The circuit described is a time of flight sonar, rather than the doppler sonar abhishek describes.

If I understand Abhishek correctly, he is not referring to Doppler. I will try and restate his idea as I understand it:

Emit a signal of (constantly) increasing frequency. Take the received signal and mix it with the transmitted signal. The time delayed received signal will be a lower frequency than the transmitting signal, so the mixed signal will have a beat frequency that can be found and then used to solve for the delay.

This is interesting because it would give you a continuous measure of distance, instead of the periodic measurements given by timing incident waves.

I wouldn't have a clue on implementing it.. Sounds like a problem for someone with some radio experience.

- Lee