Species classification using an acoustic sounder is important for fisheries. With schools of mixed species, it is necessary to be able to classify individual fish species from echoes, and to isolate individual fish echoes. A broadband signal, which offered the advantage of high-range resolution, was applied for this purpose, and the positions of fish were estimated using the split-beam system. The target strength (TS) spectrum of individual fish echoes was computed from isolated echoes and estimated positions. In this paper, these TS spectra were used as features of fish classification for machine learning. Also, it is well known that the TS spectra are dependent on not only fish species but on fish size. Therefore, it is necessary to classify both fish species and size using these features. We attempted to classify two species and two fish sizes using the Support Vector Machine (SVM) and Nearest Neighbor Algorithm (NNA) as machine learning. Subject species were chub mackerel (Scomber japonicas) and Japanese jack mackerel (Trachurus japonicus). The classification rates using the SVM were superior to those using the NNA. These rates were dependent on the frequency bandwidth and tilt angle. The classification rate was 71.6% with limitation of the tilt angles.
Underwater acoustic (UWA) communication is still a challenging area of research, as methods to deal with large intersymbol interference and Doppler spread are needed. We previously evaluated the performance of an orthogonal signal-division multiplexing (OSDM) scheme in a UWA communication system in a test tank and found that OSDM achieved better bit-error-rate (BER) performance than the single-carrier with decision-feedback-equalizer (DFE) method and the orthogonal frequency-division multiplexing (OFDM) method. However, no experiments have yet been conducted in an actual sea environment, which is much more complex than a test tank. In this paper, we report an initial demonstration of OSDM communication in an actual sea environment. The experiment was performed at a harbor for about 85 h by changing the transmission level. We also performed the communication using single-carrier DFE and OFDM, and confirmed that OSDM with a multichannel receiver is attractive to existing schemes; as in the test-tank results, OSDM achieved the best bit-error-rate performance compared to other schemes with practical complexity. We confirmed that OSDM can achieve a highly reliable communication for UWA channels, such as in shallow water.
To track individual fish within a dense school, it is necessary to isolate the echoes from multiple individual fish and then estimate their positions accurately. The signals from a broadband split-beam system offer the advantage of high-range resolution for this purpose. The range resolution and localization accuracy of this system were investigated by measuring the echoes from spheres in a tank. A tracking method exploiting the high-range resolution of the broadband signal was proposed. Echoes from schools of Japanese anchovy were measured in the ocean. Individual fish in these schools were measured separately and tracked using the proposed method. The technique allowed for accurate tracking of individual fish within dense schools. The estimated tracks can help us to better understand fish behavior under the sea.
Computer simulations of single-target echo waveforms have been useful for fisheries acoustics. In previous studies, the use of conventional scientific echosounders has been assumed, and therefore a narrowband tone-burst wave has been used as the transmit signal. However, computer simulation methods for recently developed broadband echosounders have not been fully established. In this paper, a method for the computer simulation of broadband single-target echo waveforms using a linear frequency-modulated signal is presented. This simulation method takes into account not only the backscattering amplitude of a target, but also transducer directivities and transmission losses. The simulated broadband single-target echo waveforms will be applicable to various studies. In this study, the broadband signal processing procedure for volume backscattering strength was verified using broadband multiple-target echo waveforms, which were generated by summing a large number of simulated single-target echo waveforms.