Student/intern wondering?

Is their a way to accurately (3inch max tolerance) measure distance between multiple devices in a single story building. Using only RF signals? Imagine a square building with any random amount of wall/rooms etc. I have a master device located in known area of the building and any number of "slave devices" in unknown areas of the buiding, but slave devices are capable of "talking" to the master device or any slave device per request of the master. Now take the building and divide it into four even parts so that their is section 1, section 2, section 3, and section 4. The master device is located in section 1 , so I guess this is my question: Is their a way that I can calculate the distance of the slave devices from the master device to the degree that I could know what section the slave devices are in. Is it possible with RF signals, through walls and obstacles and such?

In a word, no. It sort-of works in an open field.

Thanks. I will search for other methods.

In theory it would be possible if you had very stable RF transponders on each device and a means of measuring response times with an accuracy of ~ 1/4 nanosec. But in practice that's not likely.

alec_t said:

In theory it would be possible if you had very stable RF transponders on each device and a means of measuring response times with an accuracy of ~ 1/4 nanosec. But in practice that's not likely.

Click to expand...

In theory a single measurement will only provide you with distance information, which would only allow you to discriminate position as being somewhere on an arc.

Unless the base unit has a directional antenna array, in which case you could narrow it down to a chord. This is neglecting all the junk in the way e.g. walls, cupboards.

If there's four base units (one in each room), then it should be easier to determine which room the mobile units are in, without having to resort to space-age transponders; just a simple signal strength comparison will provide enough info. Accounting for the building structure and contents could be done by learning signal strength patterns for each room using a learning algorithm, e.g. SVM, ANN, etc.