Abstract:Since the conventional second order cyclic statistics algorithms degenerate severely in the impulsive noise environment, this paper takes the -stable distribution as the noise model and presents an algorithm of Direction of Arrival (DOA) estimation based on the fractional lower order cyclic correlation. The algorithm transforms the DOA estimation of the wideband cyclostationary signals into the narrowband problem with the “center frequency” , by using the phase shift characteristics of the fractional lower order cyclic correlation. Simulation results indicate that the new algorithm is superior to the classical SC-SSF.

Abstract:It is known that choosing the appropriate MIMO modes in varied situations can achieve better transmission effects. We investigated the switch technology of the MIMO modes in TDD LTE downlink for the purpose of reaching the best average throughput as well as spectrum efficiency, and of enhancing the system’s transmission rate. The PDSCH simulation platform of TDD LTE downlink control channel is built to simulate the MIMO transmission modes. By comparing the simulated results in different modes the switch threshold is determined and an adaptive algorithm of switching MIMO modes is proposed, which render the user optimal the transmission efficiency in different situations.

Abstract:In order to achieve complex motion and navigation control, it is important to access space attitude information of control object. This paper presents an efficient low-cost design of attitude measurement system, among which their main features are: monitoring the attitude angles by fusing one tri-axial accelerometer and one tri-axial gyroscope integrated in ADIS16355 (MEMS inertial sensor module); collecting information and calculating attitude angles by using ARMv7 processor Cotex-M3. The makeup of the whole attitude measurement system, function of ADIS16355 and their installation on the object are presented. One kalman filter is then designed as sensor data fusion method for measuring real-time attitude angles. Finally, experimental results are presented based on the comparisons between fused data and reference data from an AHRS500GA instrument, showing that the proposed fusion method gives quite accurate attitude information of the controlled object and the whole attitude measurement system works reliably.

Abstract:A new approach is proposed for cross terms suppression for multi-component FM signals in space-time parameter estimation with antenna array . The method utilizes spatial domain information in multi-channel and employs a 2-dimension product filter to reduce cross terms of time-frequency distribution. The filtered TFD(Time-Frequency Distribution) can be applied to estimate DOAs(direction-of-arrival) of FM signals by time-frequency MUSIC(Multiple Signal Classification). Simulation results are presented verifying the efficacy of this method.

Abstract:The extensive application of low probability of intercept (LPI) radars caused most of the interception of radar signals with low SNR. The performance of the common parameter estimation algorithms of radar signals will be inevitably corrupted in low SNR. To estimate the parameters of radar signals combined pseudo-random binary phase code (PRBC) and linear frequency modulation (LFM) in low SNR, a novel method was proposed based on the poly-phase filter bank and high-order cumulant.Firstly,it realizes the uniformity signals in the frequency domain rapidly by the poly-phase filter bank.Then the Gaussian noise can be suppressed by using the short-time estimate of the third-order cumulant on each narrow frequency band.The estimation of the whole time-frequency matrixes can be obtained with envelope detection.Finally,the chirp rate, code-width and code-length of compound signal can be estimated by Radon transform of the Time-Frequency images. At the same time,the carrier frequency, band width, start-stop frequency can be estimated by extract the frequency curves. Simulation results show that all parameters of the compound signal can be evaluated precisely when SNR is more than -11dB.

Abstract:Abstract：To correctly classify advanced radar emitter signals, a novel approach using image feature of Choi-Williams time-frequency distribution for radar emitter signal recognition is proposed, which transforms the classification of emitter signals into image processing and image recognition. Time-frequency images of radar Emitter signals are obtained by using Choi-Williams distribution, and then these images are transformed into grayscale images. A series of image processing methods are employed for time-frequency image enhancement and de-noising, and the grayscale images are converted into binary images following. In addition, the areas not containing signal components from the edges of the image are removed. Finally, the centralize moments and pseudo-zernike moments are calculated as the feature for signal recognition, and the support vector machine is used to identify radar emitter signals automatically. Simulation results show that the proposed approach can achieve satisfying accurate recognition when signal-to-noise rate (SNR) varies in a large range. Even for SNR=-3dB, the proposed method which adopts pseudo-zernike moments works effectively as high as 92% recognition rate . The validity of the approach is demonstrated by experiments.

Abstract:A high speed data gathering system is designed for the seismic data acquisition system. CPCI crate is applied as the framework of the data gathering system. The number of the sample channels can be easily changed by plugging in or out the cards of the CPCI crate. Data can be gathered with high speed through the PCI bus of the CPCI crate. The use of DMA transferring simplifies the system control and improves the data gathering speed from multi-cards. Large cache is used to eliminate dead time of data gathering and improve the efficiency of DMA transaction. The experiment shows this system can satisfy 4 cables system, totally 7680 channels, requirements for data gathering system. This system features simple structure, high speed of data transferring, high performance of real-time data gathering, and can be easily applied to the other large scale data acquisition system.

Abstract:To improve the inpainting quality of the damaged videos with vertical line, scratch and characters, a video inpainting algorithm is proposed based on the sparse reconstruction of surfacelet in this paper. Since it has the three-dimensional sparse characteristics with multi-scales, multi-directions and efficient tree structure, the singular surface of a video can be captured effectively by the surfacelet transform. In the proposed scheme, the global optimization objective function of the original video is constructed based on surfacelet transform and solved using the relaxation algorithm to achieve the video inpainting. The proposed algorithm makes unnecessary the complex pretreatments such as segmentation, edge detection and so on, and improves the inpainting quality of texture and structure information in the video. Experimental results show that the inpainting quality of the proposed scheme is superior to that of the existing video inpainting algorithms.

Abstract:A Complete Parameters Extraction Algorithm (CPEA) is proposed to improve the performance of Electrical Impedance Scanning Imaging (EISI). CPEA includes two parts: modeling of a forward problem and a parameters extraction process. The modeling of the forward problem is based an electrostatic scattering theory. With the model, we present an analytical solution of the forward problem. The solution process of the forward problem can be done in a very short period compared with conventional methods using iteration process. Based on the proposed model of forward problem, parameters extraction process could be regarded as a constrained optimization problem which is solved by a modified simplex optimization algorithm. Simulation experiments show that CPEA method can not only achieve more information of cancerous lesion within breast but also be more accurate and robust compared with other information extraction algorithms.

Abstract:In this paper, a joint beamforming scheme is presented for multiuser multiple input multiple output (MIMO) system with limited feedback. The proposed scheme uses Per-User Unitary Rate Control (PU2RC) at the transmitter and the SINR-maximization-based antenna combining algorithm at the receiver, searching optimal receive combiner by using a generalized Rayleigh-Ritz method. Simulation results show the scheme makes full use of channel information, improves the performance of the conventional PU2RC system, and outperforms the zero forcing (ZF) joint beamforming schemes in terms of the sensitivity to the channel state information (CSI) quantization error, the complexity of user selection, and the sum rate in the case of low feedback rates conditions.

Abstract:Band selection can cut down the large amount of hyperspectral data and alleviate the Hughes phenomenon in supervised classification of ground objects. The generation and evaluation of subsets are two key factors for selection algorithm. A hybrid scheme of random search and heuristic search is proposed to generate the band subset. The method embeds the sequential search into the evolution optimization for better ability of the fine tune in local search space and thus behaves well in both global and local cases. In addition, the embed scheme guarantees the validity of solutions for the optimization algorithms. The performance comparisons among the proposed method, hybrid genetic algorithm (HGA), standard genetic algorithm (SGA) and sequential backward floating search (SFFS) are carried out in the experiments on band selection and classification with the hyperspectral data sets. The results show that the proposed method can obtain better subsets according to the evaluation criterion.

Abstract:Sensor-Track precision is the key factor affecting the correctness of the situation in MultiSensor data fusion. The principle of the error transfer and the correcting methods are investigated by modeling the affect of sensor measurement, coordinate conversion, the plat location and the baseline coordinate conversion. Simulation Results show that the error caused by the course and deck distortion can be added to the bearing measurement and can even be ignored when the elevation is less than 10 degree.

Abstract:The speedy and unbounded characteristic of data stream, whose application domain is unceasing increasing, enables mining technology to have a higher challenging more and more. In this paper, we propose a new online mining algorithm called CMNL-SW (Closed Map and Num List-Sliding Window) which uses two data structures, Closed Map stores closed itemsets mined and Num List stores the number of all different items. Via the simple union of item number contained within a new arriving or an old deleting transaction and intersection with certain previous closed itemsets once, it incrementally updates the current sliding window that makes the closed frequent itemsets can be output in real time based on any user’s specified thresholds. Theoretical analysis and the experimental results of the real datasets Mushroom、Retail-chain and artificially synthesized datasets T40I10D100K show that the proposed method is superior to that of state-of-the-art algorithm Moment and CFI-Stream in terms of time and space efficiency, and has good scalability as the number of transactions processed increases and adapts very rapidly to the change in data streams.

Abstract:In this paper, we introduce a new method for noise estimation. The noise estimate is given by averaging past spectral power values and using a smoothing factor that is adjusted by the signal presence probability in every bins. Presence of speech is determined by the ratio between the energy of the smoothed noisy speech and its minimum. Objective and subjective evaluation confirm superiority in quality of the enhanced speech.

Abstract:Multiple input multiple output(MIMO) radar as a new radar system has received significant attention in the past few years. This article studies space-time adaptive processing(STAP) for airborne MIMO radar, Based on multistage winner filter with generalized sidelobe canceller structure, A new reduced-rank space-time adaptive processing algorithm using priori knowledge constraints is proposed for MIMO radar. Through the use of jamming direction knowledge and clutter subspace knowledge estimated by prolate spheroridal wave functions ,the algorithm can greatly reduce computation and sample needs of airborne MIMO radar STAP and at the same time keep the performance of clutter suppression. Also, knowledge-aided effect on the convergence performance when the knowledge is mis-matched is considered. Simulation results show that when there is a certain error of knowledge , the algorithm can still effectively improve the convergence performance of STAP algorithm.

Abstract:In order to solve conflicts in suppressing noise and detecting edge. A new edge detection algorithm based on non-subsampled multi-scale product is presented according to Mallat multi-scale analysis, which uses Gaussian function and its first-derivative as low-pass and high-pass filter to enhance edge and suppress noise, then detect edge by gradient direction and updating search method. The experiments show that this approach has advantages of detecting edge in image embedded noise and high signal to noise ratio.

Abstract:Plane transformation technique is an effective algorithm in deinterleaving pulses, the key technique contains two aspects: characteristic curve searching and recognizing.To characteristic curve searching, the traditional way which depends on manpower takes time and energy ,also limits automatic operation; To characteristic curve recognizing, there is only initial description and no executive solution. Considering the weakness above, pulse points’ features are studied, automatic algorithms are designed for selecting characteristic curves, recognizing rules are presented for different types. The searching algorithm, which does not rely on intial pulse’s selection, is applicable to line and curve, the recognizing algorithm is practical and effective, finally, automatic deinterleaving pulses is realized. Simulations testify the efficiency.

Abstract:Due to data sparsity, discriminative training has not been successfully applied in the system of vocal password by now. This paper proposed a novel vocal password framework based on a specific feature processing strategy, in which the new feature was used to represent the distance measure and the problem caused by data sparsity could be solved to some extent. As a result, the vocal password was actually transferred from verification to binary classification and the discriminative training of these two models was successfully accomplished on the minimum classification error criteria. After fusing this discriminative system with the GMM-UBM system, the EER performance decreased to 4.48%, relatively 17.95% and 59.68% lower than the GMM-UBM and the DTW system respectively on the corpus including 60 speakers. This result shows us the new application of discriminative training in the vocal password system is feasible and effective.

Abstract:In this paper, the model of multi-target based on the statistical multiple-input multiple-output (MIMO) radar is analyzed, and the CRLB of multi-target localization estimation in this model is also present. The optimal rule is improved by the thesis of diversity gain and mutual interference based on the CRLB by contrasting the optimal antenna placement within the signal target case. It is shown that the existing inevitability of optimal antenna placement is demonstrated, and the non-symmetrical but horizontal placement is optimal. The performance of the proposed estimator is validated by simulation results.

Abstract:This paper proposes a new method describing human behavior in crowded scenes based on STIPs (Spatial Temporal Interesting Points). By comparing three different methods for STIPs extraction, which are Harris corner, Gabor wavelet and Hessian matrix, The scale-invariant extraction method which based on Hessian matrix is chosen in the paper. Histogram of gradient, histogram of optical flow orientation and spatial-temporal Haar feature are used to build descriptors for STIPs. Then bag-of-words model is used in normal behavior modeling. GMM based on EM estimation is introduced to produce keywords. Then each video of normal action is divided into several clips and they are described in probability vectors using keywords. All vectors construct normal behavior codebook. In testing phase, through calculating the similarity distance between the coding vector of the test sample and that of the normal, abnormal behavior can be detected when the distance exceeds the threshold. The algorithm is tested in UMN and UCF datasets, the experiments show that the proposed algorithm has effective identification for group abnormal behavior, and it has good adaptability against scale variance and illumination changing.