A CPE is designed for this use case. It has a directional antenna which will get a better signal on a point to point link. Then it would pass data to another router which you already have.
Presently my favorite CPE for shorter range 5 GHz links is the Ubiquiti Nanostation AC loco.
5 GHz is much faster and less prone to interference than 2.4, but of course the source end needs to have it.
The 1900 and 3200 are not recommended for intense WiFi usage because the wifi chips they have do not have a good open source driver.
Thanks, I can't see such a small device working in this application, I gather one needs a pair of them but I do not have control of the device or building sending the signal, it does have AC and N.
I have tried a plethora of routers and the ones that seem to work best are ones with external antennas.
Thanks for tip on Linksys 1900 and 3200, what device would you say does have good open source driver?
Is it there anything I can do in OpenWRT to improve there performance of such a driver?
In front of it, even a low-end CPE's signal will far outperform any desktop router. Desktop routers radiate signal in all directions thus wasting most of it for a point to point link.
With the stock firmware it will only connect to another AirMax device. Running OpenWrt it can connect to any 5 GHz wifi AP where you have the password.
That may well be the case but as I have explained it is not an option no matter how good it is because I am picking up a public hotspot, I do not have access to their building, their network and have no clue if they have Airmax.
What I NEED is a the best router for repeating, not a CPE device.
Right now I can pick up the signal and most of the time it is 95/65
Knowing the limits I have would you say the Netgea 7800 is any good or are it's wifi drivers no good too?
My primary goal is to improve the latency, but also to stop losing the signal intermittently
I guess another option is a better antenna, possibly outdoor?
I have been re-reading your replies in cold light of day and just noticed a few things that may not have sunk in last night.
"Running OpenWrt it can connect to any 5 GHz wifi AP where you have the password."
"5 GHz is much faster and less proIn front of it, even a low-end CPE's signal will far outperform any desktop router."
"Desktop routers radiate signal in all directions thus wasting most of it for a point to point link.ne to interference than 2.4, but of course the source end needs to have it."
The devices I am connecting to do have 5Ghz, I have the password because they are free to access with a login page.
However, they themselves are Desktop Routers
So my question now is whether using the CPE device you recommend will yield better results considering the above than me putting in say a Netgear 7800 or TP-Link of some sort?
What concerns me is there is no visible antenna and I am not sure how to point this in a direction
The TP-Link CPE devices are another option and they will connect as clients of a standard AP out of the box according to their marketing material.
As for the answer to
Yes. CPE devices are precisely designed for exactly what you're trying to do. The antennas are covered by plastic, but they're in there and they have a directionality. The instruction manual for the device should explain how to aim them.
I do not actually know where the router I am connecting to is, I position the router to where I have seen the strongest list of Wifi networks and then choose what seems to be best one, I actually have a second one configured but disabled which I use if the primary has an extended outage.
What I like about this CPE device is it goes outside so should be better.
What I need is a rotator that turns the unit to the optimum direction for a network (by mac address).
What are the client's dBm and MCS numbers with the setup you have now? That would decide if radio upgrade is really likely to help.
The antenna inside the Nanostation beams signal out perpendicular to the front side, the side with the U logo on it. It isn't extremely directional, you can be 20 degrees or so to either side and still have good results. They can usually be deployed inside a window looking out-- regular glass doesn't affect the signal but metal coated glass found in commercial buildings can severely degrade it.
I wouldn't expect a public network to have good or consistent latency, unless they've set up SQM really well (or at all) which is unlikely. That is out of your control.
Received dBm in the -60s is good. The "noise" number isn't very important.
The MCS is the modulation scheme (raw bit rate) negotiated based on the fastest rate that signal conditions permit. This is a strong 2x2 HT40 signal.
Indeed 5GHz is faster and less prone to interference, but the downside is the range tends to be shorter compared with 2.4GHz, evenmore when there are brick walls in-between. A dual-band equipment should choose the working mode according to the measured conditions, (although that is not allways the case and some times you can fine tune that)
On the other hand there are several routers (full size or repeater size) you can use as client. Depending on the distance and location, you can ponder for ones with external (and optionally removable) antennae for better suit your needs. The bright side of the Linksys E3200 is its internnal amplifiers, but YMMV.