We are looking at using a BaseBox 2 (AR9342 chipset) on a mobile equipment.
(I understand that BaseBox and mobile seem to be a diametrically opposed application)
Does someone know how the Atheros DSP expect that antennae can be used to extend coverage, rather than bandwidth ?
For discussion purposes,
Think of 2 omnidirectional antennas back to back, but separated by a grounded wall of steel that would see DIFFERENT access points.
The radio will be configured as a Wifi CLIENT
The radio will be moving at 50 km/h
My understanding of MIMO 2x2 is that it’s aimed at raising bandwidth throughput through Beamforming or Multiplexing streams on the same channel.
QUESTION:
If we use the AR9342 dual antenna radio with the antenna configured where they will see different access points to provide better coverage, will this provide
A) better COVERAGE performance to link to APs?
B) deteriorate performance because the DSP is confused and trying to use the 2 antennae together ?
MIMO improves both range and speed simultaneously, but it uses both chains reading the same signal, more or less, to do the job.
Physical separation is required to a degree, but blocking the antennas from receiving the same signals is going to break the signal reconstruction.
You’re not going to be able to connect to two access points with the car as a station, though if you set the car as the access point and two stations out in the field you may be able to pull it off. The problem is creating a network loop, either using spanning tree, routing, any of a dozen different hand off methods to swap between two physical connections is going to introduce a bit of lag.
It is an interesting problem that I’d actually really like to work on.
We’re not trying to block the signal from both antenna intentionally, we’d be mounting it on a huge piece of open pit mining equipment and we’d place one antenna on each side of the truck. The challenge is that the access points are on mobile trailers in the middle bottom of the pit and the equipment is working in nooks at the bottom edge of the pit. So if when we have a single antenna radio mounted where it is protected from falling rocks (when the trucks get loaded), then we have 180 to at best a 220 degree of RF coverage. So depending if the truck is facing right or left of the access point, then equipment will have good LOS to the access point or it will completely block the LOS to the access point.
So we’ve had good success with the single antenna Metals, but now the question is can we get better success with the dual basebox…
Now I know the “old” diversity antenna array DSP code was designed for better coverage, and I know that the new MIMO DSP code is more aiming at higher bandwidth. I was wondering if there was already wisdom out there that knew off the bat that using MIMO technology in this way would be beneficial or detrimental.
I think the only way we’ll be able to know is to actually run half a fleet with the Metal and half with the BaseBox and log the connectivity & throughput and compare. I know that if we’re using the radio in a way the DSP code is not designed to be used, we can end up with a lot of bad side effects. Another method would be to mount 2x single antenna radios but the routing then becomes more challenging + we’re effectively doubling the media access (twice more radios chewing the band)
Bigger, more power and more antenna will not always deliver a Better performance.
Maybe my question should be:
Can I configure the BaseBox 2x2 MIMO radio chains into a DIVERSITY 1x1 SISO but using both radio chains ?
(i.e. NOT creating a SISO by disabling a chain)
To make it a little more reliable you can go in to the data rates, choose your config (can’t remember the phrasing at the moment) then go the HT tabs and disable all speeds above 7. It won’t try anything above single chain speeds.
Normally I’m better with the terminology, but it’s 3am and I’m dead tired and can’t sleep.
I understand your situation a bit better and I’d do it the same way. In the first post it sounded like you were intentionally blocking the two antennas for separation which is detrimental for 802.11n and above, probably below as well.
Yeah, I know exactly what you’re talking about… I suffer from that as well from time to time when I can’t shutdown my mind.
I think we’d have to run an “experiment” to measure the difference in performances. We could measure coverage and run the fleet with the 2 chains turned on for a day, then try it with a single chain and see the performance. There are so many variables (radio firmware, radio configuration, RF reflections, RF multipath, speed of vehicle → Rate of change of S/N to AP, antenna location)