Development of a new positioning system for underwater robot based on sensor network

We are developing a new positioning system for an underwater vehicle based on the sound power decreasing with the distance transmitted and sensor network technology. Conventional positioning systems use the difference between the time sound occurs at the station and the time it is received at a vehicle. However, this is very expensive because of the high operational costs and the fact that the measuring field is not very wide. We therefore need a new positioning system with low cost, easy operation, easy maintenance, and high accuracy. Our system is expected to satisfy such requirements. At the first stage, we confirmed the principle with theoretical and experimental methods. In the research, we considered only the direct waves in the received data. These showed good precision for short distances, but also showed that it was easily affected by reflected waves, and the positioning area was limited to within only a few hundred meters when a reflective surface existed near to the receiver. These results were reported at AROB2008. Therefore, in the second stage of this study, we tried a new idea that uses a low-frequency sound and sends plural-frequency sounds simultaneously. We then tried to assess the performance of this method. This article gives the results of the simulation and the water tank experiment. With these improvements, the influence of reflected sound was suppressed and the positioning area became larger than in previous research. Then the underwater positioning system using sound power transmission loss will become useful.