Abstract
This paper presents an innovative algorithm for Doppler-shift and multipath correction in a coherent-FSK modem, which is optimized for acoustic communications in shallow water underwater networks.The final modem will be used in the ANDREA project, whose goal is to deploy a reliable network in a fish farm in Mediterranean shallow waters with both fixed and mobile nodes. As we will discuss in this scenario both multipath and Doppler-shift are critical.On the one hand, multipath distortion is produced by echoes from the sea-bottom, the surface or submerged structures -e.g. fish farm structure. On the other hand, Doppler-shift is also very important in this case attending to the proposed scenario. Doppler-shift is produced as an effect of the relative movement between transmitter and receiver, e.g. Autonomous Underwater Vehicles (AUV); but it also can be produced by sound speed changes. All these aspects must be addressed in our system to deploy a reliable network.Since sensor networks are formed by a large number of nodes, sensor nodes must be low cost. Moreover, as underwater system maintenance is very expensive, node battery lifetime must be extended. Thus, an optimal solution should run reliable but still low complexity algorithms in low power architectures. With this purpose a new coherent FSK modem was released last year but, unfortunately, neither optimal multipath nor Doppler-shift correction algorithms have been found for coherent-FSK modulation in the literature so far.The aim of this paper is to find a balance between speed and complexity, as well as to reduce power consumption, but still having a reliable modem for using multi-node networks in shallow water environments. The implementation is simple and needs very low extra resources. An additional adaptive filter is needed for multipath correction and the already present FSK demodulator is reused for Doppler-shift correction. The design has been simulated in a bellhop model and experimentally tested. Results are presented in the paper.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have