• Journal of the Korea Institute of Military Science and Technology
• /
• v.12 no.2
• /
• pp.159-169
• /
• 2009

In this paper, optimum design of distributed detection is considered for a parallel sensor network system consisting of a fusion center and two passive sonar nodes. AND rule and OR rule are employed as the fusion rules of the sensor network. For the fusion rules, it is shown that a threshold rule of each sensor node has uniformly most powerful properties. Optimum threshold for each sensor is investigated that maximizes the probability of detection under the constraint of a specified probability of false alarm. It is also investigated through numerical experiments how signal strength, false alarm probability, and the distance between two sensor nodes affect the system detection performances.

• The Journal of the Acoustical Society of Korea
• /
• v.29 no.1
• /
• pp.10-17
• /
• 2010

In a passive sonar, the improvement of detection performance by using noise cancellation is usually a important problem. In this paper, we have analyzed the own-ship noise cancellation in the two operation modes which are used in the HMS system. In the operator mode, an adaptive line enhancer(ALE) is applied to improve the tonal detection by using broadband noise cancellation and the normalized least mean square(NLMS) algorithm is applied to the design of an adaptive filter. The reference input that is correlated with a primary input can be used to remove the noise incident on the observation directionin the automatic mode. Computer simulations with real sea that data show that the proposed adaptive noise canceller has good performance in passive detection under HMS operation.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.6
• /
• pp.9-16
• /
• 1999

As a part of target motion analysis(TMA) with highly noisy bearings-only measurements from a passive sonar system, a nonlinear batch estimator is proposed to provide the initial estimates to a sequential estimator called the modified gain extended Kalman filter(MGEKF). Based on the system observability analysis of passive target tracking, a practical and effective method is suggested to determine the observer maneuvers for improved TMA performance through system observability enhancing. Also suggested is a method to determine observer location for enhanced system observability at the initial phase of TMA from various engagement boundaries which represent the relationship between observer-target relative geometrical data and system observability. The proposed TMA methods are tested by a series of computer simulation runs.

• The Journal of the Acoustical Society of Korea
• /
• v.43 no.2
• /
• pp.234-242
• /
• 2024

We propose a passive sonar signal classification algorithm using Graph Neural Network (GNN). The proposed algorithm segments spectrograms into image patches and represents graphs through connections between adjacent image patches. Subsequently, Graph Convolutional Network (GCN) is trained using the represented graphs to classify signals. In experiments with publicly available underwater acoustic data, the proposed algorithm represents the line frequency features of spectrograms in graph form, achieving an impressive classification accuracy of 92.50 %. This result demonstrates a 8.15 % higher classification accuracy compared to conventional Convolutional Neural Network (CNN).

• Journal of Ocean Engineering and Technology
• /
• v.34 no.4
• /
• pp.277-284
• /
• 2020

Underwater acoustics, which is the study of phenomena related to sound waves in water, has been applied mainly in research on the use of sound navigation and range (SONAR) systems for communication, target detection, investigation of marine resources and environments, and noise measurement and analysis. The main objective of underwater acoustic remote sensing is to obtain information on a target object indirectly by using acoustic data. Presently, various types of machine learning techniques are being widely used to extract information from acoustic data. The machine learning techniques typically used in underwater acoustics and their applications in passive SONAR systems were reviewed in the first two parts of this work (Yang et al., 2020a; Yang et al., 2020b). As a follow-up, this paper reviews machine learning applications in SONAR signal processing with a focus on active target detection and classification.

• The Journal of the Acoustical Society of Korea
• /
• v.20 no.7
• /
• pp.31-36
• /
• 2001

Passive sonar system forms the various beams in any desired directions to obtain the improvement in Signal-to-Noise (S/N) ratio, bearing detection and localization of targets, and the attenuation of interferences from other directions. The improvement of beamforming is very important to detect modern underwater targets as noise reduction technology leads to considerably low-level acoustic emissions in the long range in complex environmental sea. In this paper, we proposed the spatial cross correlation beamforming (SCCBF) algorithm using cross correlation matrix of individual hydrophone pairs of linear array sensors. By the theoretical analysis and simulation, the proposed SCCBF is demonstrated that its performances compared to conventional beamforming (CBF) output can be obtain above 3dB of array gain and about half of beam width represented the bearing accuracy in target detection. Also, this paper presents sea test result of linear passive sonar system that the proposed algorithm implemented.

• International Journal of Naval Architecture and Ocean Engineering
• /
• v.4 no.2
• /
• pp.96-111
• /
• 2012

Hull Mounted Sonar (HMS) is a long range submerged vehicle's hull-mounted passive sonar system which detects low-frequency noise caused by machineries of enemy ships or submerged vehicles. The HMS needs a sound absorption /insulation multi-layer structure to shut out the self-noise from own machineries and to amplify signals from outside. Therefore, acoustic analysis of the multi-layer system should be performed when the HMS is designed. This paper simplified the HMS multi-layer system to be an infinite planar multi-layer model. Also, main excitations that influence the HMS were classified into mechanical, plane wave and turbulent flow excitation, and the investigations for each excitation were performed for various models. Stiffened multi-layer analysis for mechanical excitation and general multi-layer analysis for turbulent flow excitation were developed. The infinite planar multi-layer analysis was expected to be more useful for preliminary design stage of HMS system than the infinite cylindrical model because of short analysis time and easiness of parameter study.

• The Journal of the Acoustical Society of Korea
• /
• v.43 no.1
• /
• pp.1-8
• /
• 2024

For measuring the performance of passive sonars, we usually consider the maximum Detection Range (DR) under the environment and system parameters in operation. In shallow water, where sound waves inevitably interacts with sea surface or bottom, detection generally maintains up to the maximum range. In deep water, however, sound waves may not interact with sea surface or/and bottom, and thus there may exist shadow zones where sound waves can hardly reach. In this situation, DR alone may not completely define the performance of each sonar. For complete description of sonar performance, we employ the concept 'Robustness Of Detection (ROD)'. In the coastal region of the East Sea, the spatial variations of water masses have close relations with DR and ROD, where the two parameters show reverse spatial variations in general. The spatial and temporal analysis of the temperature by employing the Empirical Orthogonal Function (EOF) shows that the 1-st mode represents typical pattern of seasonal variation and the 2-nd mode represents strength variations of mixed layers and currents. The two modes are estimated to explain about 92 % of the variations. Assuming two types of targets located at the depths of 5 m (shallow) and 100 m (deep), the passive sonar performance (DR) gives high negative correlations (about -0.9) with the first two modes. Most of temporal variations of temperature occur from the surface up to 200 m in the water column so that when we assume a target at 100 m, we can expect detection performance of little seasonal variations with passive sonars below 100 m.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.148-148
• /
• 2000

In this paper the IMM(Interacting Multiple model) algorithm using the MGEKF(Modified Gain Extended Kalman Filter) which modes are variances of the process noises is proposed to enhance the performance of maneuvering target tracking with bearing and frequency measurements. The state are composed of relative position, relative velocity, relative acceleration and doppler frequency. The mode probability is calculated from the bearing and frequency measurements. The proposed algorithm is tested a series of computer simulation runs.

• The Journal of the Acoustical Society of Korea
• /
• v.38 no.2
• /
• pp.177-183
• /
• 2019

In this paper, we proposed an application method of ICA (Independent Component Analysis) to passive line array sonar to separate interferences from target signals in low frequency band and compared performance of three conventional ICA algorithms. Since the low frequency signals are received through larger bearing angles than other frequency bands, neighboring beam signals can be used to perform ICA as measurement signals of the ICA. We use three ICA algorithms such as Fast ICA, NNMF (Non-negative Matrix Factorization) and JADE (Joint Approximation Diagonalization of Eigen-matrices). Through experiments on real data obtained from passive line array sonar, it is verified that the interference can be separable from target signals by the suggested method and the JADE algorithm shows the best separation performance among the three algorithms.