>Just a few FYI comments:-
> >the efficiency of PHY layer is greatly dependent on the noise and interference >factors. This is the case with every radio. While there are differences how radios perform with interference. >One can 'tune' auto everything mode, by manually locking out channels (i.e. >channels in use by other equipment of yours). If I am correct Mimosa plans to autonegotiate used channels with radios on the same l2 network. >There is another release for the PTP which is due soon 1.3 (I believe) which >is supposed to have some more >interesting improvements....including a >reduction in latency... This will be great. Latency will be an issue when there are chained links. _____ From: "Stefan Englhardt" <[email protected] <mailto:[email protected]> > To: [email protected] <mailto:[email protected]> Sent: Friday, November 6, 2015 9:24:47 AM Subject: Re: [AFMUG] B5 vs AF5X >I’ve got both radios running in different areas. The Mimosa radios are pretty >proprietary in they are actually running about 87.5% of the PHY layer for >throughput. This is with 8ms Frame size I guess. We use only 2 or 4 as latency adds up with multiple links. >At that level, I really don’t think there is that much difference between the >radios. The advantage in the world I live in is that spectrum interference is >constantly changing and that the combination of split frequencies and >auto-everything both make my life easier and seems to maximize thought. Auto-Everything still does not work for me. Still does some wired decisions. At some links the background spectrum scan still shows the own radio as interferer. Using this as information for channel selection does not help. >The AF5x radios however, have 10, 30 and 50MHz channel options whereas the >Mimosa have an 80MHz channel option. Finding the spectrum to maximize the >radios is the real key to optimizing the AF5x and they give you 3 distinct >options. With the Mimosa’s I don’t worry about it, they search all the >spectrum and make the changes in terms of channels and channel widths. We’re very limited in high power spectrum (only 120MHz with 36db EiRP) so I want to squeeze every bit out of it. This is why I look at the AF5X for some places. >What will be interesting is how both radios perform after the next firmware >releases. Both products are expecting big things. I guess you’re talkin PTMP? Would love to see this. I see both Radios have a great hardware base and have the possibility to get better with Firmware. From: Af [mailto:[email protected]] On Behalf Of Stefan Englhardt Sent: Friday, November 6, 2015 12:31 AM To: [email protected] <mailto:[email protected]> Subject: Re: [AFMUG] B5 vs AF5X >You're looking at the difference in code rates between 256QAM 3/4 code rate >(MCS8) and 256QAM 5/6 code rate (MCS9)? All >things being equal in the same >size TDD 40 MHz channel, of course the MCS9 radio will have a greater bps/Hz. No. The AF5X claims to give higher performance with (8x) 256QAM compared to MCS9 (256QAM 5/6) of an .ac radio. AF5X is no .ac radio so they seem to have a different amount of subcarriers, less overhead or other modulation scheme? http://community.ubnt.com/t5/airFiber/AF5X-Link-Calculator-Updated-Download/m-p/1255928#M20955 >Talking about the bps/Hz for a single stream, the specs for 802.11ac say that >an MCS8 channel 40 MHz wide will be 162 to 180 Mbps, while a MCS9 channel 40 >MHz wide will be 180 to 200 Mbps. >I am not sure the two can be compared directly side by side thanks to the B5's >split frequency modes of operation. They can be >better compared head to head >if you're using a single fixed TDD frequency (like, 5760 center channel). Yes. I took 40MHz to be fair as 802.11ac does MCS9 only with channels greater equal 40MHz. And I am talking real IP capacity not physical. On Thu, Nov 5, 2015 at 10:41 PM, Stefan Englhardt <[email protected] <mailto:[email protected]> > wrote: Anyone who have them both running did an unbiased comparison? We’ve several B5 links running and are quite happy with them. Good 11ac radios in a very neat package. The AF5x on the other side is a custom designed radio which seems to squeeze more mbits out of smaller channels. In a 40MHz Channel I see an aggregated thruput of 320 Mbit/s with the B5 with 4ms framesize at MCS9 (256QAM). Looking at the AF5X (inofficial) link table should do 390 Mbit/s aggregated with 2ms framesize. So they promise to do higher bandwidth with lower latency in a 40MHz Channel (at smaller channels the difference is higher as .ac do only MCS8). Is the AF5X the better radio with limited spectrum while the B5 can do more with enough spectrum available? I’ve sites where I ran out of spectrum (ETSI) so I plan to put AF5X there.
