Distance radio sensors (motion tracking)

Hi, I'm new to Arduino and electronics at all. I want to make a simple project, but I have some questions. It will be very good if someone will make some things clear.

Are there sensors that calculate distance between emitter and sensors themselves by radio waves like bluetooth with high precision (up to ~ 2 cm, small emitter size will be great)? Link to a shop or just a title would be great. IK sensors are not my option because they cannot be used from behind the obstacles. The idea is to calculate the position of an object with emitter in a room by several sensors that will receive radio signal from emitter and transfer it to a computer program.

Does it mean that I have to add inertial sensors combined with radio emitters if I want to get the rotation direction of an object with the sensors? Can be inertial sensors be used alone to calculate position (in a room)? What do I need in that case bluetooth or wi-fi module to transfer data to a "station"?

Someone told me several minutes ago that it is nearly possible but very-very expensive. So the new question that I have is do only inertial sensors help me to get position of a moving object in a room by calculating dx, dy, dz each time? How can I get relative precise initial position of an object?

How can I get relative precise initial position of an object?

Theodolite.

Radio propagation doesn't work anything like light from a bulb at this scale - its highly directional affected by all conducting metal objects that cause wave reinforcement, cancellation, reflections (multi-path reception). Particular sizes of conductor will resonate like aerials leading to strong scattering, and sources of external interference will cause issues too (accurate measurement of signal strength is not trivial).

You can rely on radio propagation to be line-of-sight in free space, but you still need to know the transmitter's polarization in order to be able to calculate distance based on signal strength, so a transmitter that's free to move/rotate in 3D still can't be measured this way - you only get order-of-magnitude information from signal strength.

What you can do is measure time-of-flight differences like GPS system does, should you happen to have the precise clocking setup between multiple receivers. This is more practical with ultrasound where the time delays are more practical, but multi-path reception is an issue with time-of-flight too - sophisticated signal processing is often needed to get good results I believe.

Setups with IR cameras fair much better, although its more complex the triangulation of an IR LED indoors works well - is used to monitor quadcopter swarms for instance, and has been taken to new levels with the Kinect.

The Score is: 1 - 0 for IR Pattern Recognition or Machine Vision. Personally there is enough to keep an older Pentium busy in either Area of IR research. Macinne vision might work BUT I think It's Going to Take More Than 1 Arduino... Even a Mega...

Doc

Thanks for your answers. To be more specific, I want to (just to try) to make simple motion tracking. I know that such equipment is based on inertial trackers. What can you advise to start with?
I think I can start with only one sensor but I have problem with fast computing of initial position.