SODAR - Sound Detection and Ranging

SODAR is, in short, a passive SONAR array. Like SONAR, SODAR is capable of determining the range and direction of a target based on audio radiation from it.

The final implentation would consist of an arrry of carefully-placed microphones, and a central control unit (a DSP chip or personal computer) to process the inputs from the microphones.

The control unit conducts real-time signal processing, and determines the range and azimuth of a point audio source. Once these numbers are computed, the controller renders a display - much like a radar scope - and places a dot on it at the location of the source.

Overview:

Figure 1 provides a block diagram of final SODAR system. Four microphones are precisely arranged in a formation commonly called a Y-Array. The number of microphones, array geometry, and distance between microphones have a strong impact on the performance of the SODAR system. Specifically, angular resolution is greatly improved with four microphones in a Y-Array, and the distance between microphones affects range and frequency performance. This will be explained more thoroughly in the Details section.

Figure 1: Block Diagram

The four input streams are fed into a control unit, which in our case was a computer running MATLAB. Each stream is compared to the stream from the central microphone by a function performing fast convolution. This essentially takes the shifted inner products of the two signals, and from the result, we can easily determine the phase shift between the two streams. When done with all three pairs of microphones, we find the time delays between when the signal was received by the central microphone and each of the peripheral microphones.