Is it possible to determine the distance between 2 devices (using phase shift, time of flight or any other method) without line of sight?

In the affirmative case what's the precision and accuracy for distances below 50cm, and where could I get good information about that particular method?

  • \$\begingroup\$ What are you trying to go through? Depending what it is, attenuation frequencies may be different. Though, you may be able to do it indirectly through reflections, but you'll need a very complex setup. In other words, more detail is needed about your particular case. \$\endgroup\$ Jan 29, 2015 at 15:05
  • 4
    \$\begingroup\$ Sound propagates round corners, so theoretically it's not impossible - but it will give you the propagation distance not the "straight line" distance. \$\endgroup\$
    – pjc50
    Jan 29, 2015 at 15:10
  • \$\begingroup\$ I'm trying to go through an human ankle, to measure the leg stride. I was looking for a simple and practical to use solution. \$\endgroup\$
    – Rui Lima
    Jan 29, 2015 at 15:17
  • \$\begingroup\$ I don't think ultrasound is an appropriate use for that. I'd look at more round-about methods of getting the same information. For example, if you can figure out the distance made per stride, then you can calculate the stride with information like the users leg height/length. \$\endgroup\$ Jan 29, 2015 at 15:35
  • 1
    \$\begingroup\$ I believe you can use the gyroscope to tell when the leg is in a stride position and use the acceleration impulse time to get the time it is in that position. I believe you can derive distance from the accelerometer via a method called the double integration method... normally the error associated is too large, but you're only interested in distances between each stride. Technically, with just the impulse time and the distance calculation, you should be able to get stride length. \$\endgroup\$ Jan 29, 2015 at 16:09

1 Answer 1


If both devices could both send and receive... then they could talk to figure it out.

One would send a reference pulse, and the other would respond as quickly as it could. As long as your timing on processing and sending was fairly predictable (low jitter), the first device could take the total time from sending the reference to receiving the response, subtract a constant processing delay that would be very easy to experimentally measure, and arrive at the trip time.

This would not give a direct distance, but the distance going around any obstacles... unless the obstacles were particularly transparent to ultrasound. Even then, the difference in propogation speed through different mediums would mess with your results. Practically, it would most likely nearly always result in the through-air trip time around obstacles.

  • \$\begingroup\$ And worse, indirect paths almost always involve multipath effects. \$\endgroup\$ Jan 29, 2015 at 17:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.