Towards higher data rates in underwater acoustic communication using MIMO

There is increased activity in the subsea domain, which requires access to robust high-speed communication. Applications range from live video streaming within marine robotics to fish-telemetry in aquaculture. Existing underwater communication solutions are often unable to meet the new and emerging demands. New solutions are required. This thesis aims to improve the data rates in underwater communication, by means of coherent acoustic communication in the very high frequency (VHF) band. VHF underwater acoustic communication is relatively unexplored terrain, but its importance is expected to increase with the growing demands for larger bandwidths and higher data rates. Knowledge of VHF propagation channels is a prerequisite to achieving high data throughput. We have in this work examined shallow-water channels in the VHF band around 250 kHz, through channel sounding. The measurements were performed at high resolution using 64 hydrophones arranged in a line array. Delay spreads of hundreds of symbols characterize the measured channels, and Doppler spreads of tens of hertz. Long and rapid fluctuating channels form the basis for processing-intensive algorithms. A standard receiver structure is the multichannel decision feedback equalization (DFE) adapted with the recursive least square (RLS) algorithm. As the number of equalizer channels grows, the complexity of RLS gets problematic. We have experimentally shown that adaption based on low complexity LMS algorithm can have the same performance as the RLS when the number of hydrophone channels is large. Based on this, a multichannel receiver based on linear equalizer adapted with the LMS algorithm is developed and implemented in a field-programmable gate array (FPGA). Combined with a 64 hydrophone line array, we have demonstrated realtime information rates of 465 kbps over 100 m, and 155 kbps over 1000 m in the VHF band. The alternative to increasing the data throughput is to use the available bandwidth to support several users; potential applications are within aquaculture fish telemetry. Based on data recorded in an Atlantic salmon pen, we have presented VHF channel characteristics and demonstrated reliable 78 kbps burst based communication over 18 m, with the potential of serving hundreds of unsynchronized fish tags.

Bruk av akustisk VHF kanal, FPGA og flerkanal utjevning for sanntids høyhastighets kommunikasjon. Potensielt forbedret fisktelemetri for oppdrettsfisk.

Underwater communication systems on the market today are reliable system able to communicate over several kilometers. But the data rates are far from the rates we have been used to in terrestrial communication. Instead of Gbps (Gigabit per second) we measure, in the best case, data rates in kbps (kilobit per second), imposed by the underwater channel characteristics such as long multipath delay spreads, limited bandwidth and large Doppler shifts. However with emerging of underwater factories, more advanced environmental monitoring platforms and AUV's (Autonomous underwater vehicles), we believe current data rates will not be sufficient and a demand for higher data rates will follow. One popular method to increase data rates in reverberant terrestrial environments, is MIMO (Multiple Input Multiple Output). In MIMO several transmitting and receiving antennas are used to cope with multipath propagation, transmitting several uncorrelated data streams in parallel. In this project we want to make a first step towards the use of MIMO in underwater communication by examining spatial diversity for acoustic transducer arrays, and different diversity coding schemes.