WelcomeTuesday, February 7th, 2023
Distance from GPS Base to Radio Repeater
Hello all, my company is contemplating setting up a CORS-ish type continuously operating base. We desire to place the GPS receiver at Location A with an elevation of about 3800 and place a repeater radio at Location B about eight miles away at an elevation of 6600. A wise man (our GPS equipment rep) said this would not be a good setup, having the repeater radio so far from the base. I'm just looking for another opinion. Or several. I can't get any help from the manufacturer, they say to go through your local dealer, and I already have his very opinionated thoughts on the matter.
I know that changing atmospheric conditions are the enemy of accuracy with RTK GPS surveying, so the farther the rover is from the base the more potential there is for reduced accuracy. But I don't see why my GPS rover would care if it is talking to a base via a radio on some distant mountaintop.
Thanks in advance for any help.
I'm guessing the real problem is the communication between the base and the repeater. Radio waves flow like water and don't like to go uphill. Set up the base and then go to your repeater location and see if you get signal. If the elevations were the other way around you would have a better chance of success.
Radio paths are usually near symmetric-the loss going one way is similar to going the other way.
Radio waves don't care about uphill and downhill. What they need at the frequency bands involved here is clear line of sight or minimal obstacles to it. They can diffract going over a ridge that blocks line of sight, but that causes some loss and uneven coverage on the other side.
The radio is sending data, it's not necessary that the radio and receiver are seeing the same atmoshpere.
The problem might be the base broadcasting at only 2 or less watts. A repeater that far removed might have reliability issues connecting. But there is another issue, you are proposing setting up a permanent radio at a high point blasting a radio signal, not sure that's OK with the FCC. I'd be very careful doing it.
May want to consider running an NTRIP caster over IP (RTK Bridge or similar) if there are cellular signals available, to get those distances.
However, that's running single baseline GNSS, at (relatively) long(er) distances.
Just tell them you need two (2) bases with NTRIP caster capabilities, and put them at A and B.
Sounds like you're planning on setting up a single base solution RTK system.
Cool! As far as using a repeater, that will depend on the configuration for the base setup. We had purchased a new Alloy, and there was a variety of ways to use it.
The license to broadcast is fairly simple, and the vendor can help walk you through although it's good to do it yourself to understand the guidelines and requirements.
All in, if you're using it as a base, and then want to use repeaters, that's just a simple set up and just selectable depending on the hardware manufacturer.
Mi left above also has some good info too if you're configuring the network solution VRS also.
My experience with radio data modems has been tempered by the trial and error and even using real physics to get results.
if you're broadcasting from the bottom of a big basin, the signal bathes the ascending elevations in the spheroidal pattern of omnidirectional output. Think about the cardioid patter of a microphone and it's graphic diagram.
That being said, if you place it high above, the cancellation near to and beneath the source is also similar.
Beaming occurs at distances and other things like Bill mentioned like propagation and summing also occur downhill from peaks and ridges. One solid solution l found was to do the 'radio dance.'
I'd pack a 50' antenna cable and when I was far away and getting less signal, I would deploy the cable and walk away from my gravity station with the GPS, and walk into the signal with my now 100ft diameter mobile antenna dealio to close that point.
either way cool beans they want to give you some effective tools.
the GPS guy is wrong. 3800 to 6600 is a great set up.
low to high works best.
prove me wrong GPS sales guy.
"Sounds like you're planning on setting up a single base solution RTK system." Yes, that is the intent. I don't know any more about NTRIP than what I just googled in the last 10 minutes. It's a bit intimidating for me. We're in a rural area, with the nearest CORS on the other side of a mountain range some 35 miles distant. Setting up more than one base is out of reach for me budget-wise, and at the moment I don't have the technical savvy to do that anyway.
We own space on a radio tower at the 6600 elevation location. That tower is already bristling with gobs of radio antennas, but I have no reason to believe another couldn't be added. We have a company we work with that would get the repeater (and the base radio) appropriately licensed.
Thanks to all for your responses.
I don't know what frequencies the existing radio tower(s) are using or what the "bristling with gobs of radio antennas" are for, but there is a better than fair chance that all the high power RF stuff pumping out of that site may interfere with your RTK radio broadcast/reception and possibly even your GNSS reception.
This sounds like a creative idea/solution, but is likely wrought with problems you won't find until you have your system up and running. And those types of problems may not have solutions once you are up and running.
We have both a UHF antenna and a GNSS antenna on our tower. We used to broadcast both UHF and to a static IP as one crew had radio modem and other cellular. UHF was limited in range.
You do not say if you are in the US and if you have good cell coverage. You appear to be in terrain that might block out radio signals. Suggest you compare the cost of static ip and a sim card in your sensor or data collector vs what you propose.
Only good reason (I can think of) to use radios is if you do not have cell coverage unless I am missing something.
I have run the Base in St George Utah for a long time (2009'ish I think). We have a base above town and a repeater on West hill:
The distance from the repeater to the base is 17 miles. The repeater is at 7700' elev, the base is at 3000'. The only problems we have had are with lightning, power (at the repeater), users accidently setting their bases on the primary frequency and dimwits taking our antennas off of buildings and leaving them disconnected.
We have dedicated frequencies (well relatively dedicated) for 453.500, 453.400 and 453.250, we also run a second repeater at Corral Hills Park that covers southern Washington County. The terrain is very mountainous and all this is really needed.
Of course, all of these systems have internet access. Both by DIP and NTRIP. However, even in this heavily populated area there are plenty of places where there is no cell service.
The base covers most of town, the repeater covers Ivans, Santa Clara, most areas west of the old airport, north to Veyo and Central, AND clear over to Tocquerville.
I also have several other 'systems' around the state. One has a internet backhaul and it has had issues with needing to be restarted occasionally.
The real issue with repeaters is the number of bytes-per-second that RTCM3.3 MSM4 generates tracking GPS+GLO+BDS+GAL with all signals enabled and 37 SV's tracked. Your radios need to have a high enough baud rate to transmit each correction twice, once from the main radio, then a 10 ms delay and again from the repeater. It is getting to where 19,200 baud with FEC off is the 'only answer' for RTCM3.3 (CMRx still has some headroom at 9600 baud.)
Another reasonable solution is to use the 'Bandwidth Limit' setting, which I have verified as working in the latest Trimble firmware sets--with most brands of rovers.
Anyway, back on topic, the real problem will be getting a frequency which is clear of other users. If you can get a licensed and unique frequency, then you won't have any problems. And while we used to (GPS only) claim that the uncertainty of solutions was 1 PPM of baseline, current antidotal evidence is that with more SV's and more frequencies, acceptable accuracies can be obtained at substantially longer distances.
Took me a bit to find this picture. This is about 5% of the antennas on West Hill:
There is a big cell tower and multiple towers of microwaves, ambulance, police, private repeaters. Way more than 100 antennas and cables.
I also found this cool google earth image looking back to town:
It is really hard to get up there, one year the power supplied died and ToddJ and I hiked up the last 3 miles in 2 feet of snow. Not sure I would make it today.
The distance from the radio to the base should not cause a GNSS precision issue (other than a little latency). I would guess that the biggest problem you will have is getting the radio signal from the CORSs base radio to the repeater. If the repeater covers your entire working area then I would suggest a directional yagi antenna to shoot get from the base to the repeater.
As others have said, if you have good cellular service in your area you may want to think about going that route. You don't need to set up a network of CORS, single base solutions work fine for smaller areas, or even an NTRIP server if you don't plan on having a lot of simultaneous users. I think I can have 5 rovers attached to my GS16 via static IP GSM internet connection. As @hpalmer said, the only reason to go with a radio setup is lack of cellular coverage. Radios are finicky things.
@mark-silver Thank you, thank you, thank you. What you describe is pretty much exactly what I am attempting. Very helpful to know that someone has already plowed the snow. My location is rural/remote enough that I think/hope I can score an FCC frequency in the neighborhood of 450 MHz.
We do have fair cell coverage, not good. It'll leave me stranded in parts of the area I wish to cover.
There is an extra latency of the time to transmit the corrections to the Base, plus the inter-message time typically about 10 mSec. If a RTCM3.x message has 325 bytes and the radio is set to 9600 baud, then the throughput will be ~960 characters per second, assuming FEC is off.
The extra latency caused by transmitting the message twice is:
(325 (ch/message) / 960 (ch/sec)) + 0.01 sec = 0.349 sec/message
The latency of sending the signal an extra 35 miles (say 20 miles to the repeater, then 15 miles back to the rover) is:
35 miles / 186,282.4 (miles/second) = 0.00019 seconds
Which I think can safely be ignored.
Actually, the latency of a message that is provided by a VRS is substantially longer than a radio or repeater. There typically is > 60 ms of time from the RT27 stream leaving a CORS station to gets to the VRS server (if you were at the station, then you could ping the server to get the round trip time, and divide by 2). Then perhaps another 100 ms for the server to combine some stations together and build a virtual correction stream for your location, then another 60 to 100 ms to cast the corrections over the cellular backhaul back to your rover.
Because the rovers are typically not running in synchronous mode (one measurement per correction) and are actually modeling the corrections in time space based on a delayed correction stream, modern rovers can produce 5 to 50 measurements per second, with very small phase delay, even though the correction stream might have a latency of 4 (or more) seconds.
@mark-silver ahh, I remember when we did some survey work NW of St George in the early 1980's and a sleepy little town with cold beer and edible food. I remember a hatch expedition trip below the fg dam on the green river in the mid 70's. Fast foward - more houses and more expeditions. And more antennas