LAUSR

It is well known that bearing estimation performance, using a moving line array of pressure sensors, can be enhanced by using a Kalman filter to jointly estimate the bearing and source frequency. This is because the bearing dependence of the Doppler can be exploited when the source frequency is known. However, it is also known, based on an observability analysis, that it cannot be done using a single moving pressure sensor. Here, it is shown that if a single moving pressure-vector sensor is used, both the bearing and roll angle can be estimated. If only the vector sensor is used, estimation of these angles can still be done, but the performance is poor. The addition of the pressure sensor allows the motion to be exploited, thus significantly enhancing the performance. This phenomenon is shown theoretically by using a Bayesian Cramer-Rao lower bound calculation. An example using simulated data is shown where the improvement is clearly demonstrated.