In this paper, we propose a method which enables us to control the variance of the coefficients of the LMS-type adaptive filters. In the method, each coefficient of the adaptive filter is modeled as an random variable with a Gaussian distribution, and its value is estimated as the mean value of the distribution. Besides, at each time, we check if the updated value exists within the predefined range of distribution. The update of a coefficient will be canceled when its updated value exceeds the range. We propose an implementation method which has similar formula as the Gaussian mixture model (GMM) widely used in signal processing and machine learning. The effectiveness of the proposed method is evaluated by the computer simulations.

Gaussian integer has a potential to enhance the safety of elliptic curve cryptography (ECC) on system under the condition fixing bit length of integral and floating point types, in viewpoint of the order of a finite field. However, there seems to have been no algorithm which makes Gaussian integer ECC safer under the condition. We present the algorithm to enhance the safety of ECC under the condition. Then, we confirm our Gaussian integer ECC is safer in viewpoint of the order of finite field than rational integer ECC or Gaussian integer ECC of naive methods under the condition.

We propose a speech analysis-synthesis and deep neural network (DNN)-based text-to-speech (TTS) synthesis framework using Gaussian mixture model (GMM)-based approximation of full-band spectral envelopes. GMMs have excellent properties as acoustic features in statistic parametric speech synthesis. Each Gaussian function of a GMM fits the local resonance of the spectrum. The GMM retains the fine spectral envelope and achieve high controllability of the structure. However, since conventional speech analysis methods (i.e., GMM parameter estimation) have been formulated for a narrow-band speech, they degrade the quality of synthetic speech. Moreover, a DNN-based TTS synthesis method using GMM-based approximation has not been formulated in spite of its excellent expressive ability. Therefore, we employ peak-picking-based initialization for full-band speech analysis to provide better initialization for iterative estimation of the GMM parameters. We introduce not only prediction error of GMM parameters but also reconstruction error of the spectral envelopes as objective criteria for training DNN. Furthermore, we propose a method for multi-task learning based on minimizing these errors simultaneously. We also propose a post-filter based on variance scaling of the GMM for our framework to enhance synthetic speech. Experimental results from evaluating our framework indicated that 1) the initialization method of our framework outperformed the conventional one in the quality of analysis-synthesized speech; 2) introducing the reconstruction error in DNN training significantly improved the synthetic speech; 3) our variance-scaling-based post-filter further improved the synthetic speech.

This paper describes a novel approach to flexible control of speaker characteristics using tensor representation of multiple Gaussian mixture models (GMM). In voice conversion studies, realization of conversion from/to an arbitrary speaker's voice is one of the important objectives. For this purpose, eigenvoice conversion (EVC) based on an eigenvoice GMM (EV-GMM) was proposed. In the EVC, a speaker space is constructed based on GMM supervectors which are high-dimensional vectors derived by concatenating the mean vectors of each of the speaker GMMs. In the speaker space, each speaker is represented by a small number of weight parameters of eigen-supervectors. In this paper, we revisit construction of the speaker space by introducing the tensor factor analysis of training data set. In our approach, each speaker is represented as a matrix of which the row and the column respectively correspond to the dimension of the mean vector and the Gaussian component. The speaker space is derived by the tensor factor analysis of the set of the matrices. Our approach can solve an inherent problem of supervector representation, and it improves the performance of voice conversion. In addition, in this paper, effects of speaker adaptive training before factorization are also investigated. Experimental results of one-to-many voice conversion demonstrate the effectiveness of the proposed approach.

As one of the most popular intelligent optimization algorithms, Simulated Annealing (SA) faces two key problems, the generation of perturbation solutions and the control strategy of the outer loop (cooling schedule). In this paper, we introduce the Gaussian Cloud model to solve both problems and propose a novel cloud annealing algorithm. Its basic idea is to use the Gaussian Cloud model with decreasing numerical character He (Hyper-entropy) to generate new solutions in the inner loop, while He essentially indicates a heuristic control strategy to combine global random search of the outer loop and local tuning search of the inner loop. Experimental results in function optimization problems (i.e. single-peak, multi-peak and high dimensional functions) show that, compared with the simple SA algorithm, the proposed cloud annealing algorithm will lead to significant improvement on convergence and the average value of obtained solutions is usually closer to the optimal solution.

In this paper, we propose an adaptive fusion successive cancellation list decoder (ADF-SCL) for polar codes with single cyclic redundancy check. The proposed ADF-SCL decoder reasonably avoids unnecessary calculations by selecting the successive cancellation (SC) decoder or the adaptive successive cancellation list (AD-SCL) decoder depending on a log-likelihood ratio (LLR) threshold in the decoding process. Simulation results show that compared to the AD-SCL decoder, the proposed decoder can achieve significant reduction of the average complexity in the low signal-to-noise ratio (SNR) region without degradation of the performance. When Lmax=32 and Eb/N0=0.5dB, the average complexity of the proposed decoder is 14.23% lower than that of the AD-SCL decoder.

In order to improve the noise robustness of automatic speaker recognition, many techniques on speech/feature enhancement have been explored by using deep neural networks (DNN). In this work, a DNN multi-level enhancement (DNN-ME), which consists of the stages of signal enhancement, cepstrum enhancement and i-vector enhancement, is proposed for text-independent speaker recognition. Given the fact that these enhancement methods are applied in different stages of the speaker recognition pipeline, it is worth exploring the complementary role of these methods, which benefits the understanding of the pros and cons of the enhancements of different stages. In order to use the capabilities of DNN-ME as much as possible, two kinds of methods called Cascaded DNN-ME and joint input of DNNs are studied. Weighted Gaussian mixture models (WGMMs) proposed in our previous work is also applied to further improve the model's performance. Experiments conducted on the Speakers in the Wild (SITW) database have shown that DNN-ME demonstrated significant superiority over the systems with only a single enhancement for noise robust speaker recognition. Compared with the i-vector baseline, the equal error rate (EER) was reduced from 5.75 to 4.01.

Spectrum sensing is a fundamental requirement for cognitive radio, and it is a challenging problem in impulsive noise modeled by symmetric alpha-stable (SαS) distributions. The Gaussian kernelized energy detector (GKED) performs better than the conventional detectors in SαS distributed noise. However, it fails to detect the DC signal and has high computational complexity. To solve these problems, this paper proposes a more efficient and robust detector based on a Gaussian function (GF). The analytical expressions of the detection and false alarm probabilities are derived and the best parameter for the statistic is calculated. Theoretical analysis and simulation results show that the proposed GF detector has much lower computational complexity than the GKED method, and it can successfully detect the DC signal. In addition, the GF detector performs better than the conventional counterparts including the GKED detector in SαS distributed noise with different characteristic exponents. Finally, we discuss the reason why the GF detector outperforms the conventional counterparts.

This paper proposes a new design criterion of adaptively scaled belief (ASB) in Gaussian belief propagation (GaBP) for large multi-user multi-input multi-output (MU-MIMO) detection. In practical MU detection (MUD) scenarios, the most vital issue for improving the convergence property of GaBP iterative detection is how to deal with belief outliers in each iteration. Such outliers are caused by modeling errors due to the fact that the law of large number does not work well when it is difficult to satisfy the large system limit. One of the simplest ways to mitigate the harmful impact of outliers is belief scaling. A typical approach for determining the scaling parameter for the belief is to create a look-up table (LUT) based on the received signal-to-noise ratio (SNR) through computer simulations. However, the instantaneous SNR differs among beliefs because the MIMO channels in the MUD problem are random; hence, the creation of LUT is infeasible. To stabilize the dynamics of the random MIMO channels, we propose a new transmission block based criterion that adapts belief scaling to the instantaneous channel state. Finally, we verify the validity of ASB in terms of the suppression of the bit error rate (BER) floor.

In this letter, a construction of asymmetric Gaussian integer zero correlation zone (ZCZ) sequence sets is presented based on interleaving and filtering. The proposed approach can provide optimal or almost optimal single Gaussian integer ZCZ sequence sets. In addition, arbitrary two sequences from different sets have inter-set zero cross-correlation zone (ZCCZ). The resultant sequence sets can be used in the multi-cell QS-CDMA system to reduce the inter-cell interference and increase the transmission data.

We propose a novel bit error rate (BER) analysis model of weighted-type fractional Fourier transform (WFRFT)-based systems with WFRFT order offset Δα. By using the traditional BPSK BER analysis method, we deduce the equivalent signal noise ratio (SNR), model the interference in the channel as a Gaussian noise with non-zero mean, and provide a theoretical BER expression of the proposed system. Simulation results show that its theoretical BER performance well matches the empirical performance, which demonstrates that the theoretical BER analysis proposed in this paper is reliable.

In the past two decades, many generalized cyclotomic sequences have been constructed and they have been used in cryptography and communication systems for their high linear complexity and low autocorrelation. But there are a few of papers focusing on the 2-adic complexities of such sequences. In this paper, we first give a property of a class of Gaussian periods based on Whiteman's generalized cyclotomic classes of order 4. Then, as an application of this property, we study the 2-adic complexity of a class of Whiteman's generalized cyclotomic sequences constructed from two distinct primes p and q. We prove that the 2-adic complexity of this class of sequences of period pq is lower bounded by pq-p-q-1. This lower bound is at least greater than one half of its period and thus it shows that this class of sequences can resist against the rational approximation algorithm (RAA) attack.

Discrete Gaussian is a cornerstone of many lattice-based cryptographic constructions. Aiming at the orthogonal lattice of a vector, we propose a discrete Gaussian rejection sampling algorithm, by modifying the dynamic programming process for subset sum problems. Within O(nq2) time, our algorithm generates a distribution statistically indistinguishable from discrete Gaussian at width s>ω(log n). Moreover, we apply our sampling algorithm to general high-dimensional dense lattices, and orthogonal lattices of matrices $matAinZ_q^{O(1) imes n}$. Compared with previous polynomial-time discrete Gaussian samplers, our algorithm does not rely on the short basis.

This paper reports a novel 3D topology optimization method based on the finite difference time domain (FDTD) method for a dielectric lens antenna. To obtain an optimal lens with smooth boundary, we apply normalized Gaussian networks (NGnet) to 3D topology optimization. Using the proposed method, the dielectric lens with desired radiation characteristics can be designed. As an example of the optimization using the proposed method, the width of the main beam is minimized assuming spatial symmetry. In the optimization, the lens is assumed to be loaded on the aperture of a waveguide slot antenna and is smaller compared with the wavelength. It is shown that the optimized lens has narrower beamwidth of the main beam than that of the conventional lens.

The quality of edge detection is related to detection angle, scale, and threshold. There have been many algorithms to promote edge detection quality by some rules about detection angles. However these algorithm did not form rules to detect edges at an arbitrary angle, therefore they just used different number of angles and did not indicate optimized number of angles. In this paper, a novel edge detection algorithm is proposed to detect edges at arbitrary angles and optimized number of angles in the algorithm is introduced. The algorithm combines singularity detection with Gaussian wavelet transform and edge detection at arbitrary directions and contain five steps: 1) An image is divided into some pixel lines at certain angle in the range from 45° to 90° according to decomposition rules of this paper. 2) Singularities of pixel lines are detected and form an edge image at the certain angle. 3) Many edge images at different angles form a final edge images. 4) Detection angles in the range from 45° to 90° are extended to range from 0° to 360°. 5) Optimized number of angles for the algorithm is proposed. Then the algorithm with optimized number of angles shows better performances.

In this paper, we consider Bayesian image denoising based on a Gaussian Markov random field (GMRF) model, for which we propose an new algorithm. Our method can solve Bayesian image denoising problems, including hyperparameter estimation, in O(n)-time, where n is the number of pixels in a given image. From the perspective of the order of the computational time, this is a state-of-the-art algorithm for the present problem setting. Moreover, the results of our numerical experiments we show our method is in fact effective in practice.

The concept of Gaussian integer sequence pair is generalized from a single Gaussian integer sequence. In this letter, by adopting cyclic difference set pairs, a new construction method for perfect Gaussian integer sequence pairs is presented. Furthermore, the necessary and sufficient conditions for constructing perfect Gaussian integer sequence pairs are given. Through the research in this paper, a large number of perfect Gaussian integer sequence pairs can be obtained, which can greatly extend the existence of perfect sequence pairs.

This work presents a hybrid formulation of the stochastic reduced order model (SROM) algorithm, which makes use of Gauss quadrature, a key ingredient of the stochastic collocation method, to avoid the cumbersome optimization process required by SROM for optimal extraction of the sample set. With respect to classic SROM algorithms, the proposed formulation allows a significant reduction in computation time and burden as well as a remarkable improvement in the accuracy and convergence rate in the estimation of statistical moments. The method is here applied to a specific case study, that is the prediction of crosstalk in a two-conductor wiring structure with electrical and geometrical parameters not perfectly known. Both univariate and multivariate analyses are carried out, with the final objective being to compare the performance of the two SROM formulations with respected to Monte Carlo simulations.

This paper studies a simultaneous wireless information and power transfer (SWIPT) system in which the transmitter not only sends data and energy to many types of wireless users, such as multiple information decoding users, multiple hybrid power-splitting users (i.e., users with a power-splitting structure to receive both information and energy), and multiple energy harvesting users, but also prevents information from being intercepted by a passive eavesdropper. The transmitter is equipped with multiple antennas, whereas all users and the eavesdropper are assumed to be equipped with a single antenna. Since the transmitter does not have any channel state information (CSI) about the eavesdropper, artificial noise (AN) power is maximized to mask information as well as to interfere with the eavesdropper as much as possible. The non-convex optimization problem is formulated to minimize the transmit power satisfying all signal-to-interference-plus-noise (SINR) and harvested energy requirements for all users so that the remaining power for generating AN is maximized. With perfect CSI, a semidefinite relaxation (SDR) technique is applied, and the optimal solution is proven to be tight. With imperfect CSI, SDR and a Gaussian randomization algorithm are proposed to find the suboptimal solution. Finally, numerical performance with respect to the maximum SINR at the eavesdropper is determined by a Monte-Carlo simulation to compare the proposed AN scenario with a no-AN scenario, as well as to compare perfect CSI with imperfect CSI.

Feature visualization is of great significances in volume visualization, and feature extraction has been becoming extremely popular in feature visualization. While precise definition of features is usually absent which makes the extraction difficult. This paper employs probability density function (PDF) as statistical property, and proposes a statistical property guided approach to extract features for volume data. Basing on feature matching, it combines simple liner iterative cluster (SLIC) with Gaussian mixture model (GMM), and could do extraction without accurate feature definition. Further, GMM is paired with a normality test to reduce time cost and storage requirement. We demonstrate its applicability and superiority by successfully applying it on homogeneous and non-homogeneous features.