I didn't really know where to put this, I am trying something relatively advanced, I am trying to create a localized gps-like system that is extremely accurate within a range of 1-3km. My plan is to send a some type of signal and start a clock as I send it, when a receiver somewhere in the field gets this signal it will immediately send a signal back. The original transmitter will then subtract the latency from the original receiver from the time between when it sent the signal and the received it, then it will divide by the speed of the signal and divide by two. This will measure distance and a system of three of these will get a 2-D location. I need one of two things, or both:
I need an extremely accurate RTC to measure the distance accurately, if I use a RTC with a 3.5 ppm, and a signal the speed of light, it will know distance with a range of 1km.
I need a signal type that is both long range, and as slow as possible. I don't know if something like this exists, It was hard to research as it kept coming up with articles about things faster than the speed of light. One note is that the signal doesn't need to transmit data, it just needs to be safe, detectable, long-range, and preferably slow.
Any other ideas on how I can accomplish my task? Thank you!
I need an extremely accurate RTC to measure the distance accurately, if I use a RTC with a 3.5 ppm, and a signal the speed of light, it will know distance with a range of 1km.
So you need an RTC that can measure time at a resolution of 1km, which a radio wave takes around 3.3uS to travel. Good luck with that.
I need a signal type that is both long range, and as slow as possible.
I think you will need to re-invent physics for that one. Slowing radio waves down would imply you could have them moving at walking pace.
Forget trying to measure time of flight of radio waves by some cunning wheeze of an idea, its extremly difficult to do electronically.
The Semtech SX1280 LoRa modules, around £5, do have time of flight measuring capability, Semtech call it ranging. It works OK in large open spaces at distances of 100m+ and with good line of sight it should be good for maybe 80km and relativly accurate, I have only tested it to 40km . The ranging is not so good where there are multiple reflections etc.
Differential GPS (DGPS) is the normal way to improve GPS accuracy.
Your method wont work because of the extremely tight timing requirements and too many unknowns, the main one being the time it takes for a receiver to actually receive whats being sent, decode it , and then retransmit it.
Unless you have some insanely good test gear, this is not easy to measure.
For the SX1280 I dont know the exact timing of how long it takes to send a packet and for the receiver to process it and turn around the reply. It must be significant compared to the actual travel time of the radio wave. For the distance measurement to be accurate this processing time has to be known and consistent.
An explanation of Semtechs approach for the SX1280 is mentioned here;
GPS has to jump through hoops to get accurate arrival time differencing, it certainly cannot rely on
the signal arrival time, since this is extremely poorly defined for a narrow band signal(*). Phase information
and correlation techniques are needed. This is all done for you in every little GPS module, and doing it
yourself would be a major project for someone used to RF circuitry.
A signal with a carrier of 100MHz has 3m wavelength, but when carrying a bandlimited signal of 10kHz,
the carrier amplitude risetime would be around 16µs, which is equivalent to 5km, or 1600 wavelengths,
how can you choose which carrier cycle is the correct one to time from? The naive approach just doesn't work...
If you don't band-limit, you interfere with every other band on the radio spectrum, which won't make you friends.