We consider the problem of recovering block sparse signals with unknown block partition and propose a better alternative to extended block sparse Bayesian learning (EBSBL). The underlying relationship between the proposed method, EBSBL and pattern-coupled sparse Bayesian learning (PC-SBL) is explicitly revealed. The proposed method adopts a cluster-structured prior for sparse coefficients, which encourages dependencies among neighboring coefficients by properly manipulating the hyperparameters of the neighborhood. Due to entanglement of the hyperparameters, a joint sparsity assumption is made to yield a sub-optimal analytic solution. The alternative algorithm avoids high dictionary coherence in EBSBL, reduces the unknowns of EBSBL and explains the effectiveness of EBSBL. The proposed algorithm also avoids vulnerability of parameter choice in PC-SBL. Results of comprehensive simulations demonstrate that the proposed algorithm achieves performance that is close to the best performance of PC-SBL. In addition, it outperforms EBSBL and other recently reported algorithms particularly under noisy and low sampling scenarios.

Traditional range-instantaneous Doppler (RID) methods for maneuvering target imaging suffer from the problems of low resolution and poor noise suppression. We propose a new super-resolution inverse synthetic aperture radar (ISAR) imaging method based on deep-learning-assisted time–frequency analysis (TFA). Our deep neural network resembles the basic structure of a U-net with two additional convolutional-upsampling layers and $l_{1}$ -norm loss function for super-resolution generation and noise suppression. The neural network is trained in advance to learn the mapping function between the low-resolution time–frequency spectrum inputs and their high-resolution references. Then, the linear TFA assisted by the pretrained network is integrated into the RID-based ISAR imaging system and is found to achieve sharply focused and denoised target image with super-resolution. Both the simulated and real radar data are used to evaluate the performance of the proposed method. Numerical experimental results demonstrate the superiority of the proposed ISAR imaging method over traditional ones.

A gas sensor provides a system with signals to monitor and respond to events which matter. However, some widely deployed sensors behave with drift, leading to inaccurate measurement. The system could malfunction due to the drift. Even more, if the system is not robust enough, the mistaken measurement could trigger some serious consequences. A drift prediction framework is necessary to contain this malfunction. On the other hand, when sensors are deployed with large quantity, it limits the maintaining cost of them, including the calibration cost. Data-driven frameworks show good potential for cheap calibration. But there are some challenges. First, sensor drift keeps occurring. The drift models need to be retrained over time. Second, the drift models for the gas sensors of same type cannot be one-for-all way because they vary due to small manufacturing discrepancy. Third, drift models for a sensor could even be different only because the sensor is under different scenarios. In this paper, we propose a new framework for drift prediction with inter and intra node knowledge transfer (DP-IINK) for widely deployed low-cost gas sensors. Knowledge transfer is employed for inter- and intra-node way, as well as for continuous and discontinuous drift. Our evaluation shows that our proposed framework achieves a mean MAPE as low as 5.18% for drift prediction with less data.

The target spectrum, which is commonly used in feature extraction for underwater acoustic target classification, can be improperly recovered via conventional beamformer (CBF) owing to its frequency-variant spatial response and lead to degraded classification performance. In this paper, we propose a target spectrum reconstruction method under a sparse Bayesian learning framework with joint sparsity priors that can not only achieve high-resolution target separation in the angular domain but also attain beamwidth constancy over a frequency range at no cost of reducing angular resolution. Experiments on real measured array data show the recovered spectrum via our proposed method can effectively suppress interference and preserve more detailed spectral structures than CBF. This indicates our method is more suitable for target classification because it has the capability of retaining more representative and discriminative characteristics. Moreover, due to target motion and the underwater channel effect, the frequency of prominent spectral line components can be shifted over time, which is harmful to classification performance. To overcome this problem, we proposed a frequency shift-invariant feature extraction method with the help of elaborately designed frequency shift-invariant filter banks. The classification experiments demonstrate that our proposed methods outperform traditional CBF and Mel-frequency features and can help improve underwater recognition performance.