Rural Cellular Internet - Opinons

[sfx2000]

sfx2000,
I'm not that smart on MIMO. I've tried to do some reading, but it is confusing to say the least. I had thought about getting another Yagi. From what I've gained from the internet, I'd have to set up both the Yagi's at opposing 45 degree angles. I'd probably be more open to two omni's though since that just seems to be an easier set up. I don't really see that much difference in RSSI or SNR switching between the Yagi and the Omni.

Either another Yagi or another Omni - if you get the Yagi, point them both in the same direction... LTE is 2T2R at a minimum, and it assumes the radios are balanced - that's why I'm recommending that the antennas are matched (with same length cables) - The RSSI/SNR is at the baseband, which is combining the links, that why you don't see much difference - the Antennas are in the analog domain, and RSSI, which can be measured thru a diagnostic tool (QXDM is one of the more common ones at around $30K a seat)...

Trust me, you'll get better performance with matched antennas...

 

[sfx2000]

Another thought...since the T1114 uses signal diversity, wouldn't I be better off with my current set up anyhow? I would think that having one Yagi and one Omni my signal would be quite diverse.

You're driving an unbalanced link, with the Yagi probably around 20 dB gain, and the omni dipole around 3 to 5dB - that's not the kind of diversity you want

Most 4G technologies use both Receive Diversity and Spatial Diversity, along with MIMO encoding - it's actually two distinct receivers, and they're calibrated in such a way to work together in a balanced fashion - this goes for both uplink and downlink - so again, see my advice above.

 

[stevech]

I had a grandfathered Verizon data plan.
But they summarily ended it.
No debate.

 

[doco2279]

Quick question....if I get another Yagi, is there any minimum distance that they have to be from each other? I was actually thinking of creating a 90 degree bracket to hold the pair, one horizontal the other vertical. I think I've seen similar set ups. I was just going to go with the Wilson enclosed Yagis. Any thoughts?

 

[azazel1024]

Generally you want/need a good full wave length's distance between the two, which isn't much. A foot or two should be more than enough (from the edges of the antenna, not from the centers).

 

[stevech]

I worked on this optimal antenna separation problem professionally for some time.
There isn't a universally optimal separation other than > 1/4 wavelength. The optimal separation depends so much on the use case, e.g., eaves-height with a link to a single distant access device with near line of sight.

I've never understood how spatial diversity is supposed to bring any significant benefit, to client devices with closely spaced antennas. These are more like MISO for spatial diversity independent of frequency diversity, if any.

 

[sfx2000]

Quick question....if I get another Yagi, is there any minimum distance that they have to be from each other? I was actually thinking of creating a 90 degree bracket to hold the pair, one horizontal the other vertical. I think I've seen similar set ups. I was just going to go with the Wilson enclosed Yagis. Any thoughts?

Depends on what frequencies they use in your area - rural is pretty much 700MHz, starting to see more carriers these days in the AWS bands in Suburbs and Urban areas... 1/2 wave at 700MHz is roughly 8 inches, AWS band is 1.7GHz for uplink, and 2.1 GHz for downlink... full wave for the downlink is roughly 6 inches...

Keep the Yagi's polarized the same direction.

 

[stevech]

AWS is sometimes called LTE+.
AT&T mostly.
But AWS' GHz frequencies bring a big disadvantage in RF propagation and building penetration versus 700MHz which is most of Verizon's LTE.

 

[azazel1024]

I worked on this optimal antenna separation problem professionally for some time.
There isn't a universally optimal separation other than > 1/4 wavelength. The optimal separation depends so much on the use case, e.g., eaves-height with a link to a single distant access device with near line of sight.

I've never understood how spatial diversity is supposed to bring any significant benefit, to client devices with closely spaced antennas. These are more like MISO for spatial diversity independent of frequency diversity, if any.

True, but 1/4 wave length isn't all that much. 3.125cm, or about 1.2 inches. In a cell phone that might be hard to achieve for a pair of antennas (though even then, if on opposite sides of the PCB/body not too hard), but on a basestation, pretty easy. On a tablet or laptop also pretty easy.

 

[stevech]

True, but 1/4 wave length isn't all that much. 3.125cm, or about 1.2 inches. In a cell phone that might be hard to achieve for a pair of antennas (though even then, if on opposite sides of the PCB/body not too hard), but on a basestation, pretty easy. On a tablet or laptop also pretty easy.

Yes, we and AT&T and others made lots of multipath delay spread measurements in several markets. Spent a LOT of money on these measurement because the conclusions had such a big impact on CAPEX (capital expenditures) by the carriers - and their desire for self-installed customer premises equipment. (A truck roll was said to be $500 all said). The work culminated to the IEEE standard used in fixed wireless.

I recall that eaves-height antennas if very directional (pointed at the access node), really reduced destructive multipath from angles way off boresight. We found that wet leaves (seasonal) were the biggest culprit at 2.5-2.6GHz (old MMDS band, now repurposed). Icy asphalt and concrete too. Change the antenna height or beamwidth and everything changed a lot. I recall something like 3/4 or 1-1/2 lambda as good for eaves height. Indoors, multipath is totally different - like the RMS delay spread is about 50nSec in typical office buildings.

So as they say, it depends. MIMO is no miracle. And spatial diversity at a client device without good spacing, isn't MIMO, it's MISO.

 

[azazel1024]

Hmmm, cool to know. I guess one of the important things is maybe that there is no such thing as too much space between antennas for MIMO (within general reason, I assume 1,000ft between antennas is maybe NOT a good thing) so long as they have equal path length from antenna to receiver? Is that generally the case?

 

[sfx2000]

AWS is sometimes called LTE+.
AT&T mostly.
But AWS' GHz frequencies bring a big disadvantage in RF propagation and building penetration versus 700MHz which is most of Verizon's LTE.

Nice thing about AWS band is one can pretty much reuse the frequency plan and spacing that is used for PCS at 1.9GHz, collocating the AWS radios and sharing the antennas - and there's more spectrum available in the AWS band - for either wider channels or more RF carriers (capacity) - and it doesn't come with the baggage that the 700MHz bands have...

Verizon, ATT and T-Mobile have been aggressively rolling out LTE in the AWS bands - it's one thing that Sprint doesn't have (but Sprint isn't hurting for spectrum in any event).