We currently use two 11Ghz brands.... Trango and Dragonwave. We are happy with there best of class range and link tools.
Again.... 11Ghz, Distance, antenna, # of 9s is not enough info for discussion. Its all about "RainFade" characteristics for an area, and modulation. Its why 70Ghz can go 7 miles in Arizona, and barely 1 mile on Mid-Eastcoast US. Same reason Travis can claim his reliable 20+ miles 18Ghz link on his mid-western network, but we on mid-east coast shoot for 3.5 miles w/ 3ft dishes in 18Ghz. Both Dragonwave's and Trango's path calc tools give pretty accurate results, and both brands live up to the spews they advertise. When we've installed our links we usually get the RSSI DB that the tool predicted we would get. Its also important to recognize some tools, will pitch uptme assuming the use of adaptive modulation, and that fall back to low modulation is considered uptime. That makes the assumption that the adaptive modulation will react properly without down time, both in switching up and switching down. And should consider than durring that time marginal packet loss might occur, and could throw OSPF or BGP off sync, if being used, which could add additional time to full restoration of routing. The fade we get during storms tends to be a bit worse than the tools predict though. Or I should say that the number of bad storms tends to be more frequent than the RainZone data states. The last two years weather around here has really gotten much worse than it typically was the rest of my life. More snow, more rain, more heat. Maybe its Global Warming in action. So again, thats why I recommend using the tools as the "best case", meaning you can do as well as the tool says, but not likely better, and possibly a bit worse. I listened to Sprint discuss some of its studies a few years back, and they brought forward the fact that most rain calculations consider that all rain drops are the same size, and are only considering an "average size" and looking at average rain fall. But in reality, rain drop size and deminsions can drastically change based on different storms, and therefore the actual fade loss can vary considerably, for two different rains of the same rain rate. Thats not factored into tools. Sprint suggested that with 70Ghz, that the tools would predict one 9 better than it really was. Its feasible to assume that at 11Ghz it may be less effected than 70Ghz. The other thing we find is that... If there is a calculation of 30 minutes yearly downtime, that would never occur in one 30 minute increment. That could translate to a 5 minute outage every other month. Or a 5 minute outage twice a month during peak rain season during second quarter, aka April Showers. That can be much more annoying to I will add... several of the places that we installed 11Ghz links really had to stay fast, at a higher modulation (QAM64/128/256), and anything less would be as bad as an outage. As well, we started out using Gigas, which did not have adapative modulation at the time. So we engineered to try to get 99.999% uptime at 128 or 256 modulation always. Where as with the Trango Apex or Dragonwave, their adaptive modulation works very well, and we can push the limits in out path budgets, because we can rely on the adaptive modulation to keep it going, in cases where we can tolerate the slower speeds for short periods. In most cases, rarely ever more than 5 minutes contiguous. If you run the Trango and Dragonwave link tools, with our Maryland rainfade and 3ft dishes, and our speed and uptime expectations, you will see the results tells us 10 miles is about as far as we should go, not to sacrifice reliabilty. And our real world usage has shown that we have more down time that we'd like to have. Sure we can go 20miles, but we dont want to be subject to the reduced %downtime or reduced speed. It really doesn't take all that much better link budget to extend 20 miles, which can be gained by accepting adaptive modulation and lower modulations during weather conditions. It also depnds on whether the link is for a customer of a backbone. When we got a few 100 people behind a backbone, 5 minutes of downtime, is plenty to get the phone ringing off the hook by a few 100 callers. And of course they are all demanding call backs with explanations that they are never going to get. Lets look at a real Dragonwave tool calc.... 11Ghz, 48 rainfade, 19.5 TX power, HC277, 40Mhz, 3ft both sides, 10 miles.... result -45dbm, 14.8 fade, Link Rain availabilty 99.992%, 58 min/yr error. Change to 20 miles.... results... rx -51dbm, 8.6db fade, link rain avail 99.939%, 1440. min/yr error. Change to 20 miles, HC110 (low speed modulation model), results..... rx -43.99, fade 31db, link rain avail 99.989, 68.8 min/yr error. Most of the uptime is gained back, relying on adaptive modulation to lower modulation. Note: min/yr error may not be proportional to rain link avail% because min/yr error is based on "total link availability" which includes both rain and multi-path. Obviously, each manufacturers tool will yield slightly different results based on what the specific radio's receive sensitivity and power level is for that brand, and what modulations that it supports. Tom DeReggi RapidDSL & Wireless, Inc IntAirNet- Fixed Wireless Broadband ----- Original Message ----- From: "Collins, Jim" <[email protected]> To: "WISPA General List" <[email protected]> Sent: Wednesday, September 29, 2010 1:49 PM Subject: Re: [WISPA] 11GHz fade margin > Tom, > > What make/model of 11 GHz gear are you using? One particular > manufacturer shows 11 GHz performing with 5 nines for spans in excess of > 20 miles using standard high performance 2.6 ft antennas. I was just > curious what your manufacturer forecasts vs real life. > > Thanks, > > James R. Collins > > 255 Pine Avenue North > Oldsmar, FL 34677 > 813-891-4774 Direct > 813-416-4039 Cell > 813-891-4712 Fax > > > -----Original Message----- > From: [email protected] [mailto:[email protected]] On > Behalf Of Tom DeReggi > Sent: Tuesday, September 28, 2010 5:26 PM > To: WISPA General List > Subject: Re: [WISPA] 11GHz fade margin > > Marco, > > In Maryland, to get 270mbps reliably, I try not to do any link in 11Ghz > beyond 10 miles or so with 3ft dishes, to get 99.999%. Rain fade > calculated > at about 18db fade in that situation. But still, in heaviest rain, I > dropped > link a few times. > > Obviously with lowest modulation, larger dishes, and lower 9 > expectations, > in dryer climates, you can go much much farther. > Using DragonWave's tool, with Greenville, TX rain data, 6ft dishes both > sides, Highpower (19.5db), 40Mhz, model HC277, you show -42dbm with > about a > 17.5db fade margin, listing 99.978% uptime. (Trango's APEX or GIGAPLUS > probably does as far if not farther, I just didn't have the Trango tool > handy while writing this) > > My point here is, your link has 17db rain margin for a 27mile link in an > > area with a higher rain rate (I think around 66mm/hr), accomplishing a > lower > fade margin than I have for my 10 mile links here in Maryland where the > rain > rate might be around 48mm/hr. So... same fade margin, but your link > three > times longer. Your link will likely drop much more frequently. But will > it? > There is a misconception that a link three times longer could have three > > times the fade, which is not true, because the rain causing the fade > rarely > covers a wide area. For example, the rain storm might just be raining > over > one mile of the path, regardless of the length of the path. What is a > critical factor is the direction of your link, and the likeliness of > whether > the Rain storm would just cross your link path once (moving > perpandicular), > or whether rain storm likely would travel along the path of your link in > > parallel. If the storm followed the path of your link, moving 1 mile at > a > time along the path from one end to the other, the duration in which the > > rain storm would effect your path would be much longer. So not only is > it > useful to predict the heaviest rain and duration in an area, but also > the > directions storms likely move. > > That was a mistake I made... I have a backhaul three cell sites in a row > 10 > miles apart, and almost always when a storm comes through, it hits each > and > every one of the three tower one at a time after each other, as the > storm > migrates. Thus, if a storm causes an outage it causes it three times, > once > for each link it hits. If my towers were aligned perpandicutlar, I'd > have > one third the amount of outages or downtime. > > So yes, the 27mile link can be accomplished with 11Ghz. But yes, you > will > have some downtime, and you need to deside if that can be tolerated for > the > link's pupose. At Full modulation the tool says 728min of outages. > You'll > have to rely on adaptive modulation, and the lower modulations speeds > during > rain and fade. At 100mbps it has 37db fade margine, the downtime drops > to > only 40min/yr, (99.997%) which is way more acceptable. > > You can do some calcs and see that if you changed the design to be three > 19 > mile hops, and the uptime would go down to only 11 min/yr w/ adaptive > modulation down to100mb. But then, you'd have 1/3 more expense. > > I guess this boils down to whether your need of capacity versus uptime > is > more important. a 100mbps 5.8Ghz or 6Ghz link will have much better > uptime > at 27miles. > If you need higher capacity, then 11Ghz will give it to you, most of the > > time 99.97% of it, but you'll have some occassional down time. > > What I'm learning is to both 1) trust the path calc tools, but 2) also > to > realize there are other factors that can degrade the real world results, > and > should look at the tool as being the best case. Thinks that can > contribute > to worse are.... antennas that move, antennas that get misaligned, noise > > that develops, cables that fail, adaptive modulation or rebooting slow > to > respond, that could result in additional or premature downtime. > > Tom DeReggi > RapidDSL & Wireless, Inc > IntAirNet- Fixed Wireless Broadband > > > ----- Original Message ----- > From: "Marco Coelho" <[email protected]> > To: "WISPA General List" <[email protected]> > Sent: Tuesday, September 28, 2010 12:51 PM > Subject: [WISPA] 11GHz fade margin > > >> I'm looking at deploying some 11GHz gear. I would like to do one path >> in two 27 Mile Hops. Using 6' dishes I show a fade margin of 19db. >> Is this adequate for 11GHz at that rage? At 5GHz - 6GHz, I would be >> fine with it. >> >> Is anyone else pushing 11GHz this far? >> >> -- >> Marco C. Coelho >> Argon Technologies Inc. >> POB 875 >> Greenville, TX 75403-0875 >> 903-455-5036 >> >> >> > ------------------------------------------------------------------------ > -------- >> WISPA Wants You! Join today! >> http://signup.wispa.org/ >> > ------------------------------------------------------------------------ > -------- >> >> WISPA Wireless List: [email protected] >> >> Subscribe/Unsubscribe: >> http://lists.wispa.org/mailman/listinfo/wireless >> >> Archives: http://lists.wispa.org/pipermail/wireless/ > > > > ------------------------------------------------------------------------ > -------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > ------------------------------------------------------------------------ > -------- > > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ > > > > > -------------------------------------------------------------------------------- > WISPA Wants You! Join today! > http://signup.wispa.org/ > -------------------------------------------------------------------------------- > > WISPA Wireless List: [email protected] > > Subscribe/Unsubscribe: > http://lists.wispa.org/mailman/listinfo/wireless > > Archives: http://lists.wispa.org/pipermail/wireless/ -------------------------------------------------------------------------------- WISPA Wants You! Join today! http://signup.wispa.org/ -------------------------------------------------------------------------------- WISPA Wireless List: [email protected] Subscribe/Unsubscribe: http://lists.wispa.org/mailman/listinfo/wireless Archives: http://lists.wispa.org/pipermail/wireless/
