摘要: |
This thesis considers the use of acoustic communications in reducing position uncertainty for collaborating autonomous underwater vehicles. The foundation of the work relies on statistical techniques for accurate navigation without access to GPS, known as Simultaneous Localization and Mapping SLAM. Multiple AUVs permit increased coverage, system redundancy and reduced mission times. Collaboration through acoustic communications can minimize navigational uncertainty by permitting the group to benefit from locally discovered information. However, the propagation of acoustic communications can be used to counter detect the system during naval operations. The thesis gives explicit consideration to tactical security in acoustic communications for a multi-AUV SLAM system. It provides initial techniques and analysis for minimizing communications between AUVs. The reduction is accomplished through a statistical method that allows for the estimation of the updated covariance matrices. Normally, SLAM techniques use expropioceptive sonar and cameras sensors and computer vision algorithms for the detection and tracking of navigational references. We propose a novel use of the acoustic modem as another sensor. It leverages the physical characteristics of underwater acoustic transmissions and the information transmitted in the signal to provide an additional measurement. We believe this is the first emphasis on minimizing communications within a multi-vehicle SLAM approach. |