisfy the FCC today, 
why should we embark on a process that will delay deployment?"

I understand this will be the subject of considerable debate at the July 
meeting in SF.  I look forward to hearing about the outcome.


>[NOTE: Interestingly, and contrary to the belief of the Wi-Fi folks, the FCC
>folks told us that in no way were they attempting to protect 802.11, nor
>were they attempting to be hostile to 802.16. But it was clear to us that
>they had not realized how problematic was their ruling.]

The division of this into 802.11 v. 802.16 camps is only relevant to the 
basic question of how multiple users can co-exist within the band.  In an 
802.16 universe, it is profoundly difficult to envision multiple users, 
particularly in mesh networks.  But even excluding mesh, without 
contention-based protocols, the incentive is to blast your transmitter all 
the time.  The issue is not 802.11 v. 802.16. Any technology, regardless of 
whether it is an IEEE approved standard, that allows multiple users to 
share the space by resolving the "I want to speak but someone else is 
speaking" problem is permitted.

WiMax is a perfectly fine technology for what it is intended-- a national, 
licensed band.  Why must every available frequency be made "WiMax friendly" 
anymore than every available frequency should be "broadcast friendly" or 
"PCS friendly."  The FCC set aside 50 MHz of spectrum, with geographic 
carve outs from the top 25 markets, to try an experiment in frequency 
coordination.  It did this based on comments from hundreds of smaller 
entities that they wanted something like this for the purpose of bringing 
broadband to rural areas and low income areas served by low-power mesh 
networks or higher-power hub-n-spoke in rural areas.  They did it in 
response to industrial users and others saying they wanted a place where 
they couldn't be squeezed out by "always on" transmitters.

>In terms of harmonization, must I really have to explain this? U.S.
>allocations are already all over the map and incongruent with global
>allocations. Do we really want another band that forces industry to come up
>with a unique technology patch only applicable to the U.S.

So leave it to the niche players and the open source folks.  The exclusion 
of the most profitable DMAs already makes it a hard sell for this band in 
the U.S.

>In any event, the rule is unworkable anyway. Why? It forces "industry" to
>come up with a scheme to avoid interference and enable others to operate,
>but it does not require the various groups to agree and to implement a
>single approach. So tell me in practical terms how such a thing is remotely
>workable? Should I spend a million bucks to design some advanced protocol
>that allows sharing across technologies, only to have it rejected by the
>Wi-Fi crowd, etc. who has no obligation to agree to the technique, no matter
>how neutral?

THIS is the critical question, and the one upon which the future of 
non-exclusive use of spectrum hinges.

>Finally, what of QoS? If wireline equivalent QoS cannot be enabled, then the
>band can never create a service competitive against cable or DSL, and
>certainly useless for mobile. These type of half-assed rules force wireless
>broadband to be the low-rent broadband backwater, never more capable of
>providing anything beyond best-effort.

Again, you re-iterate the argument for licensing.  If what you say is true, 
that market-based coordination is impossible and that QoS is impossible in 
an market-based environment, then unlicensed spectrum (or any other scheme 
for non-exclusivity) is doomed.  Period.
As an aside, there are many situations in which "best efforts" is a 
perfectly acceptable QoS, and anyone who wants better needs to find another 
provider.  The QoS argument goes back to the 1980s and 1990s, when TCP/IP 
was never going to be as good as what carriers could provide, so why 
bother?  Tuenred out that cheap, ubiquitous and end-user defined trumped 
QoS for enough people to make it useful.

Every market has licensed, fixed wireless providers.  Why must 3650-3700 
seek to replicate these conditions?  Many, many people have said they will 
be happy with what they can get NOW rather than rely on a few licensed 
carriers selling high-priced services, even if that means a step down in QoS.

And, if history is any guide, the QoS will improve as people use the band 
and try to solve the problem.  The inability to conceive the perfect answer 
today says nothing about what clever folks on the ground will develop tomorrow.

Harold


>In the end, the rules in reality do nothing but create a sort of limbo of a
>Mexican stand-off, where the one for sure it that decent use of the band
>won't happen for a long time. ...not unlike 5.4 GHz.
>
>If anyone actually read our reconsiderations (the WiMAX Forum's,
>Intel/Alvarion/Redline's, and the WCA's), then one would see that we are
>trying to create an environment where both unlicensed and licensed has an
>opportunity to thrive, an environment where Wi-Fi or WiMAX can be
>immediately deployed (in either the unlicensed or licensed domains), and an
>environment where QoS can be truly offered, ...or not. We went out of our
>way to make our filing technology neutral, though certainly usable for WiMAX
>devices. And if you truly understood the WiMAX vision of "always best
>connected", 802.11 (as well as Bluetooth and advanced cellular) plays a
>vital role.
>
>I challenge you or anyone else to show me in any of the three filings one
>sentence that favors any WiMAX (or even 802.16) over Wi-Fi (or any 802.11).
>On the contrary, one can pull ample examples where those that only can play
>using 802.11 have direct hostility and fear of any other technology. [Intel,
>as one making perhaps more off Wi-Fi than anyone else, hardly has a
>motivation to destroy 802.11 and any theory otherwise is just plain silly.]
>
>Kind regards,
>
>Patrick J. Leary
>Assist. V.P., Marketing
>Alvarion, Inc.
>cell: (760) 580-0080
>Skype: pleary
>
>-----Original Message-----
>From: Harold Feld [mailto:hfeld at mediaaccess.org]
>Sent: Wednesday, June 29, 2005 10:46 AM
>To: Network Summit Mailing List; Open Spectrum; CWN Policy; Unlicensed
>Advocates
>Subject: [Openspectrum] Spectrum updates
>
>(My apologies for any duplicates people see as a result of cross-posting)
>
>Important doings at the FCC and the Hill this summer.
>
>FIRST, a relatively minor thing but with significant applications.  The FCC
>Enforcement Bureau has declared the use, sale, or marketing of cellular
>jammers illegal.  These devices are manufactured and marketed in other
>countries, but are not made in the U.S.  In addition to invoking the
>requirement that such devices must receive Commission certification under
>Sec. 302A, the Commission invoked Sec. 333 which prohibits interference
>with any signal licensed or otherwise authorized by the Commission.  A copy
>of the declaration is available here:
><http://hraunfoss.fcc.gov/edocs_public/attachmatch/DA-05-1776A1.doc>
><http://hraunfoss.fcc.gov/edocs_public/attachmatch/DA-05-1776A1.pdf>
><http://hraunfoss.fcc.gov/edocs_public/attachmatch/DA-05-1776A1.txt>
>
>This has significant implications for what is being referred to as
>"suppression of rogue APs" and also for devices deliberately designed to
>interfere with wireless networks.
>
>SECOND, as some of you may recall, the FCC opened up the 3650-3700 MHz band
>to a new form of "licensed" access.  Although technically licensed and
>therefore under Part 90 rather than Part 15, the "license" is non-exclusive
>and grants no better rights with users based on past use or priority (no
>"first in time, first in right" usual in traditional "licensing-lie"
>schemes).  The rules also require users to use "contention based protocols"
>to share the band.  The rules permit "licensees" to operate fixed base
>stations of up to 25-watts EIRP in power.  Mobile devices may operate on
>the band at a strength of 1-watt, but must receive an enabling command from
>a stationary base station.  The band has geographic carve outs to protect
>incumbent satellite earth receiver stations.  However, because the new
>devices are a licensed secondary service, the FCC has required the
>incumbents to negotiate with those seeking to operate systems in the
>exclusion zones to operate on a non-interfering basis.  The exclusion zones
>effectively eliminate use in the top 25 market areas in the United States.
>
>Other than the incumbent users located along the coasts and in the midwest,
>the band is unused.  The rules prevent an accumulation of "junk" such as in
>2.4 GHz or 5.8 GHGz.  Staff have indicated that they regard this as a test
>case whether developers and those using non-exclusive spectrum for
>broadband can genuinely cooperate.  If this model works, it may be used in
>other areas to create opportunities for non-exclusive use alongside
>exclusive licenses (e.g., broadcasting).
>
>Intel has organized resistance to the new rules from what I will call "the
>WiMax posse."  Approximately ten Petitions for Reconsideration were filed,
>most of which advocate for altering the rules to accommodate WiMax (either
>move to true licensing or eliminate the requirement for contention based
>protocols and create a "first in time, first in right" regime so that the
>first carriers to deploy can set up WiMax networks.
>
>I have received inquiries from some tech companies as to whether the WISPs
>and the community wireless communities will "defend" the 3650-3700 MHz
>band.  The influx of hundreds of comments from WISPs and from community
>network supporters had a huge impact on our defeating the proposal from
>Intel to make this a licensed regime in the first place.  Intel is hard at
>work trying to "recapture" the spectrum, and companies considering
>developing equipment for the band are hesitant to do so unless there is a
>sufficient show of interest from the would-be beneficiaries to give them
>confidence the FCC will keep the rules.
>
>MAP will draft oppositions to the Petitions for Reconsideration.  In
>addition, I would like to urge people to file individual comments in
>opposition to the recon petitions.  For those wishing to file comments, the
>docket number remains the same, 04-151.
>Oppositions are due 15 days after the Commission puts the Petitions for
>Reconsideration on public notice in the Federal Register.  Given how fast
>the Fed Reg publication is, we are looking at late July at the earliest for
>Oppositions.  Nothing, however, prevents interested parties from filing at
>any time, as the proceeding is designated "permit but disclose."
>
>In addition to the FCC front, I would be very interested in talking to
>anyone in the open spectrum/open source community about developing open
>source solutions for the 3650-3700 MHz band.  This is a real opportunity
>for open source to get "first mover advantage" in a band.  The IETF
>presents a possible vehicle for a recognized standard that could
>subsequently be adopted by chip set manufacturers.  FCC staff have
>expressed interest in helping open source developers navigate the approval
>process (their certification people are very open and friendly).
>
>THIRD, Congress is considering the digital television
>transition.  Consensus is that this bill represents a "must pass" over the
>objections of the National Association of Broadcasters because they want
>the auction money.  From an unlicensed perspective, however, the DTV
>legislation represents both an opportunity and a threat.  On the
>opportunity side, Congress could either set aside spectrum for unlicensed
>access or order the FCC to set rules to permit unlicensed access in the
>"white spaces."  On the threat side, the NAB is seizing this opportunity to
>get legislative language in prohibiting any unlicensed access in the
>broadcast bands as a danger to the "digital transition."
>
>As members of Congress head home for the July 4 recess, it creates an
>opportunity to educate them about the importance of unlicensed spectrum
>access.  Needless to say, most memebrs of Congress have never even heard of
>"unlicensed spectrum," but education must begin somewhere.  If you can,
>take time to call the in-state office of your Senator or Representative and
>explain that you think the DTV transition should serve ALL Americans by
>getting us all access to spectrum and tell them about the issue.  Even
>better, a fax and a follow up phone call to simply raise the consciousness
>of the possibilities and to make staffers more receptive to the arguments
>of the folks in Washington fighting for unlicensed spectrum will be very
>helpful.
>
>Happy July 4,
>
>Harold
>
>_______________________________________________
>Openspectrum mailing list
>Openspectrum at listserv.media.mit.edu
>http://listserv.media.mit.edu/mailman/listinfo/openspectrum
>
>
>This mail passed through mail.alvarion.com
>
>****************************************************************************
>********
>This footnote confirms that this email message has been scanned by
>PineApp Mail-SeCure for the presence of malicious code, vandals & computer
>viruses.
>****************************************************************************
>********
>
>
>This mail passed through mail.alvarion.com
>
>****************************************************************************
>********
>This footnote confirms that this email message has been scanned by
>PineApp Mail-SeCure for the presence of malicious code, vandals & computer
>viruses.
>****************************************************************************
>********
>_______________________________________________
>Openspectrum mailing list
>Openspectrum at listserv.media.mit.edu
>http://listserv.media.mit.edu/mailman/listinfo/openspectrum