Search

Refined optimization of complex curve–linear-shaped radiators, such as traveling-wave Vivaldi antennas, can be achieved by considering simulated near fields to interpret in detail the structural influences of a design. The relationships between the space and time distributions of electromagnetic (EM) energy clusters and the geometric features are revealed with appropriate use of impulse response analysis combined with the multiple signal classification (MUSIC) algorithm. This article reports a deeper approach when applied to the adjustment of the geometric features of a traveling-wave antenna based on an analysis of near-field propagation features.

This systematic review aimed to identify how tele-mentoring systems that incorporate augmented reality (AR) technology are being used in healthcare environments. A total of 12 electronic bibliographic databases were searched using the terms “augmented reality”, “tele-mentoring” and “health”. The PRISMA checklist was used as a guide for reporting. The mixed method appraisal tool was used to assess the quality of the included experiments. The data were then analysed using a concept-centric approach and categorised primarily with regards to system performance and task performance measures. A total of 11 randomised controlled trials and 14 non-randomised designs were included for review. Both mentees and mentors assessed the system and task performance according to 25 categories. The feedback of mentees using AR tele-mentoring systems was generally positive. The majority of experiments revealed that the AR system was an effective tele-mentoring device overall and resulted in the effective performance of a procedure. Benefits included improvements in trainees’ confidence, task completion time and reductions in task errors and shifts in focus. However, the systems had limitations, including heaviness of the equipment, inconvenience, discomfort and distraction of wearing devices, limited battery life, the latency of video and audio signals and limited field of view. Implications for practice or policy: Health practitioners can apply AR technology to receive and follow real-time annotated instructions verbally and visually from remote experts. Technical developers may consider improving AR devices in terms of lighter weight, larger field of view, more ergonomic design, more stable network connection and longer battery life. Further AR-related experiments may need to explore AR tele-mentoring systems’ utility across healthcare environments with larger samples, real patient populations in remote settings, cost-benefit analysis and impacts on short- and long-term patient outcomes.

The present paper investigates the trade-offs between the energy efficiency (EE) and spectral efficiency (SE) in the full-duplex (FD) multiuser multi-input multioutput (MU-MIMO) cloud radio access networks (CRANs) with simultaneous wireless information and power transfer (SWIPT). In the considered network, the central unit (CU) intends to concurrently not only transfer both energy and information toward downlink (DL) users using power splitting structures but also receive signals from uplink (UL) users. This communication is executed via FD radio units (RUs) which are distributed nearby users and connected to the CU through limited capacity fronthaul (FH) links. In order to unveil interesting trade-offs between the EE and SE metrics, we first introduce three conventional single-objective optimization problems (SOOPs) including (i) system sum rate maximization, (ii) total power minimization, and (iii) fractional energy efficiency maximization. Then, by making use of the multiobjective optimization (MOO) framework, the MOO problem (MOOP) with the objective vector of the achievable rate and power consumption is addressed. All considered problems are nonconvex with respect to designing variables comprising precoding matrices, compression matrices, and DL power splitting factors; thus, it is extremely intractable to solve these problems directly. To overcome these issues, we develop iterative algorithms by utilizing the sequential convex approximation (SCA) approach for the first two SOO problems and the SCA-based Dinkelbach method for the fractional EE problem. Regarding the MOOP, we first rewrite it as an SOOP by applying the modified weighted Tchebycheff method and, then, propose the iterative algorithm-based SCA to find its optimal Pareto set. Various numerical simulations are conducted to study the system performance and appealing EE-SE trade-offs in the considered system.

This paper considers simultaneous wireless information and power transfer enabled full-duplex multi-user multiple-input multiple-output cognitive networks. By taking into account imperfect channel state information (CSI) of the links toward primary users (PUs), this paper aims to simultaneously optimize two design objectives, namely the achievable sum-rate and sum harvested energy in the secondary network. Hence, the design problem is modelled as a multi-objective optimization problem (MOOP) subject to the transmit power constraint at the base station and robust harmful interference constraints at the PUs. Then, to find the Pareto set, the MOOP is rewritten to a single-objective problem (SOOP) by using the modified weighted Tchebycheff method. However, it is mathematically difficult to solve the transformed SOOP due to the non-concavity of the sum-rate function and the semi-infinite nature of the robust interference constraints. To overcome this challenge, we use the difference of two convex functions (DC) technique and S-procedure theory to recast the design optimization problem as the sequential convex programming. Various numerical simulations are conducted to illustrate the Pareto optimal solution sets, the robustness of interference constraints against CSI uncertainties, and trade-offs between the achievable sum-rate and harvested energy.

This paper studies the energy efficiency (EE) optimization problem in multiple-user multiple-input–multiple-output heterogeneous wireless powered communication network (HWPCN) in which the users in small cells can harvest energy from the small cell base stations in downlink and use harvested energy for uplink transmission. We consider the joint design problem of wireless energy transmission and wireless information transmission precoding matrices subject to the transmit power constraints at each BS and each user. Since the optimization problem under consideration is highly nonlinear non-convex fractional programming in design variables, it is mathematically challenging to obtain the optimal solutions. To tackle with this difficulty, we employ the difference of convex programming to obtain the concave lower bound of the achievable sum rates. Then, we apply the Dinkelbach approach to develop an efficient iterative algorithm in which the convex optimization problems are solved. Simulation results are provided to investigate the EE of the proposed algorithm in HWPCNs.

The rapid development of the Internet and the wide spread of its applications has affected many aspects of our life. However, this development also makes the cyberspace more vulnerable to various attacks. Thus, detecting and preventing these attacks are crucial for the next development of the Internet and its services. Recently, machine learning methods have been widely adopted in detecting network attacks. Among many machine learning methods, AutoEncoders (AEs) are known as the state-of-the-art techniques for network anomaly detection. Although, AEs have been successfully applied to detect many types of attacks, it is often unable to detect some difficult attacks that attempt to mimic the normal network traffic. In order to handle this issue, we propose a new model based on AutoEncoder called Double-Shrink AutoEncoder (DSAE). DSAE put more shrinkage on the normal data in the middle hidden layer. This helps to pull out some anomalies that are very similar to normal data. DSAE are evaluated on six well-known network attacks datasets. The experimental results show that our model performs competitively to the state-of-the-art model, and often out-performs this model on the attacks group that is difficult for the previous methods.

In this paper, we introduce a Deep Neural Network (DNN) to maximize the Proportional Fairness (PF) of the Spectral Efficiency (SE) of uplinks in Cell-Free (CF) massive Multiple-Input Multiple-Output (MIMO) systems. The problem of maximizing the PF of the SE is a non-convex optimization problem in the design variables. We will develop a DNN which takes pilot sequences and large-scale fading coefficients of the users as inputs and produces the outputs of optimal transmit powers. By consisting of densely residual connections between layers, the proposed DNN can efficiently exploit the hierarchical features of the input and motivates the feed-forward nature of DNN architecture. Experimental results showed that, compared to the conventional iterative optimization algorithm, the proposed DNN has excessively lower computational complexity with the trade-off of approximately only 1% loss in the sum rate and the fairness performance. This demonstrated that our proposed DNN is reasonably suitable for real-time signal processing in CF massive MIMO systems.

Recently, fault-aware retraining has emerged as a promising approach to improve the error resilience of Deep Neu-ral Networks (DNNs) against manufacturing-induced defects in DNN accelerators. However, state-of-the-art fault-aware training techniques incur a gigantic retraining overhead due to their per-chip retraining nature for the chip’s unique fault map, which may render it practically infeasible if retraining is done on large datasets. To address this major limitation and improve the practicability of the fault-aware retraining methodology, this work proposes a novel concept of merging fault maps to effectively retrain a DNN for a group of faulty chips in a single fault-aware retraining round. The merging of fault maps enables to avoid per chip retraining and thereby reduces the retraining overhead significantly. However, the merging of fault maps brings in new challenges such as training divergence (accuracy collapse) if a high number of accumulated faults are injected into the network in the first epoch. To address these challenges, we propose a methodology for effective merging of fault maps and then retraining of DNNs. Experimental results show that our methodology offers at least 1.4x retraining speedup on average while improving the error resilience of the network (depending on the DNN models and the number of merged fault maps). For example, for the Resnet-32 model using fault map generated from 5 fault maps at the fault rate 6e-3, our methodology offers 2x retraining speedup and 0.6% classification accuracy drop against per-chip retraining.

In this paper, a tracking-error-based control design without feedback passivation stage is developed for the stabilization of physical and chemical nonlinear processes. To achieve control objectives, the system dynamics is first formulated under appropriate conditions as a relaxing (pseudo) port-Hamiltonian (PH) representation with some quadratic storage function, in which the positive semi-definite property of damping matrix may not be taken necessarily into account. Then, a reference trajectory expressed by a certain structure passing through a desired set-point (or containing an optimal profile) is suitably chosen. By adding a relevant damping injection, asymptotic and global convergence of error dynamics is guaranteed. Two case studies including the level control of a four-tank process of continuous time type and the optimal tracking of a batch polymerization reactor of discontinuous time type due to the nature of the process operation are used to illustrate the application and the effectiveness of the approach. Besides the control performance comparison with the conventional passivity-based control method and the robustness evaluation against disturbance and/or noise are included.

The paper focuses on designing precoding matrices in multi-cell multiple-input multiple-output (MIMO) simultaneous wireless information and power transfer networks (SWIPT) where the sets of users are selected for data transmission in each time slot and the unselected users are dedicated to energy harvesting. The precoding design for the SWIPT problem is formulated as a general multi-objective maximization problem, in which the sum-rate (SR) and sum harvested energy (SHE) are maximized simultaneously under the transmit power constraints. Since the objective function of the maximization problem is not concave in the design matrix variables, it is difficult to directly obtain the optimal solutions. To tackle this challenge, we recast the SR function into one more amenable by applying the connection between the minimum mean square error and achievable data rate. In addition, to deal with the non-concavity of the harvested energy function, we derive its concave minorant. Then, we develop an efficient iterative algorithm based on alternating optimization (AO) to obtain the optimal precoders. We also analyze the convergence and computational complexity of the proposed algorithm. Finally, by numerical simulation results we investigate the trade-offs between the SR and SHE.

This work proposes an advanced control strategy for an open non-isothermal homogeneous reactor by integrating alternative model of the system dynamics based on variant and invariant states with the tracking-error-based control method in the framework of port-Hamiltonian representation. More precisely, a linear transformation is used to transform the original dynamics into a new model, where the enthalpy, reaction variant and reaction invariant are the state variables. Then, the transformed model is formulated into the port-Hamiltonian structure before the tracking-error-based approach is applied to design the controllers for the purpose of stabilization. The numerical simulations for the first order reaction modelled with the continuous stirred tank reactor having the multiplicity behavior subsequently show that the proposed feedback law is able to stabilize the system at the desirable equilibrium point.

In this paper, a feature extraction method based on neural networks for Vietnamese text classification is developed. The Vietnamese online news documents are initially preprocessed to transform to lower-case documents as well as remove punctuation and stop-words. Then, the tokenization applying the combination of uni-gram and bi-gram models is conducted to generate a list of tokens for each document. As a result, polyphonic linguistic problems in Vietnamese language are diminished. The statistical term frequency - inverse document frequency model is employed to represent the lists of tokens as real value vectors. Instead of applying conventional feature selection algorithms, neural networks are conducted for dimensionality reduction. Therefore, not only the size of the term frequency - inverse document frequency vectors are reduced, but also distinctive feature vectors are created for text classification tasks. Support vector machines are utilized in the classification step. The results attained in the experiment have shown that our work using neural networks for the feature extraction outperforms other traditional methods.

This paper is concerned with the proportional fairness (PF) of the spectral efficiency (SE) maximization of uplinks in a cell-free (CF) massive multiple-input multiple-output (MIMO) system in which a large number of single-antenna access points (APs) connected to a central processing unit (CPU) serve many single-antenna users. To detect the user signals, the APs use matched filters based on the local channel state information while the CPU deploys receiver filters based on knowledge of channel statistics. We devise the maximization problem of the SE PF, which maximizes the sum of the logarithm of the achievable user rates, as a jointly nonconvex optimization problem of receiver filter coefficients and user power allocation subject to user power constraints. To handle the challenges associated with the nonconvexity of the formulated design problem, we develop an iterative algorithm by alternatively finding optimal filter coefficients at the CPU and transmit powers at the users. While the filter coefficient design is formulated as a generalized eigenvalue problem, the power allocation problem is addressed by a gradient projection (GP) approach. Simulation results show that the SE PF maximization not only offers approximately the achievable sum rates as compared to the sum-rate maximization but also provides an improved trade-off between the user rate fairness and the achievable sum rate., Comment: The article was published in International Journal of Communication Systems, volume 33, issue 14; 25 September 2020; https://doi.org/10.1002/dac.4534

This paper proposes a blended model that is suitable for recognizing off-line handwritten accented characters in general and Vietnamese characters in particular. The recognition model linearly combines four extracting methods in the feature extraction period, including Zones density, Projection histogram, Contour profiles, and Haar wavelets. The set of features obtained will be applied with Principal Component Analysis (PCA) to retain useful features, reducing the recognition time. Additionally, a Support Vector Machine (SVM) is also utilised for training and recognition. The proposed model is tested on the dataset of 21174 samples with 99 Vietnamese off-line handwritten accented characters.

International audience; Sentiment analysis or opinion mining used to understand the community's opinions on a particular product. Sentiment analysis involves building the opinion collection and classification system. Aspect-based sentiment analysis focuses on the ability to extract and summarize opinions on specific aspects of entities within sentiment document. In this paper, we propose a novel supervised learning approach using deep learning techniques for multitask aspect-based opinion mining system that support four main subtasks: extract opinion target, classify aspect-entity (category), and estimate opinion polarity (positive, neutral, negative) on each extracted aspect of entity. Using extra POS layer to identify morphological features of words combines with stacking architecture of BiLSTM and CNN with word embeddings achieved by training GloVe on Restaurant domain reviews of the SemEval 2016 benchmark dataset in our proposed method is aimed at increasing the accuracy of the model. Experimental results showed that our multitask aspect-based sentiment analysis model has extracted and classified main above subtasks concurrently and achieved significantly better accuracy than the state-of-the-art methods.

Deep Neural Networks (DNNs) - the state-of-the-art computational models for many Artificial Intelligence (AI) applications - are inherently compute and resource-intensive and, hence, cannot exploit traditional redundancy-based fault mitigation techniques for enhancing the dependability of DNN-based systems. Therefore, there is a dire need to search for alternate methods that can improve their reliability without high expenditure of resources by exploiting the intrinsic characteristics of these networks. In this paper, we present cross-layer approaches that, based on the intrinsic characteristics of DNNs, employ software and hardware-level modifications for improving the resilience of DNN-based systems to hardware-level faults, e.g., soft errors and permanent faults.

A calibration method for a high-resolution hybrid MIMO turntable radar imaging system is presented. A line of small metal spheres is employed as a test pattern in the calibration process to measure the position shift caused by undesired antenna effects. The unwanted effects in the antenna near-field responses are analysed, modelled and significantly mitigated based on the symmetry and differences in the responses of the MIMO configuration.

Deep Neural Networks (DNNs) are widely being adopted for safety-critical applications, e.g., healthcare and autonomous driving. Inherently, they are considered to be highly error-tolerant. However, recent studies have shown that hardware faults that impact the parameters of a DNN (e.g., weights) can have drastic impacts on its classification accuracy. In this paper, we perform a comprehensive error resilience analysis of DNNs subjected to hardware faults (e.g., permanent faults) in the weight memory. The outcome of this analysis is leveraged to propose a novel error mitigation technique which squashes the high-intensity faulty activation values to alleviate their impact. We achieve this by replacing the unbounded activation functions with their clipped versions. We also present a method to systematically define the clipping values of the activation functions that result in increased resilience of the networks against faults. We evaluate our technique on the AlexNet and the VGG-16 DNNs trained for the CIFAR-10 dataset. The experimental results show that our mitigation technique significantly improves the resilience of the DNNs to faults. For example, the proposed technique offers on average 68.92% improvement in the classification accuracy of resilience-optimized VGG-16 model at 1e-5 fault rate, when compared to the base network without any fault mitigation., The 23rd Design, Automation and Test in Europe (DATE 2020)

This work proposes a passivity-based approach to deal with the output-tracking-error problem for a large class of nonlinear chemical processes including non-minimum phase systems. More precisely, in that framework, the system dynamics is firstly written into the relaxing (pseudo) port-Hamiltonian representation which does not necessarily require the positive semi-definite property of the damping matrix. Then, a reference trajectory associated with a certain structure passing through a desired equilibrium point (i.e., the set-point) is chosen so that the error dynamics can be globally asymptotically stabilized at the origin thanks to the assignment of an appropriate damping injection. This method is subsequently illustrated for a benchmark of multiple reactions systems, namely Van de Vusse reaction system. The numerical simulations show the applications of the proposed approach.

In this paper, we investigate the downlink multiple-input-multiple-output (MIMO) broadcast channels in which a base transceiver station (BTS) broadcasts multiple data streams to K MIMO mobile stations (MSs) simultaneously. In order to maximize the weighted sum-rate (WSR) of the system subject to the transmitted power constraint, the design problem is to find the pre-coding matrices at BTS and the decoding matrices at MSs. However, such a design problem is typically a nonlinear and nonconvex optimization and, thus, it is quite hard to obtain the analytical solutions. To tackle with the mathematical difficulties, we propose an efficient stochastic optimization algorithm to optimize the transceiver matrices. Specifically, we utilize the linear minimum mean square error Wiener filters at MSs. Then, we introduce the constrained particle swarm optimization algorithm to jointly optimize the precoding and decoding matrices. Numerical experiments are exhibited to validate the effectiveness of the proposed algorithm in terms of convergence, computational complexity and total WSR.

This paper studies the transceiver design for multiuser multiple-input multiple-output cognitive radio networks. Different from the conventional methods which aim at maximizing the spectral efficiency, this paper focuses on maximizing the energy efficiency (EE) of the network. First, we formulate the precoding and decoding matrix designs as optimization problems which maximize the EE of the network subject to per-user power and interference constraints. With a higher priority in accessing the spectrum, the primary users (PUs) can design their transmission strategies without awareness of the secondary user (SU) performance. Thus, we apply a full interference alignment technique to eliminate interference between the PUs. Then, the EE maximization problem for the primary network can be reformulated as a tractable concave-convex fractional program which can be solved by the Dinkelbach method. On the other hand, the uncoordinated interference from the PUs to the SUs cannot be completely eliminated due to a limited coordination between the PUs with the SUs. The secondary transceivers are designed to optimize the EE while enforcing zero-interference to the PUs. Since the EE maximization for the secondary network is an intractable fractional programming problem, we develop an iterative algorithm with provable convergence by invoking the difference of convex functions programming along with the Dinkelbach method. In addition, we also derive closed-form expressions for the solutions in each iteration to gain insights into the structures of the optimal transceivers. The simulation results demonstrate that our proposed method outperforms the conventional approaches in terms of the EE.

Estimating unknown parameters in ethylene polymerization reactor using hybrid Fuzzy-SMO.Display Omitted Novel hybrid observer method combining fuzzy logic with sliding mode observer.The observer is able to estimate several parameters without redesigning the structure of the whole observer.The design is simple and easy to formulate.Provides fast and accurate estimation results in estimating the ethylene, butene and melt index in a polyethylene reactor. In this work, we proposed a novel hybrid fuzzy-sliding mode observer designed in such a manner that it can be utilized to estimate various parameters by simply using the related process model, without redesigning the structure of the whole observer. The performances and effectiveness of this hybrid observer are shown through numerical simulation based on a case study involving an ethylene polymerization process to estimate the ethylene and butene concentrations in the reactor as well as the melt flow index. It can be concluded that the proposed hybrid observer provides fast estimation with a high rate of accuracy even in the presence of disturbances and noise in the model. This hybrid observer is also compared with the sliding mode, extended Luenberger and proportional sliding mode observers to highlight its effectiveness and advantages over these observers.

In this paper, we consider the mobile edge offloading scenario consisting of one mobile device (MD) with multiple independent tasks and various remote edge devices. In order to save energy, the user's device can offload the tasks to available access points for edge computing. Data compression is applied to reduce offloaded data size prior to wireless transmission to minimize the execution latency. The problem of jointly optimizing the task allocation decision and the data compression ratio to minimize the total tasks' execution latency and the MD's energy consumption concurrently is proposed. We show that the design problem is a non-convex optimization one but it can be transformed into a convex one through a semidefinite relaxation (SDR) based approach. Numerical simulations demonstrate the outperformance of the proposed scheme compared to the benchmark one., ISEE 2019

In this paper, we present a deep learning framework applied for acoustic scene classification (ASC) recognizing the environmental sounds. Since an audio scene related to a given location potentially contains numerous sound events, only few of these events supply helpful information on the scene, which makes the acoustic scene classification task become a very complex problem. To confront this challenge, we suggest a novel architecture consisting of two basic processes. The front-end process approaches a spectrogram feature, using Gammatone filters. Regarding the back-end classification, we propose a novel convolutional neural network (CNN) architecture that enforces the network deeply learning middle convolutional layers. Our experiments conducted over DCASE2016 task 1A dataset offer the highest classification accuracy of 84.4% as compared to 72.5% of DCASE2016 baseline.

In this paper, we study the achievable sum-rate of a multi-user (MU) multiple-input-multiple-output (MIMO) wireless communication system in which one full-duplex (FD) base station (BS) serves a number of half-duplex (HD) mobile stations (MSs). The problem of interest is to design the precoding matrices of the uplinks and the downlinks to maximize the system sum rate under transmit power constraints at the BS and the MSs. An iterative algorithm based on the gradient projection (GP) and the Armijo rule is developed to maximize the sum-rate. The convergence of the algorithm and the achievable sum-rate performance of the FD MU-MIMO model as compared to those of the HD MU-MIMO one will be investigated through numerical simulation results.

This paper studies a cognitive radio (CR) network which consists of a full-duplex (FD) multi-user (MU) multipleinput multiple-output (MIMO) secondary user (SU) networks operating within the coverage of multiple primary users (PUs). It is assumed that the channel state information (CSI) matrices associated with SU systems are perfectly known whereas the CSI ones from SUs to PUs are imperfectly estimated. The problem of interest is to design robust precoding matrices at the SUs to maximize the CR sum rate subject to the SU transmit power constraints and harmful interference restrictions at PUs. Due to non-concavity of the objective function and intractability of robust PU interference constraints, the design problem is non-convex and challenging to directly solve. We exploit the difference of two concave functions to recast the sum rate objective function as a lower bounded concave one. In addition, a linear matrix inequality (LMI) transformation is used to handle the semi-infinite robust interference constraints. Then, the sequential convex programming method is carried out to iteratively solve a convex optimization problem in each iteration. The simulation results are provided to investigate the CR sum-rate (spectral efficiency) performance and the robustness against the CSI uncertainty.

This paper studies the energy efficiency (EE) optimization problem in multicell wireless networks in which each base station (BS) equipped with multiple antennas serves multiple users at the same time and in the same frequency. The problem of interest is to design the precoders to maximize the network EE subject to practical power constraints at physical layers. The resultant optimization design problem is nonconvex fractional programming and, thus, finding its optimal solution is mathematically challenging. In this paper, we use a combination of difference of convex (d.c.) programming and the Dinkelbach algorithm to iteratively solve the optimization problem. Then, by numerical simulations, we verify the convergence characteristics of the iterative algorithm and examine the EE performance of the system as compared to an spectral efficiency (SE) approach.

This letter studies the transceiver design for optimizing the energy efficiency (EE) in a secure multiple-user multiple-input-multiple-output wireless system. Such a design problem is a highly-nonlinear program whose optimal solution is hardly found. To address its suboptimal solution, an interference alignment technique is first employed to cancel information leakage and interference, and, thus, the design problem is recast into a computationally tractable concave–convex fractional program. Then, we reformulate the fractional program into convex optimization by adopting the Dinkelbach method. Furthermore, by invoking the dual problem, a semiclosed-form solution is derived for an efficient iterative algorithm. The simulation results demonstrate the superior EE performance of the proposed method over the conventional methods, especially in the high transmit power regime.

This paper investigates the precoding design for cooperative multiuser multiple-input multiple-output (MU-MIMO) muticell wireless networks with full-duplex base stations (BSs). The precoding design problem is to maximize the network energy efficiency (EE) under transmit power constraints. To handle the non-concave-convex fractional programming, we adopt the combination of the successive convex programming (SCP) method and Dinkelbach approach for a numerical iterative algorithm. By numerical simulations, we investigate the convergence characteristic of the iterative algorithm for the EE maximization under different transmit power budgets. We also examine the EE and spectral efficiency (SE) for the EE maximization and conventional SE maximization methods for various transmit power budgets under the predefined self-interference and co-channel interference levels.

Brain–computer interfaces (BCIs) are desirable for people to express their thoughts, especially those with profound disabilities in communication. The classification of brain patterns for each different subject requires an extensively time-consuming learning stage specific to that person, in order to reach satisfactory accuracy performance. The training session could also be infeasible for disabled patients as they may not fully understand the training instructions. In this paper, we propose a unified classification scheme based on ensemble classifier, dynamic stopping, and adaptive learning. We apply this scheme on the P300-based BCI, with the subject-independent manner, where no learning session is required for new experimental users. According to our theoretical analysis and empirical results, the harmonized integration of these three methods can significantly boost up the average accuracy from 75.00% to 91.26%, while at the same time reduce the average spelling time from 12.62 to 6.78 iterations, approximately to two-fold faster. The experiments were conducted on a large public dataset which had been used in other related studies. Direct comparisons between our work with the others’ are also reported in details.

This paper aims at finding an automatic approach for extracting features of the Vietnamese sign language to classify both static Vietnamese alphabet letters and their combing diacritic marks as dynamic hand gestures. A Vietnamese sign language recognition system (VSLRS) collects all images including depth images, RGB images, and skeletal join maps to extract the desired features of each hand gesture and their own movements. These characteristics are normalized and converted to build a full Vietnamese sign language combing diacritic marks. The primary features of this system are automatically extracting the hand gestures of the observed person before the Kinect device version 1, and both dynamic and static diacritic marks are able to be recognized because of the movement detection method. Multi-class support vector machines (SVMs) and the One-Against-All approach are employed to find two suitable SVMs for static and dynamic hand gesture recognition. During the recognition phase, all hand gestures are extracted, normalized, and then filtered out based on the Euclidean distance difference of hand positions in captured frames to go through the exact SVMs. The recognized letter or diacritic is the positive label of all the SVM classes. The experimental results demonstrate the proposed VSLRS recognized the Vietnamese sign language (VSL) in realtime with the high accuracy.

Observers are computational algorithms designed to estimate unmeasured state variables due to the lack of appropriate estimating devices or to replace high-priced sensors in a plant. It is always important to estimate those states prior to developing state feedback laws for control and to prevent process disruptions, process shutdowns and even process failures. The diversity of state estimation techniques resulting from intrinsic differences in chemical process systems makes it difficult to select the proper technique from a theoretical or practical point of view for design and implementation in specific applications. Hence, in this paper, we review the applications of recent observers to chemical process systems and classify them into six classes, which differentiate them with respect to their features and assists in the design of observers. Furthermore, we provide guidelines in designing and choosing the observers for particular applications, and we discuss the future directions for these observers.

© 2017 IEEE. In most Brain-Computer Interface systems, especially the P300-Speller, there must be a harmonized balance between the accuracy and the spelling time. One major drawback of the classical 36-choice P300-Speller is the slow rate of character elicitation. This paper aims to propose a real-time signal processing method to decrease the spelling time by exploiting the score margins of the ensemble Support Vector Machine classifiers during real-time P300-Speller flashes, rather than just getting the classifiers' highest scores. Our experiments were conducted on the dataset of the BCI Competition III and resulted in a successful character rate of over 96% with just approximately 15 to 20 seconds for each character spelling session. As compared with the fixed 31.5 seconds of the best original approach of the competition, our proposed method significantly reduces the required spelling time by over 30% while maintaining the desired classification accuracy.

© 2017 IEEE. Brain-computer interfaces (BCIs) are used to assist people, especially those with verbal or physical disabilities, communicate with the computer to indicate their selections, control a device or answer questions only by their mere thoughts. Due to the noisy nature of brain signals, the required time for each experimental session must be lengthened to reach satisfactory accuracy. This is the trade-off between the speed and the precision of a BCI system. In this paper, we propose a unified method which is the integration of ensemble classifier, dynamic stopping, and adaptive learning. We are able to both increase the accuracy, as well as to reduce the spelling time of the P300-Speller. Another merit of our study is that it does not require the training phase for any new subject, hence eliminates the extensively time-consuming process for learning purposes. Experimental results show that we achieve the averaged bit rate boost up of 182% on 15 subjects. Our best achieved accuracy is 95.95% by using 7.49 flashing iterations and our best achieved bit rate is 40.87 bits/min with 83.99% accuracy and 3.64 iterations. To the best of our knowledge, these results outperformed most of the related P300-based BCI studies.

© 2017 IEEE. In brain-computer interface (BCI) research, there must be a trade-off between accuracy and speed of the BCI system, especially those based on event-related potentials (ERPs). This paper proposes a novel method which can significantly increase the spelling bit rate while also maintaining the desired accuracy. We provide an adaptive real-time stopping method based on the scores of ensemble support vector machine classifiers. We apply a criteria assessment process on the classifiers' scores to dynamically stop the ERP-evoked paradigms at any flashing sequence. Our experiments were conducted on three different P300-Speller data sets (BCI Competition II, BCI Competition III and Akimpech). Our proposed framework significantly outperformed the related state-of-the-art studies in terms of character output accuracy and elicitation bit rate rise between static and dynamic stopping schemes. We improve the average bit rate by over 80% while perfectly maintaining the best original static accuracy of over 96%.

This contribution discusses open and closed reacting systems. In recent years, this subject has attracted the attention of many researchers, studying the problems of stability characterization exploiting thermodynamics potentials and reacting invariant structures description using a graph-theoretic approach. Here, we outline an approach to these problems from the expression of mass-action kinetics as potential-driven reacting forces.

This paper addresses the optimal cooperative beamforming design for multi-user multi-relay wireless networks in which the single-carrier frequency division multiple access (SC-FDMA) technique is employed at the terminals. The problem of interest is to find the beamforming weights across relays to maximize the minimum signal-to-interference-plus-noise ratio (SINR) among source users subject to individual power constraints at each relay. Such a beamforming design is shown to be a hard nonconvex optimization problem and therefore it is mathematically challenging to find the optimal solution. By exploring its partial convex structures, we recast the design problem as minimization of a d.c. (difference of two convex) objective function subject to convex constraints and develop an effective iterative algorithm of low complexity to solve it. Simulation results show that our optimal cooperative beamforming scheme realizes the inherent diversity order of the relay network and it performs significantly better than the equal-power beamforming weights.

Electroencephalography (EEG) is an emerging held of digital signal processing. EEG is electrical signals which are recorded by the sensors attached on the human scalp to detect human's brain activities along the scalp. EEG signal processing has been a technically challenging problem for researchers due to its extremely noisy nature compared to other different kinds of digital data such as voice or image. However, EEG signals also create promising research areas and applications for the human being. This paper aims at designing a brain-controlled interface using P300 component in EEG signal processing. The EEG signal's P300 component is an event related potential (ERP) component elicited by the human's brain in the process of decision making. We use the oddball paradigm presented by the P300-Speller to create the brain-computer interface for the subject to express their intention of thinking via the computer screen. The device used to record the brain signal is the Emotiv EPOC headset. We exploit the Bayesian linear discriminant analysis (BLDA) to classify the P300 signals. We conduct the experiments to help people express their thought in l-out-of-4-choice interfaces in real time. The choices, which are presented on the computer screen, could be anything from letters to images or specific symbols. The experimental results reveal the acceptable accuracy of correct classification to be 80 percent and the best bit rate of over 10 bits per minute.

In this paper, precoding techniques are investigated for secure energy-efficient multiple-input multiple-output (MIMO) systems in the presence of an eavesdropper. To tackle the non-convexity of the secure energy efficiency maximization (SEEM) problem, we employ branch-and-reduce-and-bound (BRB) approach to obtain the globally optimal solution. It is observed that the BRB algorithm suffers from highly computational cost, its globally optimal solution is importantly served as a benchmark for the performance evaluation of the suboptimal algorithms. This motivates us to develop a low-complexity approach using the well-known zero-forcing (ZF) technique to cancel the wiretapped signal. In ZF based method, the SEEM problem can be transformed to a concave-convex fractional one and, then, be solved by the combination of the Dinkelbach and bisection search algorithm. Simulation results reveal that the ZF based method can converge fast and obtain a sub-optimal performance in compared with the optimal solution of the BRB method. The ZF based scheme also outperforms the conventional secrecy rate maximization precoder design in terms of the energy efficiency.

The recovery of handwriting’s dynamic stroke is an effective method to help improve the accuracy of any handwriting’s authentication or verification system. The recovered trajectory can be considered as a dynamic feature of any static handwritten images. Capitalising on this temporal information can significantly increase the accuracy of the verification phase. Extraction of dynamic features from static handwritings remains a challenge due to the lack of temporal information as compared to the online methods. Previously, there are two typical approaches to recover the handwriting’s stroke. The first approach is based on the script’s skeleton. The skeletonisation method has highly computational efficiency whereas it often produces noisy artifacts and mismatches on the resulted skeleton. The second approach deals with the handwriting’s contour, crossing areas and overlaps using parametric representations of lines and thickness of strokes. This method can avoid the artifacts, but it requires complicated mathematical models and may lead to computational explosion. Our paper is based on the script’s extracted skeleton and provides an approach to processing static handwriting’s objects, including edges, vertices and loops, as the important aspects of any handwritten image. Our paper is also to provide analysing and classifying loops types and human’s natural writing behavior to improve the global construction of stroke order. Then, a detailed tracing algorithm on global stroke reconstruction is presented. The experimental results reveal the superiority of our method as compared with the existing ones.

This paper aims at finding an efficient approach for automatic human action recognition to classify human actions in both outdoor and indoor environments. A human action recognition system (HARS) collects video frames of human activities, extracts the desired features of each human skeleton. These characteristics are calculated, classified to build a skeleton database that can distinguish almost human gestures. This HARS converts every sequence of human gestures to the sequences of skeletal joint mapping (SJM). Then it assigns corresponding observation symbols to each SJM. Those observation sequences are used to train of hidden Markov models (HMMs) corresponding to seven actions: standing, walking, running, jumping, falling, lying, and sitting. Baum-Welch and forward-backward algorithms are employed to find optimal parameters of each HMM. During a recognition phase, each human gesture sequence is converted to an observation sequence and put into seven optimized HMM models. The current action can be identified by finding a model with the highest probability. The experimental results show that the proposed HARS offers high accuracy of action recognition in real-time.

Electroencephalography (EEG) which is the electrical signal recorded by the sensors attached on the human scalp to detect brain activities has been the emerging trends in digital signal processing. As compared to processing other types of digital data such as speech or audio signals, EEG signal processing is more challenging. However, EEG signals have practically found a wide range of important applications. In this paper, we propose a design of a brain-computer interface (BCI) using EEG's P300 component to a control application. First, we use the Emotiv EPOC headset to capture the raw EEG signals. Then, we adopt a classification algorithm by invoking support vector machine along with the selected extracted features to classify the two-class EEG trials (with and without P300 component). The algorithm is developed to help people express their selection of one among four commands. The experimental results are provided evaluate the classification accuracy.

This paper presents two-hop relay gain-scheduling control in a Wireless Sensor Network to estimate a static target prior characterized by Gaussian probability distribution. The target is observed by a network of linear sensors, whose observations are transmitted to a fusion center for carrying out final estimation via a amplify-and-forward relay node. We are concerned with the joint transmission power allocation for sensors and relay to optimize the minimum mean square error (MMSE) estimator, which is deployed at the fusion center. Particularly, such highly nonlinear optimization problems are solved by an iterative procedure of very low computational complexity. Simulations are provided to support the efficiency of our proposed power allocation., Comment: Submitted to IEEE ICCE 2016

It is known that the design of optimal transmit beamforming vectors for cognitive radio multicast transmission can be formulated as indefinite quadratic optimization programs. Given the challenges of such nonconvex problems, the conventional approach in literature is to recast them as convex semidefinite programs (SDPs) together with rank-one constraints. Then, these nonconvex and discontinuous constraints are dropped allowing for the realization of a pool of relaxed candidate solutions, from which various randomization techniques are utilized with the hope to recover the optimal solutions. However, it has been shown that such approach fails to deliver satisfactory outcomes in many practical settings, wherein the determined solutions are found to be unacceptably far from the actual optimality. On the contrary, we in this contribution tackle the aforementioned optimal beamforming problems differently by representing them as SDPs with additional reverse convex (but continuous) constraints. Nonsmooth optimization algorithms are then proposed to locate the optimal solutions of such design problems in an efficient manner. Our thorough numerical examples verify that the proposed algorithms offer almost global optimality whilst requiring relatively low computational load.

In this paper, interference alignment (IA) is designed for secure multiuser multiple-input multiple-output systems in the presence of an eavesdropper. The proposed IA technique designs the transmit precoding and receiving subspace matrices to minimize both the total inter-main-link interference and the wiretapped signals. With perfect channel state information is known at the transmitters (Txs), the cost function of the optimization problem is alternatively minimized over the precoding matrices at the Txs and the receiving subspace matrices at the receivers (Rxs) and the eavesdropper. The feasible condition, the proof of convergence of the proposed IA approach are provided. Numerical results reveal a significant improvement in terms of average secrecy sum rate of our IA algorithm as compared to the conventional IA design., Comment: submitted to the international conference on Advanced Technologies for Communications (ATC) 2015