How does the ePMP compare to the new Ubiquiti Airmax AC system? We have been getting a lot of information on both the Force 110 from DoubleRadius as well as the Ubiquiti Airmax AC system. What are everyone's thoughts? Obviously, the ePMP is shipping right now and the Airmax AC units aren't out yet, so it is hard to compare from a real world standpoint. Have folks been switching over from the original Airmax products to ePMP?
1 Like
The two are pretty different systems at the moment, or at least until Ubiquiti fleshes out the AC firmware more.
Some main points:
- ePMP is a PtP and PtMP 802.11n based product that covers both FCC/US 2.4GHz, and 5GHz covering 5.1-5.2,5.4,5.7GHz.
- Airmax AC is a a PtP 802.11ac based product that currently only covers the 5.7GHz band, IIRC they're still waiting for certification for other 5GHz bands. They're PtMP firmware is still in beta IIRC.
- There are a number of confusing options for Airmax AC... e.g. there's one that just does PtP, another that does PtMP... or is expected to do PtMP when the firmware arrives, there's also ones with the 'air prsim' and others that don't have it, there's others with an off load chip and others that don't. All in all, I find it quite confusing, and UBNT's marketing materials don't help. You just have to read ALL the threads very carefully in the UBNT forums and look for UBNT staff comments to figure out what does what.
- Airmax AC is not backwards compatible with older Airmax TDMA, supposedly they'll introduce compatibility, but there's no timeline IIRC. Based on Ubiquiti's track record, this might never come to pass, or possibly will, but will be very limited/kinda broken (like they did with the old 802.11b/g radios to make them compatible with Airmax).
- There is no 2.4GHz Airmax AC product, and there probably never will be.
- ePMP supports frequency re-use and GPS synchronization, Airmax AC doesn't, and Ubiquiti has mentioned that it never will support re-use or GPS on the Airmax AC line
- In order to get any benefit from "AC' e.g. 256QAM, your SnR needs to be very good, otherwise, it performs just like any other 802.11n radio. In a high noise environment, AC is pretty useless.
3 Likes
Cambium Network equipment works as advertised and is rock solid when deployed in the field.
Ubiquiti is nothing more than an over hyped, under performaning paper weight.
3 Likes
@test123 wrote:Cambium Network equipment works as advertised and is rock solid when deployed in the field.
Ubiquiti is nothing more than an over hyped, under performaning paper weight.
@That's a pretty wide sweeping generatlization there @test123. Ubiquiti is pretty rock solid when properly deployed too.
1 Like
I take a look at the UBNT .ac line to see how it works. I have none running until now. They address some deficit of wifi based gear. They include RF Filters to the APs. There is a PTP and a PTMP version. The PTP has a wider filter for 80 MHZ channel. The PTMP has a sharper filter and is limited to 40Mhz Channel size. The filters limit the usable Band. The lite AP and all CPEs don't have filters. This filters reduces near channel interference not on channel interference. So it helps with colocation even with foreign gear but does not allow frequency reuse.
The second interesting stuff is the additional ASIC which handles TDMA protocol. The CPU's in Wifi gear are quite weak to handle this. With higher data rates and smaller packets this might help to run smooth.
The third interesting stuff is a second listening wireless card to do frequency scans in background. This helps identifying free channels without interrupting communication. So in case of a DFS event the AP can jump right into a free channel.
Hi,
I don't agree with you. I have +3000 ubiquty airmax on the field and since one year I'm not installing anymore.
They are correct for the starting WISP (cheap and easy) but they are crap for the demanding WISP. Also they have a name: what to install? install ubiquti.. everybody is doing... it's like microsoft but from far is not the best...
I have set-up the new areas with epmp1000 and ubiquti areas I'm upgrading with albentia wimax. The epmp1000 hardware is rock solid. the performance is somehow predictible. In ubiquti nothing is predictible. Performance don't degrate with distance. Ubiquti due to the ack protocol degrades A LOT. There are lots of out of control parameters which affect the overall AP performance. The airmax is vaporware.. it works like CTS/RTS protocol but it seems there are no other benefits..
Remember I'm talking about demanding WISP with challenging customer setups.. If you only have an AP with 10 SMs at few hundreds meters, then install ubiquty, it is the cheapest and will work acceptable.
Regards
Antonio
3 Likes
Hi Antonio,
I guess you respond to my post. I don't claim UBNT to work good. I have only once tested M5 Rocket and SM in an early phase. Firmware was horrible so we did not use it. We have heard from others that the product matured and is working quite well. We also recognized this faulty GPS sync.
What I am saying is that they addressed some points and offer interesting features. If this is usable now or in the future is up to testing. If you look in their forum everything is fine. If you look at the release notes you find some nasty bugs fixed shortly. So may be this needs some time as the M5 did.
I'll throw a sector up and look how it is doing and if promises matches reality.
Stefan
Hi,
Him forum is plenty of fanboys. It's a shame and very frustating nobody of the "experts" in him forum have never a problem as simple of a poe failure.
As told, if you have a very simple setup, near customers, very good signals, nothing strange, it works. This scenario is very limited in real world.
Also is very frustating the high rate of device failures: flash memory, power supplies, etc.. affecting both APs or CPEs.
1 Like
@roanwifi wrote:Hi,
I don't agree with you. I have +3000 ubiquty airmax on the field and since one year I'm not installing anymore.
They are correct for the starting WISP (cheap and easy) but they are crap for the demanding WISP. Also they have a name: what to install? install ubiquti.. everybody is doing... it's like microsoft but from far is not the best...
I have set-up the new areas with epmp1000 and ubiquti areas I'm upgrading with albentia wimax. The epmp1000 hardware is rock solid. the performance is somehow predictible. In ubiquti nothing is predictible. Performance don't degrate with distance. Ubiquti due to the ack protocol degrades A LOT. There are lots of out of control parameters which affect the overall AP performance. The airmax is vaporware.. it works like CTS/RTS protocol but it seems there are no other benefits..
Remember I'm talking about demanding WISP with challenging customer setups.. If you only have an AP with 10 SMs at few hundreds meters, then install ubiquty, it is the cheapest and will work acceptable.
Regards
Antonio
I have never seen performance degrade with distance with Ubiquiti. If you're installing within specs you should see full rates. Obviously, if you install a -70 you're going to get the modulation that comes with that signal. That's just physics.
By "challenging customer setups" I feel like you are saying "improper installs". Are you installing customers without clear LOS at distances that exceed modulation levels for receive signals at those distances? That's not challenging customers, that's poor engineering.
On the other hand, if you're saying ePMP works fine in those types of setup, then that's a win for ePMP.
@roanwifi wrote:Hi,
Him forum is plenty of fanboys. It's a shame and very frustating nobody of the "experts" in him forum have never a problem as simple of a poe failure.
As told, if you have a very simple setup, near customers, very good signals, nothing strange, it works. This scenario is very limited in real world.
Also is very frustating the high rate of device failures: flash memory, power supplies, etc.. affecting both APs or CPEs.
I think probably the "fanboys" don't have the problems other people do because they follow best practices. I've got probably 25-30 Ubiquiti sectors running, with 30 and 40 subs on each one. We regularlly push 60megabits/second through the sectors with low latency.
I won't install anyone below a -70, and AMC and AMQ need to be above a 75% on the CPEs, as well as the AP AMC and AMQ need to be above 80%.
Are you saying ePMP allows you to just install a bunch of "terrible" clients and everything works well.
Hi
I have more than 50 rocket m5 aps. Lots of them are at 40-50 customers each. For sure I follow best practices. I'm in the business for +10 years.
My target RSSI is -55 so you may imagine what is AMC,AMQ,CCQ, signal levels, etc. My "no install" is lower than -70 and only for very far customers (which have differentiated APs for "near" customers).
For example: an AP with 50 customers, all between 4 Km distance, 20 MHz. All at ~ -55 .. -60 dBm at CPE, -60..-65 seen at AP. The AP is unable to move more than 15-20 Mbit/s REAL TRAFFIC which is a mix of streaming, browsing and voip. Others APs which CPEs are at few hundreds meters, the performance will go up quickly. Idem for APs with CPEs at > 13-14 km. The AP is unable to move more than 8-10 Mbit/s...
This isn't the Ubiquiti forum so I won't digress us further off the original topic other than to say that something is not right with that setup.
Your return signals should be equal to your signal at your CPEs (e.g. a -55 at CPE equated to around a -55 at your AP). You should definately be able to pass more traffic 20megabits.
This the reason I'm in the cambium forum now and don't waste more time with ubiquti.
With the same parameters and RSSIs as ubiquity:
-ePMP1000 with ~25 users => + 20 Mbit/s and ~2500pps
-PMP450 with ~90 CPEs => +60 Mbit/s and ~8000pps.
I never seen an ubiquity AP in PtMP with more than ~2500pps and this is absolutely related to distances. Apart of impredictive performance based on customer data (rssi, distance, etc..) this is what really kills the AP performance.
2 Likes
@roanwifi wrote:This the reason I'm in the cambium forum now and don't waste more time with ubiquti.
With the same parameters and RSSIs as ubiquity:
-ePMP1000 with ~25 users => + 20 Mbit/s and ~2500pps
-PMP450 with ~90 CPEs => +60 Mbit/s and ~8000pps.
So hold on a second. Are you saying the ePMP1000 is only good for about 25 users and 20Mbit/s? That sounds like it's no better than the Ubiquiti products. Am I misunderstanding?
@roanwifi wrote:This the reason I'm in the cambium forum now and don't waste more time with ubiquti.
With the same parameters and RSSIs as ubiquity:
-ePMP1000 with ~25 users => + 20 Mbit/s and ~2500pps
-PMP450 with ~90 CPEs => +60 Mbit/s and ~8000pps.
I never seen an ubiquity AP in PtMP with more than ~2500pps and this is absolutely related to distances. Apart of impredictive performance based on customer data (rssi, distance, etc..) this is what really kills the AP performance.
This does include the .ac system ?
>>So hold on a second. Are you saying the ePMP1000 is only good for about 25 users and 20Mbit/s? That >>sounds like it's no better than the Ubiquiti products. Am I misunderstanding?
I may explained wrong. What I would say is that this is the more loaded ePMP1000 system I have at the moment. The ePMP1000 are in new areas we're still developing.
Regarding .ac I have not used and don't plan to do. There are too much incertidumbre about kind of AP (lite, no lite, airmprism, no airprism), kind of SM, king of SW, etc... In my area, 256qam is unusable due being impossible to get such CINR. Also channels bigger than 20 MHz are very hard to find and too sensitive to interference.
Also GPS syncronization is mandatory due to heavy usage of 5 Ghz spectrum and need to reuse channels.
I still can't follow you. Are you saying your ePMP1000 system will only support 25 users? Or that that's just all you have on the system currently but you think you can put more on?
@roanwifi wrote:>>So hold on a second. Are you saying the ePMP1000 is only good for about 25 users and 20Mbit/s? That >>sounds like it's no better than the Ubiquiti products. Am I misunderstanding?
I may explained wrong. What I would say is that this is the more loaded ePMP1000 system I have at the moment. The ePMP1000 are in new areas we're still developing.
Regarding .ac I have not used and don't plan to do. There are too much incertidumbre about kind of AP (lite, no lite, airmprism, no airprism), kind of SM, king of SW, etc... In my area, 256qam is unusable due being impossible to get such CINR. Also channels bigger than 20 MHz are very hard to find and too sensitive to interference.
Also GPS syncronization is mandatory due to heavy usage of 5 Ghz spectrum and need to reuse channels.
I just mounted a .ac instead of a test ePMP and tested with a Nanobeam AC 19 2km apart. Sorry cambium you have to do .ac. This is a dirty installation with an Omni Antenna but the SM managed to get 256QAM on a 20MHZ Channel and pushed 140MBit UDP/100MBit TCP on the downlink. ePMP had much smaller numbers.
@ste wrote:I just mounted a .ac instead of a test ePMP and tested with a Nanobeam AC 19 2km apart. Sorry cambium you have to do .ac. This is a dirty installation with an Omni Antenna but the SM managed to get 256QAM on a 20MHZ Channel and pushed 140MBit UDP/100MBit TCP on the downlink. ePMP had much smaller numbers.
@ste - those are closer to the numbers I'm used to seeing with Ubiquiti gear, granted you're using AC and I'm using airMax N gear, but still I'm used to seeing 60-70megabits on a 20MHz channel with regular gear and yes upwards of 100-130 on AC.
I'm still trying to fully understand what was being said by the other poster.
So you mounted a Point to Point with .ac equipment? 2 Km apart? and it was modulating at 256qam because you have enough signal strenght and CINR.
One interesting test would be know how speed scales on distance (2km, 10km, 20km etc), keeping the link budget in order to maintain 256qam and CINR... If the .ac protocol uses ARQ then bandwidth impact should be minimal. If this .ac system uses the ACK mechanism like airmax the performance will degrade quickly as each km is 3.3microseconds to travel the packet, each way.. These silences in the airtime kill the performance very quickly.
Also important to know if such speeds changes with large packets or small packets. One way or two way.
Anyway this scenario is very different to having a P to Multipoint because lot of useful things of wifi n like packet aggregation which work very well on PtP don't perform too much due to having multiple MACs in play in PtMP... multiple customers at different distances, etc.
Have a look at this: https://albentia.wordpress.com/2015/06/01/wimax-latencia-capacidad-interferencias-wisp/ to know the differences of a real TDD like albentia pmp450 or epmp1000 compared to airmax..... (sorry only in spanish)