info on using rssi data to do antenna tracking?

rwgast_logicdesign

ok so ive been looking into using rf as a means of tracking as suggested by another member. i guess the plan is to use rssi to determine the direction and strength of the signal. im not sure what ill be using for beacons probably nordic 2.4 based boards.

.
for right now i want to track my bots location using its hc05 bluetooth. the data sheet says you can run an at command to get rssi data. now i dont know what rssi data looks like or what kind of alogorythms are used to determine distance or direction..

i searched google for rssi based tracking and turned up a few rc sites and pdfs that were super vague. i was hopping someone could shed some light on what to do with the rssi to turn it into useful tracking data

Phil Pilgrim (PhiPi)

RSSI is just a number that indicates the relative strength of the incoming signal in no particular units. It says nothing about direction and, frankly, not enough about distance to provide an accurate range value. However, if you had a directional antenna and a way to steer it, you could determine the bearing to the signal source by rotating the antenna to maximize the RSSI value. By doing this with multiple sources, you could then triangulate a position.

-Phil

xanadu

Using RSSI as a beacon works great under certain conditions. You might be referencing my post from awhile back. You need a signal that is weak enough to measure a difference in line of sight, and the cheapest of the XBees work well. I've been meaning to update that project with the directional antenna.

http://youtu.be/QgN8kpS3DME

The best way to track something would be multiple methods, using just RSSI alone will be cumbersome

rwgast_logicdesign

well it looklike the boe dar is a very good start to what im tryimg to do. the problem is alit of the hardware used in it isnt being sold anymore. i was kinda hoping to learn from it.

Mark_T

For any accuracy you look for nulls (signal drop outs), not maximum signal, since these are much better defined (unless you use a _very_ high-gain
antenna like a long Yagi or parabolic dish. You need a carefully engineered antenna away from all other influences, perhaps a small
2-element dipole antenna array stuck up a pole with a stepper motor to rotate it. You'd need to calibrate the null-points of the antenna
and try to make sure the nulls are insensitive to changes in polarisation. And you'll need some way of identifying which null point is which... Bit tricky
I think, good luck.