High Spectral-Efficiency Noncoherent Underwater Acoustic Communication for Seafloor Observation Network

Noncoherent underwater acoustic communication is robust to channel conditions, but has low spectral efficiency. In order to improve its spectral efficiency, we propose to adopt a combination of an irregular recursive convolutional code (IrCC) with the on-off keying (OOK) modulation. The IrCC is optimized under a flat Rayleigh fading channel model, assuming the receiver has no knowledge about the channel state information. An iterative threshold estimation (ITE) algorithm is designed to improve the iterative decoding performance. For a fixed spectral efficiency of 0.5 bit/s/Hz, simulations show reliable communication is achieved at a signal-to-noise ratio (SNR) of 11 dB, and the gap to the channel capacity is about 4.4 dB. The proposed scheme is also verified by experimental data collected in a seafloor observation network where over the frequency band of 6–10 kHz, a data rate of 1400 bps is achieved at a communication distance of 2500 m.