I'm pretty okay w/ networking, but a noob at tweaking a point-to-point radio installation.
Equipment:
2x Routerboard 411r controllers each w/ a Mikrotik 900MHz radio @ 20ft
2x 3Com 4-port 10/100 switches
2x 12v 3A PoE injectors
Cable length from injector to radio is 30'. Both panels are 1/4 (.25) mi apart shooting through moderately thick groves of pine trees. One panel is 20ft up a 95' radio tower. It can be raised to 80'. Other panel (my house) the panel is mounted on 10' of mast next to the upstairs-bedroom window. Using WDS bridge (tried EoIP tunnel but they wouldn't associate w/ each other).
I used GPS to aim them at each other. Plot 1 in front of my panel, plot 2 in front of antenna on tower. They are dead-on aim-wise and at the exact same height.
SNR is -80, Overall Tx CCQ varies between 95% and 100%. Noise floor is -87.
I feel my SNR could be better, and the noise floor could be lower since this is a relatively remote-location install on top of a small mountain. I have neighbors, but were spaced pretty far apart so I don't know whats up w/ the noise floor being as high as it is. Is that a normal value for the noise floor?
Also, I looked a little and probably could have searched more...but I don't understand what an "overall tx ccq" is.
I'm extending a wifi-broadband link from the antenna on my neighbors tower (my antenna, his tower) to my house. Its a 3mbit plan, and I get all 3mbit (sometimes a little more) and an average ping from Maine->NYC of 20ms. When I do a continuous ping to google.com it stays between 15ms and 25ms...but has occasional fits of 150ms - 300ms. The climax of the ping fits usually consists of 4 or 5 failed pings, and then back to normal. I'm wondering if a better SNR would help with this.
Short version: I'm wondering if my noise floor should be better, wondering what "overall tx ccq" means, and wondering if a -80db SNR could be the cause of occasional fits of network lag and unreturned pings.
Thank you!
PS I bolded what I suspect are negative factors causing the bad SNR values.
EDIT: I think its also worth noting the TX-rate hops around quite a bit. It usually hits all of the values (1, 2, 5.5, 11) and not in order. It doesn't build-up to 11 and go back to 1...it just hops around. I've tried turning the panels slightly to find the best average SNR, but they always end up back where they started. I think its the trees.