Relative positioning system ?

[amith.srivatsa]

Hello everyone,

I am working on building a relative position system for a robot using 433 MHz transcivers.

To measure the position of the robot,(................principle similar to GPS............) I intend to use an active device(433 Mhz transceiver) on the robot and 3 repeaters(433 Mhz transceivers) placed on for example trees, now, to measure the distance, I plan to use "data delay time" principle.

So, I should know "at what point of time" I transmitted my signal from the active device" & then "at what point of time" I received back my transmitted signal after modification by the repeaters. Using this data, I can find the relative position.

Do you know how to do this ? Also which transceiver modules have this feature ?

Many thanks in advance
Amit

 

[kender]

Get a good text on GPS and follow it. BTW, you'll need at least 4 transmitters (just like you need 4 GPS satellites), because you will need to solve for time on the roving unit too. In some sense your setup is simpler than GPS: no Doppler and you know the PRN codes of your transmitters in advance.

 

[FvM]

GPS is using a sophisticated receiver design to achieve high time resolution with relative small signal bandwidth. I don't think, that any usual UHF transceiver is particularly suited for a similar application. Depends of course on the intended distance accuracy, that wasn't mentioned yet.

The most simple method is possibly a private DGPS setup: A pair of a fixed and a mobile GPS receiver and a radio link for the correction signal. Unfortunately, it only works in the open country.

 

[amith.srivatsa]

Yes, I forgot to mention an important parameter, the accuracy required is 1cm when the rover is not moving and about 300cms when the rover is moving.

Also, the LOS of the transceivers is for 500m.

I am a beginner, have no previous RF experience...For now, I am selecting a good 433MHz transceiver IC & a good CPLD/FPGA.

The plan is to implement the algorithm to calculate the distances between the "active device"(on the rover) and the repeaters(on trees) using CPLD/FPGA. Then, to make a PCB design with the Trancseiver & CPLD/FPGA. The calulated distances are then to be given to a microcontroller which calculates the relative position of the rover based on which it can navigate autonomously.

Will this work ? Suggestions wuold hell me a lot.

 

[FvM]

1 cm sounds illusional.

 

[kender]

1 cm sounds illusional.

1cm was done at 50m with a well-controlled setup and in "vanilla" environment. The application was to determine positions of horse raiders on the race track (hippodrome). It can be done at 500m provided good environment.

 

[FvM]

I don't doubt, that it's possible with respective techniques. But not with a usual 433 MHz transceiver.

P.S.: Do you know, what kind of RF technology (carrier frequency, modulation, receiver design, antenna) was used to achieve the said 1 cm accuracy?

 

[kender]

Do you know, what kind of RF technology (carrier frequency, modulation, receiver design, antenna) was used to achieve the said 1 cm accuracy?

It was a reverse GPS system: the rovers were transmitting and "satellites" were receiving. Frequency: 1.5GHz. Same frequency as GPS, however, the transmitter was orders and orders of magnitude closer to the receiver, so the interference from celestial GPS signal was negligible. Antenna on the rover was a patch, antenna on the receiver was either patch or helix. Modulation was BPSK.