This letter addresses the adaptive classification of the Interference Covariance Matrix (ICM) structures in radar applications. This is an essential issue when the design assumptions do not perfectly match the actual operating scenario due to environment uncertainties. Thus, in this letter, we propose a classifier capable of identifying the ICM structure as either complex Hermitian or real-valued symmetric. To this end, a Bayesian approach is employed by assuming a suitable model for the probability density function of the unknown ICM. This classification problem is firstly formulated in terms of a binary hypothesis test and the posterior probability is maximized to devise the classifier. Furthermore, the classifier resorts to secondary data only which are obtained from the adjacent cells around the cell under test and share the same ICM structure as the primary data. The illustrative examples conducted on simulated data have confirmed the superiority of the proposed classifier compared with its state-of-the-art non-Bayesian counterparts.

In this paper, we address the problem of detecting the signal of interest in the presence of Gaussian clutter with symmetric spectrum. To this end, we exploit the spectral properties of the clutter to transfer the binary hypothesis test problem from complex domain to real domain. Then, we devise and assess a detection strategy based on the so-called two-step Generalized Likelihood Ratio Test (GLRT) design procedure. Finally, a preliminary performance assessment, conducted by resorting to simulated data, has confirmed the effectiveness of the newly proposed detector compared with the traditional state-of-the-art counterparts which ignore the spectrum symmetry.

In the present letter we investigate the problem of adaptive detection and range estimation for point-like targets buried in partially homogeneous Gaussian disturbance with unknown covariance matrix. To this end, we jointly exploit the spillover of target energy to consecutive range samples and the oversampling of the received signal. In this context, we design a detector relying on the Generalized Likelihood Ratio Test (GLRT). Remarkably, the new decision scheme ensures the Constant False Alarm Rate (CFAR) property with respect to the unknown disturbance parameters. The performance analysis reveals that it can provide enhanced detection performance compared with its state-of-art counterpart while retaining accurate estimation capabilities of the target position.

In this paper, we propose a sliding-mode-based stochastic distribution control algorithm for nonlinear systems, where the sliding-mode controller is designed to stabilize the stochastic system and stochastic distribution control tries to shape the sliding surface as close as possible to the desired probability density function. Kullback-Leibler divergence is introduced to the stochastic distribution control, and the parameter of the stochastic distribution controller is updated at each sample interval rather than using a batch mode. It is shown that the estimated weight vector will converge to its ideal value and the system will be asymptotically stable under the rank-condition, which is much weaker than the persistent excitation condition. The effectiveness of the proposed algorithm is illustrated by simulation.

Variable length coding (VLC) is widely used in the entropy codec part of various video codec standards. Its parallelism is difficult to be developed due to its inherent bit-serial characteristic. As it becomes the system bottleneck of real-time video decoding, this paper presents a bitstream co-processor which greatly accelerates the codec of VLC by designing corresponding bitstream instructions. With SuperV DSP as the research prototype, the area cost of the bitstream co-processor only increases by 1.71% while the performance of the application of entropy decoding is improved about 73% on average.

A delay slot scheduling framework for VLIW architectures is presented in this paper. In this framework, an assembly data dependence graph is proposed to describe the data dependences among instructions in a basic block. An assembly control flow graph is also proposed to describe control relations among basic blocks. With the help of predicate analysis technology, majority invalid control flows are detected. Along the edges in these two graphs, instructions are selected to be filled into delay slots. Additionally, a scheme to balance the local scheduling and the global scheduling is proposed in this paper. This framework is evaluated by several experiments on the SuperV-DSP processor. And the results demonstrate the effectiveness of the new approach.

This letter deals with the problem of adaptive detection in partially-homogeneous Gaussian disturbance with unknown but persymmetric structured covariance matrix. Since no uniformly most powerful test exists for the problem at hand, we devise and assess two detection strategies based on the Rao test and the Wald test design criteria. Remarkably, both detectors ensure the constant false alarm rate property with respect to both the structure of the covariance matrix as well as the power level. The preliminary performance assessment, conducted by resorting to simulated data, has confirmed the effectiveness of the newly proposed detectors.

In this paper we consider the problem of detecting multiple point-like targets in the presence of steering vector mismatches and Gaussian disturbance with unknown covariance matrix. To this end, we first model the actual useful signal as a vector belonging to a proper cone whose axis coincides with the whitened direction of the nominal array response. Then we develop two new robust adaptive detectors resorting to the two-step generalized-likelihood ratio test (GLRT) design procedure without assignment of a distinct set of secondary data. Finally, a performance assessment, conducted by Monte Carlo simulation, show that the proposed detectors achieve a visible performance improvement over their natural counterparts.

Most of the existing spherical array modal beamformers are implemented in the frequency domain, where a block of snapshots is required to perform the discrete Fourier transform. In this paper, an approach to real-valued time-domain implementation of the modal beamformer for broadband spherical microphone arrays is presented. The microphone array data are converted to the spherical harmonics domain by using the discrete spherical Fourier transform, and then steered to the look direction followed by the pattern generation unit implemented using the filter-and-sum structure. We derive the expression for the array response, the beamformer output power against both isotropic noise and spatially white noise, and the mainlobe spatial response variation in terms of the finite impulse response (FIR) filters' tap weights. A multiple-constraint problem is then formulated to find the filters' tap weights with the aim of providing a suitable tradeoff among multiple conflicting performance measures such as directivity index, robustness, sidelobe level, and mainlobe response variation. Simulation and experimental results show good performance of the proposed time-domain broadband modal beamforming approach.

Visual attention detecting has drawn great research interest due to its useful application in the area of multimedia. Recently, a fast spectral residual approach in frequency domain was proposed to find the Region of Interest in an image. It was proved to be able to give better performance than the classic approaches. In this paper, we propose a new model for Region of Interest detection based on the phase spectrum. First, we analyze the reason that the salient region can create “pop-out” in the feature maps of the phase spectrum in spatial domain. Second, in order to obtain better feature maps, we propose to blur the image and then get the feature maps based on the phase spectrum of the blurred image. Last, we present a model to generate Region of Interest accordingly. Experimental results show that our approach achieves better and more robust results than that from the state-of-the-art.

In this paper, we introduce a practical real-time speech enhancement system based on the TI TMS320VC5509A digital signal processor (DSP). A stereo audio codec is used as the analog input/output (I/O). The algorithm for noise reduction is an improved and optimized spectral-subtraction method. The detailed design of the hardware and software is described and the experiments show an improved performance. Low power consumption and high efficiency make the whole system suitable for portable devices.

The goal of this letter is to derive robust adaptive beamformers via generalized loading. In the proposed methods, Hermitian matrices are loaded on sample covariance matrix, and this is different from those methods based on the well-known diagonal loading approach. Furthermore, the computation of the loaded matrix is fully automatic, which is scarce in the literature. Numerical examples show that our methods are more robust to errors on array steering vector and sample covariance matrix than other tested parameter-free methods.

In this letter, we propose shrinkage-based Capon (S-Capon) and APES (S-APES) spectral estimators by minimizing the mean-square error (MSE) of standard Capon and APES in a linear regression framework. The proposed methods are shown to give more accurate spectral estimates but lower resolution than the methods they based on. We combine Capon with the proposed S-Capon and S-APES to overcome the resolution limit of shrinkage-based methods for estimation of both frequency and amplitude of spectral lines. The so-obtained Capon-SCapon and Capon-SAPES spectral estimators, which have about the same computational complexity as Capon, are compared with the Capon-APES (CAPES) by numerical examples. Simulations show that the Capon-SCapon performs similarly to CAPES in a wide range of signal-to-noise ratio, and the Capon-SAPES always gives more accurate spectral amplitude (less bias and lower MSE) than CAPES.

In this letter, we develop a fully automatic robust linear receiver technique for joint space-time decoding and interference rejection in multi-access MIMO systems that use orthogonal space-time block codes and erroneous channel state information (CSI). The proposed receiver does not need any a priori knowledge of channel estimation errors and has a simple closed form. Numerical examples show that our method usually gives good performance in case of non-perfect CSI and/or low sample sizes when compared with other tested linear receivers.

The paper investigates the combinatorial mechanism of a general case of exponential backoff (EB), with factor, r, and retransmission cutoff, with traffic sources consisting of an infinite number of stations in ideal channel conditions. A new, simple, but more exact, infinite-station bidimensional discrete-time analytical model based on a Markov chain, to reflect the actual behavior of this combinatorial mechanism, is presented. By means of the proposed model, we provide an extensive performance evaluation, and new analytical results are given. We also obtain the analytical expressions for the maximum saturation throughput, which converges to a non-zero constant as the station number goes to infinity. The accuracy of the analysis is verified by elaborate simulation results. In addition, the packet rejection rate is discussed for various values of system offered load and initial minimum contention window.

Controlled sharing of the wireless link by different packet streams is critical in transporting multimedia traffic across wireless networks. Consequently, it has been treated as providing support for quality of service (QoS) in medium access control (MAC) protocols. This paper proposes a distributed class-based access scheme whose allocation policy adopts token passing packet scheduling, designed to handle efficiently different service classes of traffic in heavy load wireless networks. The traffic is divided into three classes by QoS specifications and packet characteristics, and served with asymmetric exhaustive discipline. The probability generating function based on embedded Markov chain for this discrete-time system is derived. We employ basic queuing analysis methodologies, calculate the average buffer size of each class queue at polling instants and then apply Little's law to find the average delay. Experiments are conducted to validate the new strategy, and the performance is analyzed and evaluated. Finally, we visualize and compare how the relative variable parameters effect on system characteristics.

Scheduling at packet level is one of the critical technologies in transporting multimedia traffic across wireless networks for providing quality of service (QoS) differentiation and guarantees. Also QoS support in wireless scheduling is dictated by the network architecture and service mode. Moreover, wireless communication poses special problems and difficulties that do not exist in wireline networks. This paper develops a self-organizing, multihop, mobile wireless network, which relies on a token passing access scheme for multimedia support. In the proposed network architecture, mobile nodes are grouped into clusters by QoS specifications and packet characteristics. The clusters are independently controlled, and are dynamically reconfigured as node move. Hence, the network is robust in the face of topological changes caused by node motion, node failure, and node insertion/removal. We model it as a discrete-time system with asymmetric exhaustive service policies by imbedding Markov chain, and apply qualitative and quantitative analysis to validate if the new strategy can provide service at a certain desired performance. Through simulations we show that this architecture provides an efficient, stable infrastructure for the integration of different types of traffic in a dynamic wireless networks. In addition, how the relative variable parameters effect on system characteristics is discussed and compared.

A neuron-based algorithm for solving the complex principal components analysis problem and its application to adaptive bearing estimation are presented. The authors specify the bearing estimation problem in a narrowband version and use the eigen-decomposition method to achieve high resolution. Both input data and eigenvectors that span the signal subspace are complex values. So it is important to extract the complex principal components from the complex input data sequence. Previous methods do not offer complex algorithms. To overcome this problem, the authors propose a linear neural network with complex weights which is a generalized and modified version of E. Oja's (1985) and S.Y. Kung and K.I. Diamantaras's (1990) methods, and they use their own method to estimate the direction of arrival (DOA).>

This paper describes a new method of adaptive array processing for the unknown frequency characteristics. In the practical situations, it is easy to obtain the estimation of the power spectrum in the observed direction instead of the frequency characteristics. Then the maximum entropy spectral analysis can be incorperated with the adaptive array processing. The predicted coefficients of MESA can be obtained. The predicted coefficients can be used as the constrained conditions of the algorithm for the adaptive array processing. In this paper the above method is further expanded and the adaptive array processor is presented for the more general cases, in which the signal characteristics in the observed direction are completely unknown. A series of computer simulation experiments have been conducted.