Yeah, I saw somewhat of an explanation from ubnt at one point... I don't remember the details, but I'm pretty sure FEC was part of the reason.
For an $800 radio, they're good at what the do and they have their place, but people shouldn't expect them to perform the same as a $4000 radio. On May 25, 2017 9:15 PM, "Eric Kuhnke" <eric.kuh...@gmail.com> wrote: I'm going to eat my own words a little bit here. Just compared the AF11FX datasheet side by side with the SAF Integra for a theoretical 40 MHz single polarity FDD link. The AF11FX is something like 252Mbps, the SAF is a claimed 315 Mbps in the same 40 MHz wide. My theory is that ubnt has done something that is lower cost with the radio circuitry that doesn't allow it to very closely approach the edges of a 40 MHz channel without excessively strong signal leaking over the edge of a channel mask (as viewed by a spectrum analyzer), and possibly has additional FEC which is opaque to the end user. But that makes total sense when comparing $799/unit to $4000/unit+. https://dl.ubnt.com/datasheets/airfiber/airFiber_AF-11FX_DS.pdf https://www.saftehnika.com/files/downloads/4e6954be-a416- e611-a0d1-0050569a8c0f/Integra%20series%20DS%20v1.39.pdf On Thu, May 25, 2017 at 7:12 PM, Bill Prince <part15...@gmail.com> wrote: > I don't know that anyone has really examined the overhead required by OFDM > (AF11) versus the overhead required by the various other licensed vendor > proprietary modulation schemes. I think this would be the primary issue. I > know we get over 500 Mbps through a Dragonwave on a single 50 MHz channel. > Two channels should theoretically get over 1 Gbps. The AF11 comes out of > the chute doing dual polarity, so that would be the comparison. > > > bp > <part15sbs{at}gmail{dot}com> > > > On 5/25/2017 7:06 PM, Eric Kuhnke wrote: > > If all you can get on a particular path is a theoretical single 40 MHz > wide FDD channel pair, one polarity, I don't see how the 1024QAM bps/Hz > efficiency would be significantly worse than a competing single polarity > product (SAF Integra, etc) running in the same channel size. Unless you are > counting more expensive competing products that advertise header > compression and very different Mbps rates for 64-byte vs much larger packet > sizes. > > It's very cost effective so I will forgive it many things, my main problem > is that it can't actually *use* near the full width of an 80 MHz channel. > > On Thu, May 25, 2017 at 6:26 PM, George Skorup <george.sko...@cbcast.com> > wrote: > >> Yeah. Cost is one thing, but if all you can get is a single polarity on a >> particular path, the AF11 is probably one of the last things I'd look at. >> Congestion is a problem around here. >> >> >> On 5/25/2017 8:21 PM, Seth Mattinen wrote: >> >>> On 5/25/17 18:12, Mathew Howard wrote: >>> >>>> >>>> We're running the full 56mhz/MIMO... I haven't been able to get them to >>>> run at 1024qam yet (antennas still need to be fine tuned, it wasn't ideal >>>> weather conditions when we put them up, so I'm hoping we'll be able to get >>>> a bit more out them), so they're only at around 550Mbps capacity (and I've >>>> verified the link will do around 500Mbps with real traffic). >>>> >>>> >>> >>> Only 500 meg with two channels? Crap, I have an old Exalt that can do >>> that with only one channel at 256QAM. >>> >> >> > >
