A5-360-18 OMNI dBi?


#1

Since the A5-360-18 OMNI is really 4 antennas inside, does anyone know if these 4 antennas are 18 dBi each?


#2

Given that they are beam forming, my guess is that the individual elements are ~15db and they turn on/off phase-matched elements to steer the beam. The combined power between active elements creating the combined directive power of 18db.


#3

Each panel is 18dBi gain


#4

Now its my understanding that each panel is a single chain. wouldn’t this seriously hinder any clients ability to connect with 2x mimo? Meaning most clients would be half speed with SISO? Also wouldn’t this make GPS sync on the same tower null and void?


#5

This is official from Mimosa:

  1. Each sector is 18dBi.
  2. Currently only Wifi Interop mode on the A5 is a 4x4 MIMO radio. With TDMA it will be acting as 4x4 MU-MIMO.
  3. So the A5 is rated to 100 clients. The max client number I’ve seen from out beta tester is 38 clients on a single A5 in WiFi Interop mode. With the release of of the TDMA firmware we fully expect the client number per A5 to rise. However, keep in mind that latency can rise when adding high number of clients to a A5.

#6

Yes, true 18 dBi.

There’s some overlap between the panels, but practically speaking in the environments it’s intended for, which is quite short range, sub 1km, preferably 600m for full rate performance, there’s a ton of multipath from the “quad directional” MIMO streams. In practice in all the fielded deployments, each client is “seeing” 2, and quite often more streams. It’s also why we’ve seen some users get far better Near LOS and foliage penetration.

One field example we saw with a WISP who called our support, they aimed their C5 at the closest A5-14, but their happened to be a second A5-14 in the opposite direction of the C5 too with the same SSID/Key settings, and really no LOS. The C5 saw that one first behind it, and connected to it weakly with 2 streams too, until we figured out he accidentally connected to the wrong A5! The power of near-in directional multipath reflections.

I would not look at the A5 like you would compare an equivalent sector and longer distances, it really is designed for close in suburban and urban environments, whereas B5c multipoint and A5c clearly are for the distance beyond the aforementioned.


#7

Regarding sync on the A5, when it’s available end of summer, remember these are designed for Micro-PoP style deployments, so they’re meant to be distributed throughout neighborhoods. In that style deployment, you’ll be able to reuse the same channel in GPS sync throughout the neighborhood typically. The A5 distributed devices are never transmitting at different times, so they don’t interfere with each other. The only possible self-interference could come from C5 directional clients shooting “past” the A5 and hitting another A5 in the same beamwidth, even that should still work with our AGC technology so long as the signal received at the unintended A5 is down at least 10 dB versus the other clients on that AP.


#8

Thank you very much for explaining that. I am coming from a much different environment. I am very rural with long distance between. I would not be able to use the multi-path in a lot of situations. I liked the 18dbi gain but will need to go with a sector in most situations.


#9

Unfortunately less then 5 percent of my CPE’s are under 1 mile from tower. Typically range from 1 to 10 miles. Will there be a higher gain C5? I am fine with connectorized AP’s but have no desire to use connectorized CPE’s.


#10

Jaime,

This is good info to have on the website so the use case can be understood. For us, we sometimes put up 13 db 5ghz omnis for 1-5 mile clients on a new site until there is enough demand for 4 sectors. We were hoping the A5-18 would fit the bill for these sites but we are also excited to try it in town in more dense areas to compete with cable and fiber.

Do you still think we’d be ok using the A5-18 in place of the current 13db omnis we use now for rural 1-5 mile shots as long as we use connectorized C5c with 25+ gain antennas? We don’t always need a full sector build for our very rural sites. I’m talking maybe 15-20 clients max on an omni and we would just use a 20mhz channel.

Thanks


#11

I’d defer to someone from Mimosa for the “real” answer, however as long as you have a high enough signal strength, and are using TDMA to overcome “hidden node” syndrome, distance shouldn’t matter much. The C5c coupled with a 16-24" 5ghz dish should be sufficient to get the gain levels required for a -50-55db signal level at those distances, assuming unobstructed LOS and Fresnel zone clearances.


#12

@Jaime we use Ubiquiti 120 degree 19 dBi sector and have clients at the 20 mile range using a 30 dBi CPE dish and getting 20+Mbps. So I don’t understand how an 18dBi OMNI can only serve within 1km, that’s ridiculous.


#13

Remember there is only 1 stream in each 90-degree direction with small petaloid overlaps. In short range, the signal on the CPE will pick up the second stream easily, especially with multipath, that’s been proven incredibly well in the 6 months of customer field testing.

In long range with pure LOS, it likely will be down to 1 stream, and with 4 degree downtilt you’ll need to be high enough up to counteract the tilt since it’s fixed elevation.

That said, of course it will work further, but we don’t have field data about actual performance at longest distances, which is what the A5c and B5c multipoint were designed for - performance at distance. Granted we’re being conservative, but it’s not the same as multiple streams in a single sector.


Is A5 any better than a standard 802.11 AP?
#14

And there you have it, a logical reason why it works, but is not a good idea at longer distances. Wait for the B5c firmware.


#15

So what kind of distances can we expect from A5c and B5c? I do not really like idea of a SM/SU with connectors. Would much rather have an integrated one with higher gain antenna, and cost effective.


#16

Having an OMNI or even Sectors in a populated area is ridiculous, clients can get FibreOP or high speed Cable.
There is absolutely no value, no play, no opportunity to deploy this type of wireless service in a dense population base. When Cable & Fibre ends outside of city/town limits (minimum 6km+) then yes there is an opportunity and clients will signup. I agree with @Matthew2 that an integrated 25dBi or higher dish is the solution for the homes and businesses to install with.


#17

A5c and B5c we have clients asking for 10-20 miles in many cases, we’re not limiting the distances so it’s all down to the size antenna to the connected radio. I certainly understand the desire for integrated solutions, we like them too, we just couldn’t get it all done at once so flexibility was key to start with.

Can you give me a sense of your preferred antenna gains and the weighting of how often you expect to use them (25, 27, 30, 34, etc,) and preferred antenna styles (reflector/horn, parabolic, grids, etc.)?


#18

@Jaime all of our dishes are parabolic, customers really dislike yagi or grid looking units. We use 25 dBi parabolic dishes, some with radome due to close proximity of the Atlantic ocean and harsh winds. Our 15km+ clients we use 27 and 20km+ 30’s.


#19

In addition to the current C5, it would be great to have a “medium gain” (about 25-27db) for our 2-8 mile customers and “high gain” (30-31db) for our 8-20 mile customers. Obviously the C5c coupled with an appropriately sized antenna accomplishes this, but nothing beats an integrated client for simplicity of installation and relative immunity to future coax issues (costly truck rolls).


#20

Understood, we’re keeping this in mind with our next gen client roadmap as we look at clients for the 8x8 MIMO solutions. Obviously we focused on solutions to cover the whole range, we’ll begin to optimize that approach for simplicity of installation and maintainability at the different distances.