Broadband MIMO sonar system: a theoretical and experimental approach

MIMO systems have raised a lot of interest in the recent years. The radar community pointed out the multiple advantages of MIMO systems such as diversity gain for target detection, angle of arrival and Doppler estimation. Coherent processing also allows super-resolution for target localisation. We explore in this paper broadband MIMO sonar systems. In the current literature, the channel matrix has been computed using point scatterer models. The limitation of such models for sonar (and broadband sonar in particular) is explained and a model based on the target form function is proposed. We will show that this deterministic MIMO model can naturally be extended to a statistical model. It has been shown that broadband sonars offer in situ great capability for target classification. Using widely separated transducers, it has been shown that channel matrices are decorrelated from one another, which means that the views of the potential target are independent. MIMO systems improve the process of classification thanks to these multiviews.
We present experiments done in our tank (Width x Length x Depth: 3 x 4 x 2 m) with a broadband MIMO system (2 transmitters, and 4 receivers). The transmitters cover the frequency band of 30 kHz to 150 kHz. We demonstrate experimentally the advantage of Time Reversal for MIMO system by focusing the energy on the target independently of the medium. We propose a pseudo Time Reversal technique which focuses the energy directly back to the receivers increasing the SNR by a factor of N where N is the number of transmitters.