Search

For underwater acoustic (UWA) communication in sensor networks, the sensing information can only be interpreted meaningfully when the location of the sensor node is known. However, node localization is a challenging problem. Global Navigation Satellite Systems (GNSS) used in terrestrial applications do not work underwater. In this paper, we propose and investigate techniques based on matched field processing for localization of a single-antenna UWA communication receiver relative to one or more transmit antennas. Firstly, we demonstrate that a non-coherent ambiguity function (AF) allows significant improvement in the localization performance compared to the coherent AF previously used for this purpose, especially at high frequencies typically used in communication systems. Secondly, we propose a two-step (coarse-to-fine) localization technique. The second step provides a refined spatial sampling of the AF in the vicinity of its maximum found on the coarse space grid covering an area of interest (in range and depth), computed at the first step. This technique allows high localization accuracy and reduction in complexity and memory storage, compared to single step localization. Thirdly, we propose a joint refinement of the AF around several maxima to reduce outliers. Numerical experiments are run for validation of the proposed techniques. [ABSTRACT FROM AUTHOR]

The derivation of low-complexity of recursive least square (RLS) adaptive filtering algorithms is discussed.

The problem of providing simultaneous high data rate communication links to multiple moving railway trains from high-altitude platforms (HAPs) using a 'smart' antenna array is addressed. Solutions are provided subject to system-level constraints: the number and position of interferers (trains), motion of both the trains and the HAP, and link budget for broadband transmission in the 31/28 GHz bands, assuming line-of-sight propagation. Techniques are studied for estimation of the number of trains present, accurate direction-of-arrival (DOA) estimation, and tracking of multiple trains for the purposes of beamforming (BF) and reliable attribution of DOA estimates to trains. A range of train scenarios, including trains crossing and passing through tunnels and shadowed stations, is considered. It is shown that extended Kalman filtering (EKF) applied to DOA estimates, in addition to providing improved accuracy of positional estimation of trains, also provides much more reliable attribution of DOA estimates to trains compared to DOA estimation alone--particularly when trains cross or pass closely. Methods for adapting EKF for tracking trains passing through tunnels and shadowed stations are proposed, and it is shown that tracking can be initialized to cope with both slow variations in train velocity and sudden HAP motion. BF that is based on null steering is shown to be beneficial in HAP-train data communications, even for small number of trains, providing increased stability of signal-to-interference-plus-noise ratio. Based on the analysis, array signal processing techniques are proposed based on an iterative root-MUSIC DOA method, extended Kalman tracking, and Capon BE Index Terms--Array signal processing, high-altitude platforms (HAPs), smart antennas, stratospheric platforms, tracking filters.

The low-complexity delayless acoustic echo cancellation methods are proposed and investigated based on frequency-domain spline-identification. The results have shown that these methods have provided cancellation performance better than that of the fast offline projection (FAP) algorithm, especially in double-talk and noisy environments, with a lower complexity.

Clarke's model of time variations of path amplitudes in multipath fading channels with Doppler scattering is used for the investigation of polynomial spline approximation of stationary random processes. The results reveal that local spline approximation is attractive for implementation, for both low processing delay and small approximation error.

This paper concerns the estimation of a frequency offset of a known (pilot) signal propagated through a slowly fading multipath channel, such that channel parameters are considered to be constant over the observation interval. We derive a maximum-likelihood (ML) frequency estimation algorithm for additive Gaussian noise and path amplitudes having complex Gaussian distribution when covariance matrices of the fading and noise are known; we consider in detail the algorithm for the white noise and Rayleigh fading, in particular, for independent fading of path amplitudes and pilot signals with diagonal autocorrelation matrices. For the latter scenario, we also derive an ML frequency estimator when the power delay profile is unknown, but the noise variance and bounds for the path amplitude variances are specified; in particular, this algorithm can be used when path delays and amplitude variances are unknown. Finally, we consider frequency estimators which do not use a priori information about the noise variance; these algorithms are also operable without timing synchronization. All the frequency estimators exploit the multipath diversity by combining periodograms of multipath signal components and searching for the maximum of the combined statistic. For implementation of the algorithms, we use a fast Fourier transform-based coarse search and fine dichotomous search. We perform simulations to compare the algorithms. The simulation results demonstrate high accuracy performance of the proposed frequency estimators in wide signal-to-noise ratio and frequency acquisition range. Index Terms--Discrete Fourier transforms, fading channels, fast Fourier transforms, frequency estimation, maximum-likelihood estimation, multipath channels, Rayleigh channels.

For classification of underwater acoustic signals, we propose a novel sparse anisotropic chirplet transform (ACT) to reveal fine time–frequency structures. The signal features in the form of a time–frequency map are fed into a deep convolutional neural network, referred to as a time–frequency feature network (TFFNet), which brings flexibility to signal classification. The TFFNet is based on a novel efficient feature pyramid enhancing feature (EFP) maps by aggregating the context information at different scales. To remove the gridding artifacts on enhanced feature maps, a form of aggregating transformation, a forward feature fusion, is utilized to merge the forward feature maps. Main contributions of this work are a novel sparse ACT, a TFFNet classifier, and an EFP with forward feature fusion. Experimental results demonstrate that the sparse ACT provides a high-resolution time–frequency representation of underwater signals and the TFFNet improves the classification performance compared to known networks and two machine learning methods (random forest and support vector machine with radial basis function kernel) on two real data sets, an underwater acoustic communication signal data set and whale sounds data set. [ABSTRACT FROM AUTHOR]

An adaptive estimation algorithm based on the Lorentzian norm is proposed for echo cancellation in vehicle hands-free communication systems and video teleconferencing systems, namely the affine-projection Lorentzian (APL) algorithm. By minimizing the Lorentzian norm of the a posteriori error vector with a suitable constraint on the weight vector and providing a dynamic Lorentzian-norm-controlling parameter, the proposed APL algorithm achieves robustness against impulsive disturbances and speeds up convergence for colored input signals. The computational complexity of the APL algorithm is analyzed and a fast recursive filtering method is employed to reduce its complexity. The stability analysis, based on energy-conservation arguments, shows that the APL algorithm converges. Furthermore, its tracking behavior is also investigated and a step size optimizing the tracking performance is derived. Simulation results agree well with the theoretical analysis. Simulation results for channel estimation and in-car echo cancellation scenarios demonstrate that the APL algorithm achieves better performance compared to the maximum-correntropy-criterion, affine-projection-generalized-maximum-correntropy, affine-projection-sign, affine-projection-like M-estimate, and Lorentzian adaptive filtering algorithms in various impulsive interference environments. [ABSTRACT FROM AUTHOR]

A kernel affine projection-like algorithm (KAPLA) is proposed in reproducing kernel Hilbert space in non-Gaussian environments. The cost function for the developed algorithm is constructed by using the correntropy approach and Gaussian kernel to deal with nonlinear channel estimation. The devised algorithm can efficiently operate in the impulse noise. As a consequence, the proposed KAPLA algorithm provides good performance for nonlinear channel equalization in impluse-noise environments. Simulations results in different mixed noise environments verify the superior behavior of KAPLA compared to known algorithms. [ABSTRACT FROM AUTHOR]

This work considers the processing of acoustic data from a multibeam forward-looking sonar (FLS) on a moving underwater platform to estimate the platform's attitude and trajectory. We propose an algorithm to produce an estimate of the attitude trajectory for an FLS based on the optical flow between consecutive sonar frames. The attitude trajectory can be used to locate an underwater platform, such as an autonomous underwater vehicle, to a degree of accuracy suitable for navigation. It can also be used to build a mosaic of the underwater scene. The estimation is performed in three steps. First, a selection of techniques based on the optical flow model is used to estimate a pixel displacement map (DM) between consecutive sonar frames represented in the native polar (range/bearing) format. The second step finds the best match between the estimated DM and DMs for a set of modeled sonar sensor motions. To reduce complexity, it is proposed to describe the DM with a small parameter vector derived from the displacement distribution. Thus, an estimate of the incremental sensor motion between frames is made. Finally, using a weighted regularized spline technique, the incremental interframe motions are integrated into an attitude trajectory for the sonar sensor. To assess the accuracy of the attitude-trajectory estimate, it is used to register FLS frames from a field experiment data set and build a high-quality mosaic of the underwater scene. [ABSTRACT FROM AUTHOR]

In underwater acoustic (UWA) communications, Doppler estimation is one of the major stages in a receiver. Two Doppler estimation methods are often used: the cross-ambiguity function (CAF) method and the single-branch autocorrelation (SBA) method. The former results in accurate estimation but with a high complexity, whereas the latter is less complicated but also less accurate. In this paper, we propose and investigate a multibranch autocorrelation (MBA) Doppler estimation method. The proposed method can be used in communication systems with periodically transmitted pilot signals or repetitive data transmission. For comparison of the Doppler estimation methods, we investigate an orthogonal frequency-division multiplexing (OFDM) communication system in multiple dynamic scenarios using the Waymark simulator, allowing virtual UWA signal transmission between moving transmitter and receiver. For the comparison, we also use the OFDM signals recorded in a sea trial. The comparison shows that the receiver with the proposed MBA Doppler estimation method outperforms the receiver with the SBA method and its detection performance is close to that of the receiver with the CAF method, but with a significantly lower complexity. [ABSTRACT FROM AUTHOR]

Underwater acoustical communication channels are characterized by the spreading of received signals in space (direction of arrival) and in time (delay). The spread is often limited to a small number of space-time clusters. In this paper, the space-time clustering is exploited in a proposed receiver designed for guard-free orthogonal frequency-division multiplexing with superimposed data and pilot signals. For separation of space clusters, the receiver utilizes a vertical linear array (VLA) of hydrophones, whereas for combining delay-spread signals within a space cluster, a time-domain equalizer is used. We compare a number of space-time processing techniques, including a proposed reduced-complexity spatial filter, and show that techniques exploiting the space-time clustering demonstrate an improved detection performance. The comparison is done using signals transmitted by a moving transducer, and recorded on a 14-element nonuniform VLA in sea trials at distances of 46 and 105 km. [ABSTRACT FROM PUBLISHER]

In practical implementations of estimation algorithms, designers usually have information about the range in which the unknown variables must lie either due to physical constraints (such as power always being non-negative) or due to hardware constraints (such as in implementations using fixed-point arithmetic). In this letter, we propose a fast (i.e., whose complexity grows linearly with the filter length) version of the dichotomous coordinate descent recursive least-squares (RLS) adaptive filter which can incorporate constraints on the variables. The constraints can be in the form of lower and upper bounds on each entry of the filter, or norm bounds. We compare the proposed algorithm with the recently proposed normalized non-negative least-mean-squares (N-NLMS) and projected-gradient normalized LMS (PG-NLMS) filters, which also include inequality constraints in the variables. [ABSTRACT FROM AUTHOR]

We develop adaptive beamforming algorithms that are robust against sensor failure and ill-conditioning in the autocorrelation matrix (common in low-rank interference scenarios). Both goals are achieved simultaneously through the use of l1 regularization. The algorithms are based on the complex adaptive reweighting homotopy technique. We also develop iterative versions of the algorithms that take advantage of properties of homotopy l1 solvers and dichotomous coordinate iterations to reduce considerably the computational complexity, compared with other regularization methods. [ABSTRACT FROM PUBLISHER]

In this paper, we consider orthogonal frequency-division multiplexing (OFDM) transmission in fast-varying underwater acoustic channels. We demonstrate on experimental data that reliable communications can be achieved without any guard interval (such as cyclic prefix or zero padding) and with a superimposed pilot. Such OFDM transmission possesses a high spectral efficiency, but incurs severe intersymbol and intercarrier interference, and interference from the superimposed pilot. We propose a receiver that can efficiently deal with the interference and has a relatively low complexity as most of its operations are based on fast Fourier transform and local spline interpolation. The receiver is verified in an experiment with a transducer towed by a surface vessel moving at a high speed; a complicated trajectory of the transducer resulted in a severe Doppler distortion of the signal received on a single hydrophone. The performance of the proposed receiver is investigated for different parameter settings and compared with an ideal receiver with perfect channel knowledge, operating in interference-free scenarios, and mimicking the signal-to-noise ratio (SNR) of the experiment. The proposed receiver has provided error-free detection of encoded data at data rates of 0.5 b/s/Hz at a distance of 40 km and 0.33 b/s/Hz at a distance of 80 km, approaching the performance of the ideal receiver with a less than 3-dB loss in SNR. [ABSTRACT FROM AUTHOR]

Locating multiple underwater acoustic sources is a problem that can be solved using antenna array beamforming based on the matched field (MF) processing. However, known MF beamforming techniques fail to provide good performance for multiple sources, a high noise power, and/or when the sources are close to each other. This paper proposes an MF technique for solving the localization problem. The proposed technique exploits formulation of the localization problem in terms of sparse representation of a small number of source positions among a much larger number of potential positions. The sparse representation is formulated as the basis pursuit de-noising (BPDN) problem for complex-valued variables. The solution is found as a joint solution to a set of BPDN problems corresponding to the set of source frequencies subject to the joint support. The joint BPDN problem is efficiently solved using the Homotopy approach and coordinate descent search. For further reduction in the complexity, a position grid refinement method is applied. Using simulated and real experimental data, it is shown that the technique can provide accurate source localization for multiple sources. The proposed technique outperforms other MF techniques in resolving sources positioned closely to each other, tolerance to the noise and capability of locating multiple sources. [ABSTRACT FROM PUBLISHER]

The matched-phase coherent broadband matched-field (MF) processor has been previously proposed and shown to outperform other advanced broadband MF processors. It has been previously proposed to search the matched phases using the simulated annealing, which is well known for its ability of solving global optimization problems while having high computational complexity. This prevents simultaneous processing of many frequencies, and thus, limits the processor performance. We propose to use a novel iterative technique, the phase descent search (PDS), for searching the matched phases. This technique is based on coordinate descent optimization which is mainly applicable to solving convex problems. In this work, we investigate its application to the phase search problem, which is a nonconvex problem. We show that the PDS algorithm obtains matched phases similar to that obtained by the simulated annealing, and has significantly lower complexity. Therefore, it enables to search phases for a large number of frequencies and significantly improves the processor performance. The proposed processor is applied to real data from the 1996 Shallow Water Experiment (SWellEx-96) for locating a moving acoustic source at distances between 1 and 9 km with a step of about 150 m. At each distance, one 1-s snapshot with 13 frequencies is enough to provide accurate localization of the source well matched to global positioning system (GPS) measurements. [ABSTRACT FROM PUBLISHER]

We propose a new method of modeling the signal transmission in underwater acoustic communications when the transmitter and receiver are moving. The motion-induced channel time variations can be modeled by sampling the transmitter/receiver trajectory at the signal sampling rate and calculating, for each position, the channel impulse response from the acoustic-field computation. This approach, however, would result in high complexity. To reduce the complexity, the channel impulse response is calculated for fewer (waymark ) positions and then interpolated by local splines to recover it at the signal sampling rate. To allow higher distances between waymarks and, thus, further reduction in the complexity, the multipath delays are appropriately adjusted before the interpolation. Because, for every time instant, this method only requires local information from the trajectory, the impulse response can recursively be computed, and therefore, the signal transmission can be modeled for arbitrarily long trajectories. An approach for setting the waymark sampling interval is suggested and investigated. The proposed method is verified by comparing the simulated data with data from real ocean experiments. For a low-frequency shallow-water experiment with a moving source that transmits a tone set, we show that the Doppler spectrum of the received tones is similar in the simulation and experiment. For a higher frequency deep-water experiment with a fast-moving source that transmits orthogonal frequency-division multiplexing (OFDM) communication signals, we investigate the detection performance of a receiver and show that it is similar in the simulation and experiment. [ABSTRACT FROM PUBLISHER]

In this letter, we propose a low-complexity channel-estimate based adaptive linear equalizer. The equalizer exploits coordinate descent iterations for computation of equalizer coefficients. The proposed technique has as low complexity as \cal O(Nu(K+M)) operations per sample, where K and M are the equalizer and channel estimator length, respectively, and Nu is the number of iterations such that N_u \ll K and N_u \ll M. Moreover, with dichotomous coordinate descent iterations, the computation of equalizer coefficients is multiplication-free and division-free, which makes the equalizer attractive for hardware design. Simulation shows that the proposed adaptive equalizer performs close to the minimum mean-square-error equalizer with perfect knowledge of the channel. [ABSTRACT FROM AUTHOR]

In this letter, a new low-complexity implementation of the affine projection (AP) adaptive filtering algorithm is proposed and investigated by simulation. The proposed algorithm uses a novel low complexity recursive filtering technique and filter update that is incorporated in dichotomous coordinate descent (DCD) iterations. If the projection order is significantly smaller than the filter length L, the complexity of the proposed DCD-AP algorithm is as small as about L multiplications per sample. [ABSTRACT FROM AUTHOR]

• We have proposed kernel least lncosh (KLL) and quantized kernel least lncosh (QKLL) algorithms in the reproducing kernel Hilbert space under non-Gaussian environment. • The online vector quantization (VQ) is used to quantize the input space to construct a robust adaptive filtering algorithm. • The mean square convergence analysis of the QKLL algorithm has been conducted. • The performance of the KLL and QKLL algorithms is demonstrated by the Mackey–Glass (MG) time series prediction and nonlinear channel equalization. This paper introduces the kernel least lncosh (KLL) algorithm, in which the lncosh (logarithm of hyperbolic cosine) cost function is successfully applied in the reproducing-kernel-Hilbert space. The online vector quantization (VQ) is then used to quantize the input space to construct the algorithm which can curb the growth of network size. As a result, the quantized kernel least lncosh (QKLL) algorithm is developed, which is robust in non-Gaussian environments. The sufficient condition for mean-square convergence of the QKLL algorithm has been conducted. The performance of the KLL and QKLL algorithms is demonstrated by the short-term chaotic time-series prediction and non-linear channel-equalization (NCE). [ABSTRACT FROM AUTHOR]