Tom, You correctly identify the Achilles Heel of modern day UL -- the survival of the nastiest phenomenon. The Part 15.247 rules give equal standing to all types of systems, regardless of how spectrally gluttonous or abusive. The problem with this is that it rewards downward innovation (i.e. dumb and inefficient), offering no incentives for developers to invest R&D to come up with more efficient, higher performing PMP techniques. There is no reward for high performance, especially in PMP where devices are most vulnerable to interference. This is a genuine reason why there is so little real investment in PMP UL. I am not talking about the creative, small market developing and tinkering that goes on, but rather the multi-millions of serious R&D investment such as that seen on the licensed side.
As well, the logical extension of this problem is that WISP operators themselves are not rewarded in a spectrum sense by using the most efficient systems. [I should insert a note here that recognizes that bringing to market a system that might be considered spectrally abusive so that it itself survives, all while conforming perfectly within the regulations, may be considered to be an entirely sound, even smart, competitive strategy -- the rules do not require me to play well with others, so I'm going to do everything I can to make sure I do not, within the rules of course. However, markets are not fond of investing in R&D to get around artificial problems, i.e. problems created by easily manipulated regulations.] Back in 2002 I was one a few panelists representing the UL BWA market (Steve Stroh was there too on another panel) on the FCC's Spectrum Policy Task Force. The panel I was on was to recommend and debate the evolution of the UL bands. Most were up there thinking WLAN, not WMAN, and they did not even understand the implication of their proposals in the outdoor, metro scale world of wireless. A few were up there saying we needed more power for rural. My input regarded changing the rules to allow for some type of sliding higher power rules based on better efficiency, and that efficiency could come any number of ways, through better and more narrow, high quality (good emitters, without lots of spurious noise) antennas, higher sensitivity and intelligence, better capacity per MHz (especially better packet per second type efficiency), etc. The better one performed, the higher power allowed. To do this, in my head I was thinking that a base line point of measurement would be some type of low performing product connected to an omni. Put that same radio on sector, you get more power. Put a more efficient radio on that omni, you get more power. Put a really efficient radio on a well-performing on a well-performing, narrow beam antenna and gets lots of power. Etc. Etc. I am sure smart people can come up some type of algorithm that incorporates most of the variables that and make something efficient, while leaving room for the formula to advance to accept new techniques that create efficiency. Such a rule would give operators incentive to employ the best systems since such would require the least number of cells and such. And suppliers would be continually encouraged to invest and innovate, because we'd know that the market would be encouraged to support new technology for the rewards in power and performance to be gained. Another great thing about these proposed rules was that they are technology neutral (the FCC does not like to specify technology these days). The rules simply would have been some type of math formula where the answer was always a reference to allowed power output (EIRP). I also wanted a registration rule, very similar to what the FCC called out for 3650MHz. I wanted it low fee (but enough to provides monies for rules enforcement), non-exclusive, and open, with only registration (not licensing) of infrastructure nodes and their locations. At the time, I really do not think most in the room or at the Commission fully grasped what I was getting at. Maybe I did not articulate it well enough (though it was all captured on video and remains archived on the FCC web site). But, I still think that such rules have the best chance of bringing about a positive revolution in UL BWA. Regards, Patrick -----Original Message----- From: Tom DeReggi [mailto:[EMAIL PROTECTED] Sent: Wednesday, April 19, 2006 5:06 PM To: WISPA General List Subject: Re: [WISPA] UL WiMAX update Patrick, Forgive the "Me to", but Patrick, GREAT POST on UL WIMAX! Your post did not only address Wimax, but also addressed several of the big delimna's for WISPs and the FCC. How to coexist. This industry has grown to the point that MANY WISPs/Players have significant amounts invested in theior strategies. There becomes a big conflict of, "does a Operator/Provider support what protect's their investment or what is best for the industry?". If the Provider has an Ego, they could argue that the survival of their own company could be the best thing for the industry, and therefore favor rules that protect himself (the existing provider). What I'm getting at is technology that protects and preserves spectrum or technology that allows last man standing. The problem is that 5.8Ghz has the most to offer UL WiMax, but supporting it means wiping out all the existing player's investments in infrastructure. In order to support it, the providers must be in a possition to afford/finance/re-invest in it themselves. WiMax probab would have Thrived in 3650, jsut because the technology could be deployed without interfering with any one elses investment in infrastructure or clientel that need to stay up reliably. One of the big problems is that new advanced modulations and technologies that allow super fast speeds, such as QAM64, also require much higher SNRs. Signals that can easilly be squashed by DSSS technologies which operate at much closer SNRs. One of the problems for the industry is that users of the less efficient systems, are rewarded, allowing them to be more survivable. It actually puts the small player in control. The guy that would rather stay at 10 mbps than risks unsurvivable service at 30 mbps. I think one of the things that would really help the industry is new rules, that would allows the more efficient systems to operate at higher power levels, so that they could survive the less efficient systems that operate on closer SNRs. The FCC started to tackle this mentality by adding new rules allowing smart antennas (more directional base antennas) to broadcast at 7db higher output. But I'd argue that possibly the same should be allowed for high modulation gear, so the faster gear could survive the slower. However, there is a catch-22 with that, allowing higher power causes more interference between the neighbor cells of same technology. The point I'm making is that the success of WiMax and efficient radios very well may depend on the availabilty of a band, that is protected from DSSS type gear that has better SNR or pre-existing rules that may compromise its efficient use. The example you used of WiMax Scheduled MAC being designed for License, also applies potentially to gear that requires large SNR to survive, where it may not have been adequately designed to survive the typical noise of pre-existing technologies in the spectrum of Unlicensed. In order for PMP high capacity backhaul to effectively work (as market demand and wareness grows), it should also have predictable consistent capacity. That is a principle that is the opposite of Contention based. I guess the point I'm really saying is that it is somewhat a no win situation. Can the conflict be resolved (contenstion vs scheduled vs customers' need )? I believe in evolution and survival of the fittest, and the FCC as well has always showed support for technologies that showed innovation and better use as technology adapts. The idea is as smarter technology gets designed, it will evolve to take the place of the lesser product that market pressure will phase out. The problem is, is that how it will really play out? Maybe the gear that is the most resilent (even if it delays innovation and progress) will survive apposed to the technology that is more efficient, faster, and honorable to edicate? At what point should the FCC step in and say, lets make rules that protect spectrum for a specific purpose or technology, apposed to licensing it to an individual or leaving it as a free for all and evolution's rule? It will be interesting to see what the UL WiMax TG or 802.16h TG come up with. Its alway better, if the problem can be solved with engenuity and inteligence instead of legislation. Tom DeReggi RapidDSL & Wireless, Inc IntAirNet- Fixed Wireless Broadband ----- Original Message ----- From: "Patrick Leary" <[EMAIL PROTECTED]> To: "'WISPA General List'" <[email protected]> Sent: Tuesday, April 18, 2006 11:49 AM Subject: RE: [WISPA] UL WiMAX update > Well George, ready for long answer that may not actually answer your > question? I'd prefer to give you the full story. > > First,...so, is Alvarion building UL WiMAX? Of course, and I personally > see > lots of potential for it. When will it come? A few things are in line > first, > so there is no firm date but we'll have it roughly around the same time as > other main suppliers. If I could give a better and more useful date, I > would.UL > > Second, WiMAX is not a simple story. Here are the issues revolving around > it: > > 1. 3650MHz is a better UL band for WiMAX than 5.8GHz: > Vendors and operators know that this band is more favorable for a scaled > BWA > deployment than 5.8GHz for both reasons of physics, higher power > allowances, > and less interference. So far, the only UL profile for WiMAX is > 5.725-5.850GHz. But most vendors are not eager to invest too much in that > profile while 3650MHz is up in the air. If 3650MHz goes UL, as it most > likely will, at least in part, then that would take the wind out of 5.8GHz > WiMAX's sales and a new profile will have to be created to support > 3650MHz. > > 2. The UL profile is limited to upper 5GHz only: > The UL WiMAX profile excludes 5.25-5.35GHz, as well as 5.47-5.725GHz. That > is 355MHz of spectrum that the WiMAX Forum so far does not support. Who > wants to build a UL WiMAX network that only uses 5.8GHz? The profile needs > to be broadened. > > 3. The scheduled MAC of 802.16 is designed for licensed: > The reality is that the 802.16 MAC was originally developed for licensed > LMDS bands. In order to push through a standard quickly, when 802.16 was > amended to be applicable to sub-11GHz frequencies, they co-opted that same > MAC. Now it's a great MAC...for licensed. Scheduled MAC's are highly > efficient, but they are intended to be used in licensed where the only > interference risks are self-inflicted. With a scheduler, when your slot > comes to talk, you talk, regardless of what is happening in the spectrum. > In > the UL world where there is contention for the spectrum, a scheduler > results > in lost packets AND hurts the other systems already in the air. > > The IEEE knows this is a problem, so they formed a new task group about 9 > months ago called 802.16h, or TG H. The charter of this task group is to > come up with a mechanism that somehow enables UL co-existence of systems > using shared (UL) spectrum. The idea of the TG is to find some type of > technology neutral soft patch that can be overlaid atop not just any .16 > device, but any 802.11, or even proprietary system. Alvarion chairs this > TG. > It is a tough nut, because we and the IEEE are trying to make this a joint > TG with the 802.11 crowd, but so far the 802.11 groups in the IEEE refuse > to > joint. The challenge is that the TG can come with some super slick > technique, maybe some time sharing mechanism, but unless other systems in > the air adopt it, it will not be as effective as it would otherwise be. > > Suppliers are aware of all this and it adds to the reluctance to release > UL > WiMAX as it exists today. > > 4. The UL WiMAX profile was designed for PMP backhaul, NOT last mile > access: > Most may not be aware of this, but if you take note that the > channelization > options in the 5.8GHz UL profile are 10MHz and 20MHz, you come to realize > that the intention is to make big pipes. Consider that the current > efficiency of WiMAX is a bit better than 3.5Mbps NET usable throughput per > megahertz used and you'll see that in UL WiMAX you can create pipes > delivering over 70Mbps NET in a 20MHz channel. Then note that the last > mile > centric licensed profiles deal in 3.5MHz and 7MHz wide channels. You > quickly > begin to realize that UL WIMAX is intended for backhaul only, for things > like mesh clouds, hotspots, and outdoor PMP enterprise bridging. > > What does this mean? This means that the market is scrambling to build > residential CPE for UL WiMAX. Instead, the CPE will be that you would > expect > at the remote end of an enterprise bridge or backhaul. In other words, we > are not talking about sub-$200 devices. > > 5. There will be no indoor only, self-install UL WiMAX CPE: > Unlike licensed WiMAX, for which the power and bands are suitable to > support > a no-truck-roll CPE, we have no such luck in 5GHz. This leaves us with the > same installation paradigm we live under today in the UL world. > > 6. UL WiMAX profile in only supported in the fixed WiMAX standard of > 802.16-2004. There is no profile for 802.16e-2005: > While we and a handful of others remain excited about fixed WiMAX, most of > the large telecom suppliers are bypassing it entirely and going straight > to > 802.16e-2005. Now, and this is key, while the -2005 standard is about > mobile, IT CAN be used also for fixed and it WILL be the basis of nomadic > and portable (semi fixed, self-install) CPE. So that is where all the big > R&D money is at now and vendors planning to participate in the main WiMAX > market (the 802.16e-2005 world) have to invest to stay ahead. This makes > 802.16-2004, and all the profiles that go along with it, including the UL > profiles, a lesser priority, at least relative to 802.16e-2005. > > > So the net result of all these issues is that the suppliers are cautious > and > not certain about the market size for UL WiMAX, or even who will be the > big > buyers. Is there a large enough market to drive down prices? Who knows, > currently the main growth in the UL BWA market is happening at the Wi-Fi > muniwireless level and there is a sense that this needs to play out, even > while WiMAX may make a good PMP backhaul solution for those projects. > > I continue to hear that some vendors out there continue to say something > like, "we have UL WiMAX just around the corner!" the fact is that as of > the > last WiMAX Forum meeting (in Paris last month), not a single vendor had > yet > submitted UL WiMAX product for certification testing. And keep in mind > that > it takes three to submit before any testing can even begin. > > I do not hear enough suppliers being blunt to WISPs and others about all > these issues revolving around UL WiMAX and the WiMAX Forum itself needs to > be more clear about the different types of WiMAX. > > Also, as operators you really have to ask yourself, what do I want out of > UL > WiMAX? Can you get that with other or current gear? Finally, when UL WiMAX > product first hits the market, if it comes from a new entrant with no > current customers to piss off, approach it with healthy skepticism and see > if it addresses the issues put forth above. Heck, do that even if it comes > from respectable long term players like us. > > Lastly, I'd advise that while you keep an eye out for progress on the UL > WIMAX front, you continue to deploy and put your faith in current > technology. From our end, solutions like BreezeACCESS VL are developed > specifically for the UL world and they are mature (over 300,000 installed > units) and ever improving. We will continue to invest in VL for the > foreseeable future, including making sure it supports ALL the 5GHz ranges > so > operators have choice and can scale. And to lend weight to the "continue > to > invest" claim, we have a firmware version entering beta right now that > will > blow the doors off you. We have basically stripped it down and rebuilt it. > Not only is it enabling a massive packet per second increase, but we have > added a prioritization feature that allows an operator to prioritize > certain > types of traffic network wide. For example, it can be configured to let > all > voice (or video) traffic from all CPEs be transmitted first, with less > time > sensitive traffic like basic data released second. This is true no matter > the users' placement geographically in the network -- on the edge or near > the base station. At the same time, we have a starvation prevention > algorithm to make sure other apps are not starved out. This all happens > dynamically. We have tentative (lab) data showing this improves concurrent > voice sessions per sector from 40'ish to over 250. We'll see how this > plays > out in our voice and video betas. > > In addition, we have added QinQ VPN ability, so individual clients can > create secure VPN tunnel within an operators own VPNs. And there are a > host > of other new features. > > Obviously, this is the one example I can speak about with authority, but I > am sure the other vendors will continue to invest and support their own > current UL technologies for a long time to come. > > Regards, > > Patrick Leary > AVP Marketing > Alvarion, Inc. > o: 650.314.2628 > c: 760.580.0080 > Vonage: 650.641.1243 > -----Original Message----- > From: George [mailto:[EMAIL PROTECTED] > Sent: Monday, April 17, 2006 5:52 PM > To: WISPA General List > Subject: Re: [WISPA] Quick note of hello > > Welcome back Patrick > > How is Alvarion doing concerning UL WIMAX? > > George > > > Patrick Leary wrote: >> Hi all, >> >> >> >> I just wanted to drop you guys a note that I have re-subscribed after >> being off the list for maybe two years. Hope all is well. >> >> >> >> Patrick Leary >> >> AVP Marketing >> >> Alvarion, Inc. >> >> o: 650.314.2628 >> >> c: 760.580.0080 >> >> Vonage: 650.641.1243 >> >> Skype: pleary >> >> >> >> > > -- > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ > > > > > > -- > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ -- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
'Patrick Leary'.vcf
Description: Binary data
-- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
