Speed of sound and air temperature estimation using the TDOA-based localization framework

Spatially distributed acoustic sensors find increasingly new applications in speech-based human-machine interfaces. One well researched topic is the localisation of an emitting source from Time-Difference-Of-Arrival (TDOA) measurements. This manuscript shows how to exploit the source localization framework and its state of the art techniques to accurately estimate the actual sound speed and the air temperature by measuring TDOAs of an unknown located acoustic source. Simulations and experiments show the validity of the proposed method.