Johnson M., Carneal J., Gillett P.

There are applications where, for practical reasons, microphone arrays must be integrated into/onto a structure. As an example there are efforts to instrument soldiers with sensors and vehicles with sensors in order to detect and localize noise events in the environment. In both of these cases the microphones in the array are unlikely to be equally spaced and in both cases the sound they measure will be diffracted around their support platform. It is postulated that the diffracting support platform is actually an advantage and in order to demonstrate this, the case of a microphone array mounted on a diffracting cylinder is used as an example. The cylinder was chosen as it has a relatively simple analytical solution for calculating the diffraction. Specifically this paper presents a theoretical comparison of the performance of two geometrically identical cylindrical microphone arrays with and without the diffracting cylinder present. An inverse method is used to compensate for the diffraction and allows the system to be conditioned for noise. Conditioning is particularly important at low frequencies where the array becomes small as compared to a wavelength. The performance of the array is evaluated using its white noise gain, its point spread function and its singular values. It is shown that the diffracting array performs better than its non-diffracting counterpart at low frequencies and equivalently at high frequency. Experimental data measured in an anechoic chamber using a 12 channel diffracting and non-diffracting array are used to validate the theory.