Search

Clinical sleep diagnosis traditionally relies on polysomnography (PSG) and expert manual classification of sleep stages. Recent advancements in deep learning have shown promise in automating sleep stage classification using a single PSG channel. However, variations in PSG acquisition devices and environments mean that the number of PSG channels can differ across sleep centers. To integrate a sleep staging method into clinical practice effectively, it must accommodate a flexible number of PSG channels. In this paper, we proposed FlexSleepTransformer, a transformer-based model designed to handle varying number of input channels, making it adaptable to diverse sleep staging datasets. We evaluated FlexSleepTransformer using two distinct datasets: the public SleepEDF-78 dataset and the local SleepUHS dataset. Notably, FlexSleepTransformer is the first model capable of simultaneously training on datasets with differing number of PSG channels. Our experiments showed that FlexSleepTransformer trained on both datasets together achieved 98% of the accuracy compared to models trained on each dataset individually. Furthermore, it outperformed models trained exclusively on one dataset when tested on the other dataset. Additionally, FlexSleepTransformer surpassed state-of-the-art CNN and RNN-based models on both datasets. Due to its adaptability with varying channels numbers, FlexSleepTransformer holds significant potential for clinical adoption, especially when trained with data from a wide range of sleep centers. [ABSTRACT FROM AUTHOR]

Beaconing is a fundamental task in IoT networks, where each node tries to locally broadcast a packet to all its neighbors. Unfortunately, the problem of Minimum Latency Beaconing Schedule (MLBS), which tries to obtain the fastest and collision-free schedule, is not well studied when the IoT devices employ the duty-cycled working mode. The existing works have rigid assumptions that there exists a single channel and can only work in a centralized fashion. Aiming at making the work more practical and general, in this paper, we investigate the first distributed method for the MLBS problem in Multi-channel asynchronous Duty-cycled IoT networks (MLBSMD problem). The MLBSMD problem is first formulated and proved to be NP-hard. To avoid collisions locally, several special structures are designed which can work in O (Δ 2) time, where Δ denotes the maximum node degree in the network. Then, a distributed beaconing scheduling method that can compute a low-latency and collision-free schedule is proposed with a theoretical bound, taking the active time slots of each node into account. Finally, the extensive simulation results demonstrate the effectiveness of the proposed algorithm in terms of latency. [ABSTRACT FROM AUTHOR]

In this paper, a multi-channel speech enhancement structure combining beamformers and lightweight neural networks is proposed. By using a first-order differential beamformer to improve the signal-to-noise ratio of the target speech, the burden on the subsequent neural network is reduced, thus lowering the complexity of the required neural network. A two-stage gated recurrent neural network is employed, where the first stage recurrent neural network processes the channel characteristics of the speech and the second handles the frequency domain features of the speech. The frequency band division combined with convolution is used to further compress the network scale. With suitable loss function, the speech enhancement performance of the model is further improved. The proposed model structure is trained, evaluated, and validated using simulated multi-channel microphone array datasets generated from the public TIMIT dataset. The results demonstrate that our model achieves good multi-channel speech enhancement performance with relatively small parameter and computational requirements when compared to popular existing approaches. [ABSTRACT FROM AUTHOR]

Speech enhancement in 3D reverberant environments is a challenging and significant problem for many downstream applications, such as speech recognition, speaker identification, and audio analysis. Existing deep neural network models have shown efficacy for 3D speech enhancement tasks, but they often introduce distortions or unnatural artifacts in the enhanced speech. In this work, we propose a novel two-stage refiner system that integrates a neural beamforming network and a diffusion model for robust 3D speech enhancement. The neural beamforming network performs spatial filtering to suppress the noise and reverberation; while, the diffusion model leverages its generative capability to restore the missing or distorted speech components from the beamformed output. To the best of our knowledge, this is the first work that applies the diffusion model as a backend refiner to 3D speech enhancement. We investigate the effect of training the diffusion model with either enhanced speech or clean speech, and find that clean speech can better capture the prior knowledge of speech components and improve the speech recovery. We evaluate our proposed system on different datasets and beamformer architectures, and show that it achieves consistent improvements in metrics like WER and NISQA, indicating that the diffusion model has strong generalization ability and can serve as a backend refinement module for 3D speech enhancement, regardless of the front-end beamforming network. Our work demonstrates the effectiveness of integrating discriminative and generative models for robust 3D speech enhancement, and also opens up a new direction for applying generative diffusion models to 3D speech processing tasks, which can be used as a backend to various beamforming enhancement methods. [ABSTRACT FROM AUTHOR]

Investigation of the proposed Silicon gate all around dielectric window spaced-multi-channel (Si-GAA-DWS-multi-channel) MOSFET is carried out through 3D-ATLAS TCAD simulator. The proposed device parameter has been compared with the existing devices, mainly with Silicon-Nanowire-Dielectric Pocket Packed MOSFET (Si-NW-DPP FET) and Silicon-Nanowire MOSFET (Si-NW FET). The Si-GAA-DWS-Multi-channel FET is found to be the better device than the other two structures with a threshold voltage of 0.40 V, an on current of 0.326 × 10–3 A, off current of 5.9 × 10–13 A, on/off current ratio of 5.52 × 108, sub-threshold swing of 63.4 mV/dec and DIBL of 26.25 mV/V. Impact of interface fixed charges and multi state defects have been studied for the Si-GAA-DWS-multi-channel MOSFET. Presence of fixed negative charges and multi state defects at semiconductor-oxide interface causes a step in the potential profile which results in the shift of threshold voltage, degradation of drain current, and Ion/Ioff ratio. [ABSTRACT FROM AUTHOR]

Semantic segmentation of aerial images plays a pivotal role in extracting detailed information about land cover, infrastructure, and natural features. Traditional single-channel segmentation models struggle to harness the rich information present in multi-channel aerial images, such as multi-spectral or hyperspectral data. While the DeepLabV3 + architecture has shown remarkable success in semantic segmentation tasks by exploiting multi-scale context and atrous convolutions, its performance on aerial images remains suboptimal due to the unique challenges of this domain. With the motive of addressing the difficulties in the existing semantic segmentation techniques for aerial images, the adoption of deep learning techniques is utilized. A multi-objective derived Adaptive Multichannel deeplabv3 + (AMC-Deeplabv3 +) with a new meta-heuristic algorithm called Improved Beluga whale optimization (IBWO) algorithm is proposed in this paper. Here, the hyperparameters of Multichannel deeplabv3 + are optimized by the IBWO algorithm and this model intends to set new benchmarks in the accuracy and contextual understanding of aerial image segmentation by integrating multi-channel data processing techniques and preserving spatial context. The proposed model attains improved accuracies of 98.65% & 98.72% for dataset 1 and 2 respectively and also achieves the dice coefficient of 98.73% & 98.85% respectively, with a computation time of 113.0123 s. The evolutional outcomes of the proposed model show significantly better than the state-of-the-art techniques. [ABSTRACT FROM AUTHOR]

Emotion recognition through the fusion of multi-modal sources, including voice and visual cues, holds paramount importance in comprehending human affective states. This research introduces an inventive paradigm, seamlessly uniting the potential of adaptive normalization fusion (ANF) and alpha Gaussian dropout (AGD) within the architecture of a multi-channel convolutional neural network (MCNN). ANF employs a sequence of pre-processing steps, beginning with batch normalization, followed by min–max scaling using the sigmoid function, and further augmented by alpha Gaussian dropout (AGD)—a fusion of Gaussian dropout and alpha dropout. These techniques are seamlessly integrated into multi-channel convolutional neural network (MCNN) architecture, aimed at improving emotion classification accuracy. This synergy fortifies the network's robust generalization capacities while effectively mitigating overfitting concerns. Rigorous experimentation and comprehensive evaluations across datasets that encompass voice and visual-based emotion recognition underscore the remarkable efficiency of the proposed methodology, culminating in notable strides in accuracy and generalization. The culmination of this research delivers a notable achievement, reflecting an impressive 97% accuracy in face emotion recognition and an exceptional 98% accuracy in speech emotion recognition. [ABSTRACT FROM AUTHOR]

In ubiquitous computing environment, data is collected by a number of devices and transmitted to a central server which is trusted to provide the information to the user. In order to cope with security threats targeting the central service providing system, there is a growing trend to use unidirectional data transmission gateway to completely block the possibility of intrusion of an external network and an internal network. There is a difficulty in that a physical unidirectional data transmission gateway up to now can use only a single session and does not support a private protocol. In this paper, we propose transmission control protocol based service which is a constraint in unidirectional security gateway system was being available, and unidirectional security gateway using multi-channel transmission method for transmission efficiency in multi-proxy operation and transmission control protocol's reliability and enhance transmission efficiency of the unidirectional security gateway and shows the experimental results of proposed method. [ABSTRACT FROM AUTHOR]

The automatic artery/vein (A/V) classification in retinal fundus images plays a significant role in detecting vascular abnormalities and could speed up the diagnosis of various systemic diseases. Deep-learning methods have been extensively employed in this task. However, due to the lack of annotated data and the serious data imbalance, the performance of the existing methods is constricted. To address these limitations, we propose a novel multi-channel multi-scale fusion network (MMF-Net) that employs the enhancement of vessel structural information to constrain the A/V classification. First, the newly designed multi-channel (MM) module could extract the vessel structure from the original fundus image by the frequency filters, increasing the proportion of blood vessel pixels and reducing the influence caused by the background pixels. Second, the MMF-Net introduces a multi-scale transformation (MT) module, which could efficiently extract the information from the multi-channel feature representations. Third, the MMF-Net utilizes a multi-feature fusion (MF) module to improve the robustness of A/V classification by splitting and reorganizing the pixel feature from different scales. We validate our results on several public benchmark datasets. The experimental results show that the proposed method could achieve the best result compared with the existing state-of-the-art methods, which demonstrate the superior performance of the MMF-Net. The highly optimized Python implementations of our method is released at: https://github.com/chenchouyu/MMF_Net. [ABSTRACT FROM AUTHOR]

The Wireless Sensor Network (WSN) is one of the emerging competences to have possible appliance on wide varieties including investigating about the nature of surroundings, elegant places, therapeutic system, and study of robots. Energy is efficient in deliberation of essential invent for WSN. In WSNs, packet loss may occur due to network congestion, packet collision, bad link quality, buffer overflow, and low energy levels. Retransmitting the lost packets again requires more energy consumption and delay. Ensuring data reliability and maintaining minimum delay with improving energy efficiency are challenging issues in a resource-constrained sensor networks. The variation of link quality and other node status impacts the end-to-end delay of the sensor nodes in the network. On the other hand, the sensor nodes have energy limitations and it is a great concern to extend the network lifetime. To deal with these issues, a novel and simple routing mechanism SCORE BASED LINK DELAY AWARE ROUTING (SBLDAR) is proposed. The proposed SBLDAR protocol selects the appropriate forwarder nodes by enhancing the forwarder node selection method using multiple parameters such as residual energy, distance, delay, Received Signal Strength Indicator, remaining delivery ratio and the Expected Transmission Count of the nodes. Based on the aforementioned parameters the protocol estimate and assign a score for every sensor nodes. The score is a factor describes the stability of the sensor node. The high score of the node denotes the high stability of the node. The proposed SBLDAR protocol implemented in NS2 simulation and compared it with existing protocols MPCBHM, QTSAC, RBDCEER, AdvMMAC and DEEHCB. From the simulation evaluations, we found that SBLDAR is able to achieve an improved network lifetime over the current protocols while maintaining the average data transmission delay. In the simulation, the SBLDAR achieved almost 75% more throughput and saved 60% of total consumed energy compared with compared protocols. In addition, the proposed SBLDAR protocol eliminated 40% of transmission delay when compared with MPCBHM, QTSAC methods and 85% of transmission delay compared with RBDCEER, AdvMMAC and DEEHCB methods. [ABSTRACT FROM AUTHOR]

In this paper, a parallel computing method is proposed to perform the background denoising and wheezing detection from a multi-channel recording captured during the auscultation process. The proposed system is based on a non-negative matrix factorization (NMF) approach and a detection strategy. Moreover, the initialization of the proposed model is based on singular value decomposition to avoid dependence on the initial values of the NMF parameters. Additionally, novel update rules to simultaneously address the multichannel denoising while preserving an orthogonal constraint to maximize source separation have been designed. The proposed system has been evaluated for the task of wheezing detection showing a significant improvement over state-of-the-art algorithms when noisy sound sources are present. Moreover, parallel and high-performance techniques have been used to speedup the execution of the proposed system, showing that it is possible to achieve fast execution times, which enables its implementation in real-world scenarios. [ABSTRACT FROM AUTHOR]

The rising user needs for high speed applications and services have necessitated the requirement of a communication link that can provide the data rate comparable to optical fiber. Free Space Optical (FSO) Communication is an admired means of communication from past few years. To achieve seamless and error free transmission at high speed for heterogeneous multimedia applications at desired bit rates, novel advancements in optical communication technology are being explored along with FSO. However, several constraints such as atmospheric turbulence, uneven terrain, dynamic climatic conditions cause several nonlinearities which limit the connectivity. In this work, a multi channel multi beam FSO communication system by incorporating Reed Solomon (RS) Codes has been proposed at 30 Gbps. The Modified Gamma-Gamma model is being applied for channel turbulences. The application of Forward Error Correction (FEC) coding in the designed system offers ease of error detection and correction with improved performance at higher data rate for implementation over longer link lengths.The system performance has been evaluated for (i) Number of errors corrected, (ii) Bit Error Rate (BER), (iii) Received power, (iv) Geometric path losses and (v) Signal to Noise Ratio (SNR). At an acceptable BER of 10–9, the RS (255, 127, 129) encoded Four channel Four beam FSO communication link is effectively implemented for a distance of 3500 m with − 32.5 dBm received low optical power and 25 dB SNR. [ABSTRACT FROM AUTHOR]

The Taiji program and the LISA (Laser Interferometer Space Antenna) program are proposed to realize the gravitational wave detection within the frequency band of 0.1 mHz-1 Hz through space laser interferometry. As a key optical component of the interferometric system, the telescope are required the dimensional stability better than 1 pm / Hz and its deformation between the primary and the secondary mirrors needs to be small enough to reduce wavefront distortion. Therefore, in the case of satisfying the structural strength, the telescope supporting structural material also needs to meet the performance of the low coefficient of thermal expansion (CTE) and high thermal conductivity. In this paper, we fabricate a telescope supporting frame using the carbon/silicon carbide (C/SiC) composite, and build a high-precision and multi-channel thermal deformation interferometry system. Using this system, the uniform thermal expansion experiment and the temperature gradient thermal expansion experiment are carried out on the telescope supporting frame. The results show that the axial CTE and the transverse CTE of the C/SiC composite is about - 8.3 × 10 - 7 /K and 3.3 × 10 - 5 /K, respectively, which indicate that the design and process of the C/SiC composite used may not be mature and need to be further studied. The high-precision multi-channel thermal deformation measurement interferometry system designed in this paper can be used for the subsequent thermal performance measurement of the telescope prototype. [ABSTRACT FROM AUTHOR]

In this work, we have proposed a quadruple gate multi-channel junctionless MOSFET structure that acts as a label-free biosensor to identify neutral and charged biomolecules by dielectric modulation procedure. The device has been analytically modelled through the solution of two dimensional Poisson's equation using conformal mapping to account for the fringing fields. As various bio targets are immobilized inside the cavity on a thin adhesive silicon dioxide film, the device electric properties such as the surface potential, drain current, threshold voltage change differently. The threshold voltage shift has primarily been utilised as the sensing parameter for the detection and identification of a specific biomolecule. The biosensing characteristics have been discussed by varying the dielectric constant and charge of biomolecules with analysing non-ideal issues like temperature variation and steric hindrance. A DNA-based variability study has also been done with real-time analysis. All the results obtained from the analytical model have been validated against SILVACO TCAD simulation data to substantiate our developed model. The structure exhibits near ideal subthreshold slope and low threshold voltage roll-off even at sub − 20 nm regime and high, reliable sensitivity for a wide range of bio-targets suggesting the possible application in short channel biosensor regime. [ABSTRACT FROM AUTHOR]

This paper presents a single-stage single-switch multi-channel LED driver with high extendibility. The PFC module adopts a buck–boost circuit, which can achieve a high-power factor when it works in the discontinuous current mode (DCM). At the same time, the leakage inductor energy can be recycled by adding a capacitor at the primary side to reduce the voltage spike and ringing of the switch. The DC–DC module adopts a forward converter that plays the role of electrical isolation. Using capacitor charge balance, the converter can achieve output current balancing of all the channels. This converter can extend the number of output channels while using fewer capacitors than traditional LED drivers based on capacitor charge balance. In addition, the PFC module is integrated with the DC–DC module while sharing a common switch, which makes the control simple. This paper provides the concrete structure, working principle, and circuit characteristic analysis of the proposed topology. Finally, a 100 W prototype was built to verify its feasibility. [ABSTRACT FROM AUTHOR]

The aim of quantum metrology is to exploit quantum effects to improve the precision of parameter estimation beyond its classical limit. In this paper, we investigate the quantum parameter estimation problem with multiple channels. It is related but not limited to the following two important and practical quantum metrology problems: (i) Quantum enhanced metrology with control, whose aim is to improve the precision of quantum sensing by utilizing feedback or open-loop control; (ii) Practical quantum metrology where the underlying evolution of quantum probes may change from a unitary dynamics to an open system dynamics, owing to the inevitable decoherence during the quantum sensing operation. For various kinds of quantum multiple channels, the corresponding quantum channel Fisher information is derived. To demonstrate the results, some illustrative examples are given. [ABSTRACT FROM AUTHOR]

The thirst for internet access in moving aircraft grows rapidly by air passengers as human's daily activities rely on digital dependency. Currently, in-flight Wi-Fi is provided by satellite-based and cellular-based systems, where the line-of-sight problem, lack of ground infrastructure, high cost, and long delay negatively affect the internet connectivity. To alleviate these demerits associated with the existing systems, the Aeronautical Ad hoc Network (AANET) has been developed as a complementary system. Internet of Things is being realized over the sky with the aid of AANET, whereby the moving flights can share their live information to the corresponding ground stations while they move across oceanic or remote areas. Establishing IoT in AANET for civil aviation systems has a great challenge in providing reliable and efficient data delivery between flight and ground stations due to their unique characteristics. This paper proposes Cross-layered Multi-Channel Automatic Dependent Surveillance based Load Sharing Routing algorithm to improve the performance of civil AANET. The ever-increasing need for proper utilization of available bandwidth is a key factor to introduce multi-channel operation in this work. Instead of allocating channels before acquiring routes, this work attempts to couple the process of packet routing followed by channel assignment to improve the network routing performance by increasing the ratio of successful packet delivery, reducing the overall delay and traffic overhead through a cross-layered approach. The primary aim of the proposed work is to enhance the load balancing, bandwidth utilization and to reduce the delay associated with the queuing of the AANET systems. The simulation setup is carried out by using QualNet 5.2 simulator and the experimental results have been obtained for packet delivery ratio, end-to-end delay, Traffic overhead in various scenarios. The result shows that the proposed work outperforms the existing routing methods. [ABSTRACT FROM AUTHOR]

Wire electrical discharge machining (EDM) is the most important approach to cutting difficult-to-machine materials and components with complex shapes in the manufacturing industry. However, the multiple demands for high material removal rate, high surface quality, and low energy consumption require contradictory working conditions and restrict the further improvement of the performance of WEDM. This paper introduced a novel powder-mixed multi-channel WEDM method using the multi-channel discharge effect to meet the conflicting requirements. The multi-channel discharge effect utilizes the equipotential characteristics of the semiconductor powder mixed in the dielectric to disperse discharge energy and therefore provides a feasible solution to resolve the above contradictions. New working principles and machining mechanisms were discovered and verified by the simulation and experimental results. Comparative experiments show that the new powder-mixed multi-channel discharge WEDM method significantly reduced surface roughness and thermal defects while maintained a similar material removal rate as conventional WEDM. [ABSTRACT FROM AUTHOR]

Wireless sensor networks (WSNs) have become an important component in the Internet of things (IoT) field. In WSNs, multi-channel protocols have been developed to overcome some limitations related to the throughput and delivery rate which have become necessary for many IoT applications that require sufficient bandwidth to transmit a large amount of data. However, the requirement of frequent negotiation for channel assignment in distributed multi-channel protocols incurs an extra-large communication overhead which results in a reduction of the network lifetime. To deal with this requirement in an energy-efficient way is a challenging task. Hence, the Reinforcement Learning (RL) approach for channel assignment is used to overcome this problem. Nevertheless, the use of the RL approach requires a number of iterations to obtain the best solution which in turn creates a communication overhead and time-wasting. In this paper, a Self-schedule based Cooperative multi-agent Reinforcement Learning for Channel Assignment (SCRL CA) approach is proposed to improve the network lifetime and performance. The proposal addresses both regular traffic scheduling and assignment of the available orthogonal channels in an energy-efficient way. We solve the cooperation between the RL agents problem by using the self-schedule method to accelerate the RL iterations, reduce the communication overhead and balance the energy consumption in the route selection process. Therefore, two algorithms are proposed, the first one is for the Static channel assignment (SSCRL CA) while the second one is for the Dynamic channel assignment (DSCRL CA). The results of extensive simulation experiments show the effectiveness of our approach in improving the network lifetime and performance through the two algorithms. [ABSTRACT FROM AUTHOR]

In hand-drawn sketch recognition, the traditional deep learning method has the problems of insufficient feature extraction and low recognition rate. To solve this problem, a new algorithm based on a dual-channel convolutional neural network is proposed. Firstly, the sketch is preprocessed to get a smooth sketch. The contour of the sketch is obtained by the contour extraction algorithm. Then, the sketch and contour are used as the input image of CNN. Finally, feature fusion is carried out in the full connection layer, and the classification results are obtained by using a softmax classifier. Experimental results show that this method can effectively improve the recognition rate of a hand-drawn sketch. [ABSTRACT FROM AUTHOR]

Fine-grained recognition is a very challenging issue, since it is difficulty to mine discriminative and subtle feature for objects with similar visual appearance. Because massive manual annotations (e.g., bounding box for discriminative regions) are time-consuming and labor-consuming, existing methods designed single form of attention model outputted discriminative regions in a weakly supervised way. In this paper, we proposed a novel method named a Spatial-Channel Aware Attention Filters (SCAF) to address the weakly supervised fine-grained recognition problem. Compared with the previous attention models, SCAF can obtain attentions-aware features from two dimensions, i.e., the spatial location of image and the channel of feature maps. With the proposed SCAF, the model can enhance the discriminative regions on both spatial and channel dimensions simultaneously. In addition, the multi-channel network multi-level structure are designed to extract richer regional features. Moreover, focal loss is introduced to balance the samples' distribution of fine-grained image dataset. Comprehensive and comparable experiments are conducted in publicly available datasets, and the experimental results show that our method can achieve the state-of-the-art performance on fine-grained recognition tasks. For instance, we achieve 99.370%, 80.749% accuracy on two underwater datasets respectively, i.e., Fish4Knowlege and Wild Fish. [ABSTRACT FROM AUTHOR]

Salient object detection is a hot spot of current computer vision. The emergence of the convolutional neural network (CNN) greatly improves the existing detection methods. In this paper, we present 3MNet, which is based on the CNN, to make the utmost of various features of the image and utilize the contour detection task of the salient object to explicitly model the features of multi-level structures, multiple tasks and multiple channels, so as to obtain the final saliency map of the fusion of these features. Specifically, we first utilize contour detection task for auxiliary detection and then utilize use multi-layer network structure to extract multi-scale image information. Finally, we introduce a unique module into the network to model the channel information of the image. Our network has produced good results on five widely used datasets. In addition, we also conducted a series of ablation experiments to verify the effectiveness of some components in the network. [ABSTRACT FROM AUTHOR]

In this paper, we develop a joint channel assignment and adaptive Multi-Input–Multi-Output (MIMO) mode selection protocol for MIMO-based multi-channel cognitive radio networks. The proposed protocol utilizes the dynamic access capabilities of cognitive radio devices along with the multiplexing and/or diversity gains offered by MIMO systems in order to improve network throughput and communication reliability. The proposed protocol attempts at enhancing system performance by selecting the best MIMO operating modes (multiplexing, diversity, or hybrid multiplexing-diversity mode) for each idle channel according to the surrounding RF environment, and then selects the channel that provides the highest throughput. The network performance of our protocol (the overall network throughput and MIMO efficiency) is investigated under various traffic and network conditions through simulations. Compared to diversity-mode channel assignment, multiplexing-mode channel assignment, and adaptive mode selection with random channel selection protocols, our proposed protocol reveals significant performance improvement in network throughput under different traffic and network conditions. [ABSTRACT FROM AUTHOR]

The existence of glass defects seriously affects the quality of glass products. Traditional glass defect recognition methods based on convolutional neural networks (CNNs) suffer from a long training time and low recognition accuracy. By introducing convolutional auto-encoder (CAE) into a CNN, we propose an auto-encoding convolutional neural network, and pre-train the convolution kernel by CAE, thereby reducing the training time caused by randomly initializing the convolution kernel. At the same time, in order to cope with the over-fitting problem caused by a small sample of glass defect dataset, we employ a fuzzy support vector machine (FSVM) instead of the Softmax classifier to classify glass defects. Furthermore, we propose a multi-channel auto-encoding convolutional neural network model to deal with misidentification of inclusions and tumor type defects due to small differences in feature space. The model takes a defect image and its enhanced image as input, and averages the output of each channel as the final output, thus achieving accurate recognition of glass defects. Experimental results show that the glass defect recognition method based on auto-encoding convolutional neural network can greatly reduce the training time while maintaining the same classification accuracy rate. The glass defect recognition method based on multi-channel auto-encoding convolutional neural network achieves a recognition rate of 95% for both inclusion and tumor type defects. As a result, and the overall recognition rate is increased from 92.6% using a single channel to 97%. [ABSTRACT FROM AUTHOR]

There are well-established chemical and turbidity anomalies in the plumes occurring vicinity of hydrothermal vents, which are used to indicate their existence and locations. We here develop a small, accurate multi-channel chemical sensor to detect such anomalies which can be used in deep-sea at depths of more than 4 000 m. The design allowed five all-solid-state electrodes to be mounted on it and each (apart from one reference electrode) could be changed according to chemicals to be measured. Two experiments were conducted using the chemical sensors. The first was a shallow-sea trial which included sample measurements and in situ monitoring. pH, Eh, CO32− and SO42− electrodes were utilized to demonstrate that the chemical sensor was accurate and stable outside the laboratory. In the second experiment, the chemical sensor was integrated with pH, Eh, CO32− and H2S electrodes, and was used in 29 scans of the seabed along the Southwest Indian Ridge (SWIR) to detect hydrothermal vents, from which 27 sets of valid data were obtained. Hydrothermal vents were identified by analyzing the chemical anomalies, the primary judging criteria were decreasing voltages of Eh and H2S, matched by increasing voltages of pH and CO32−. We proposed that simultaneous detection of changes in these parameters will indicate a hydrothermal vent. Amongst the 27 valid sets of data, five potential hydrothermal vents were targeted using the proposed method. We suggest that our sensors could be widely employed by marine scientists. [ABSTRACT FROM AUTHOR]

Wireless networks' applications are increasing gradually necessitating their performance to enhance. Evolution of these networks over time indicates the need for algorithms which can operate based on their dynamic nature. Wireless mesh networks provide Intranet and Internet access for different applications in various environments. It is expected that the traffic load will be high on these networks. As gateway nodes are responsible for the traffic load transmission, gateway selection is known as one of the important research issues in that it can lead to optimized use of the network capacity and reduce congestion effects. In addition, utilizing multi-radio multi-channel architecture is one of the promising methods for increasing performance and decreasing interference. Channel assignment determines the most appropriate channel-radio associations for transmitting and receiving data through different channels simultaneously. Taking into account this architecture, this paper was written to propose a distributed gateway selection algorithm along with a cross-layer concept which predicts environment dynamics by learning automata. Experimental results demonstrate that the proposed method in various configurations on average improves packet delivery ratio 17.66%, throughput 5.36%, network overhead ratio 6.34%, and average end-to-end delay 15.94% higher than reinforcement learning-based best path routing algorithm (RLBPR), the best studied algorithm; therefore, it leads to more efficient utilization of network capacity compared to nearest gateway, minimum load index, expected transmission count, best path to best gateway, and RLBPR algorithms. [ABSTRACT FROM AUTHOR]

This paper presents a novel approach for automatic segmentation of MS lesion including both of number and volume. The novelty includes the combination of the multiplicative intrinsic component optimization algorithm (Li et al. in Magn Reson Imaging 32:913–923, 2014) in bias field correction and normal tissue segmentation simultaneously, and the development of a multi-atlas and multi-channel (MAMC) segmentation approach. The first research focus is the classification of brain tissue into white matter, cerebrospinal fluid and gray matter in T1-w image and FLAIR image. The second research focus is the segmentation of MS lesion in white matter region using atlas. In label fusion, the coefficient as a specific weight is assigned to target label image based on the correlation function between atlases. This novel MAMC approach is evaluated by 20 training cases obtained from Medical Image Computing and Computer Aided Intervention Society 2008 MS Lesions Segmentation Challenge. The numerical results are presented in terms of accuracy, specificity and absolute volume difference. A comparison of MAMC approach and other conventional approaches is presented in terms of the true positive rate and the positive predictive value. Furthermore, the total lesion volume is calculated and compared with expert delineation. It can be seen that the MAMC approach is able to acquire a larger mean value of the Dice similarity coefficient than the other conventional approaches do. Therefore, this novel approach is an added value for the clinical evaluation of MS patients. [ABSTRACT FROM AUTHOR]

Feature vectors and similarity measures are the two key issues of most existing collaborative filtering (CF) algorithms. In item-based CF algorithms, the feature vector is often defined as the ratings of all users for a given item. For a recommender system with n users, m items, and c ratings, the length of the feature vector is n; hence, the time complexity of the similarity computation is O(n). Consequently, the overall time complexity is O (m 2 n 2) , which may be computationally prohibitive for recommender systems with millions of users. In this paper, we define the multi-channel feature vector (MCFV), which is a vector of channel length c, and calculate the similarity between items using the respective MCFVs. Each element of an MCFV corresponds to the number of users with respective ratings for the item. The time complexity for the similarity computation is O(c), and the overall time complexity is O (m 2 n c) when the k-nearest neighbors and weighted average algorithms are used. Experiments were conducted on four movie recommender systems, where n ranges from a few hundred to half a million, and c is five. Results show that the recommendation algorithms using our new similarity measure are significantly faster than their counterparts without sacrificing prediction accuracy in terms of mean absolute error and root mean square error. [ABSTRACT FROM AUTHOR]

Currently, the main shooting equipment used for circular-screen movie production is the multichannel reflective camera system. However, this equipment is cumbersome, the playout method is complex, and the cost of production is very high. To overcome these shortcomings, this study proposes a panoramic video method for producing circular-screen movies. Three key problems of using the panoramic video method are studied and solved: panoramic video flattening, the hardware and software components of a panoramic circular-screen movie playout system, and a playout algorithm for panoramic circular-screen movies. The study gives an example of a panoramic circular-screen movie and compares the features of panoramic circular-screen movies and traditional circular-screen movies. The comparison shows that the equipment used for shooting panoramic circular-screen movie is lightweight, simple to operate, and low cost. These features will promote the development of circular-screen movies. [ABSTRACT FROM AUTHOR]

In this paper, we propose and demonstrate simultaneous phase regeneration of four different channels of QPSK signal based on phase sensitive amplification. The configuration can be divided into two parts. The first one uses four wave mixing in high nonlinear fiber (HNLF) to generate the corresponding three harmonic conjugates precisely at the frequency of the original signals. The other one uses optical combiner to realize coherent addition which is aimed at completely removing the interaction in phase regeneration stage. The simulation results suggest that this scheme can optimize signal constellation to a large extend especially in high noise environment. Besides, optical signal to noise ratio (OSNR) can improve more than 3 dB while the bit-error-rate (BER) reaches 10–3 with a constant white noise and 15° phase noise. [ABSTRACT FROM AUTHOR]

Wireless Mesh Networks (WMN) is a key backhaul technology used in 802.11 networks to provide ubiquitous coverage to isolated areas that require high-speed connectivity. The multi-radio feature of WMN has enabled the mesh routers to derive the full benefits of multiple channels for providing parallel transmissions in a single collision domain. However, co-channel interfering links badly affect the channel capacity and force the mesh routers to switch the radio interface to other less interfering channel. In dynamic channel assignment, if the channel switches occur frequently, the traffic disruptions lead to excessive packet delays and drops. These problems are mostly observed in specific dense areas, where traffic saturation occurs. The existing schemes lack in properly identifying the bandwidth starved links. Therefore, the focus of this paper is to enhance the throughput and minimize the packet drops by critically identifying the bottleneck links and prioritize them for better channel assignments. The proposed metric exploits the statistical inference on dropped packets to determine the effect of interference on the achievable capacity of the links. The traffic load and the effective capacity are collectively used to identify the saturated links. The proposed metric has been evaluated through extensive simulations. The results demonstrate the validation of proposed metric with a considerable increase in performance. [ABSTRACT FROM AUTHOR]

In a multi-channel SQUID measurement system, adjusting device parameters to optimal condition for all channels is time-consuming. In this paper, an intelligent control system is presented to determine the optimal working point of devices which is automatic and more efficient comparing to the manual one. An optimal working point searching algorithm is introduced as the core component of the control system. In this algorithm, the bias voltage Vbias is step scanned to obtain the maximal value of the peak-to-peak current value Ipp of the SQUID magnetometer modulation curve. We choose this point as the optimal one. Using the above control system, more than 30 weakly damped SQUID magnetometers with area of 5×5mm2 or 10×10mm2 are adjusted and a 36-channel magnetocardiography system perfectly worked in a magnetically shielded room. The average white flux noise is 15μΦ0/Hz1/2. [ABSTRACT FROM AUTHOR]

Wireless body area networks (WBANs) support the inter-operability of biomedical sensors and medical institutions with convenience and high-efficiency, which makes it an appropriate solution for the pervasive healthcare. Typically, WBANs comprise in-body or around-body sensor nodes for collecting data of physiological feature. Therefore, the efficient medium access control (MAC) protocol is a crucial paramount to coordinate these devices and forward data to the medical center in an efficient and reliable way. However, the extensive use of wireless channel and coexistence of WBANs may result in inevitable interference which will cause performance degradation. Besides, contention-based access in single channel in WBANs is less efficient for dense medical traffic on account of large packet delay, energy consumption and low priority starvation. To address these issues above, we propose a multi-channel MAC (MC-MAC) scheme to obtain better network performance. Considering the characteristic and emergency degree of medical traffic, we introduce a novel channel mapping and selection mechanism, cooperating with conflict avoidance strategy, to organize nodes to access available channels without collisions. In addition, we have evaluated the performance of MC-MAC and the standard IEEE 802.15.6 via simulation and hardware test. The test is conducted by hardware platform based on prototype system of WBANs. Both of the analysis and simulation results show that MC-MAC outperforms the IEEE 802.15.6 in terms of packet delay, throughput, packet error rate and frame error rate. [ABSTRACT FROM AUTHOR]

With the innovation of Cognitive Radio Wireless Mesh Network’s, Opportunistic Spectrum Allocation can possibly moderate spectrum lack, by letting Secondary User’s a chance to sense and use unused bits of opportunistic licensed spectrum without any unfavourable effect on the Primary User’s. In Cognitive Radio Wireless Mesh Network’s, the Medium-Access-Control protocols characterizes the utilization of spectrum proficiently by allocating the channels among the users. Majority of proposed Medium-Access-Control protocols are utilizing Common-Control-Channel for dealing with the assets of Secondary-Users. The major downsides of these Medium-Access-Control protocols are broadcasting of Control-channel when substantial number of Secondary-User exists. In contrast with these Medium-Access-Control protocols, we develop an algorithm Time-Slotted-Allocation-Medium-Access-Control (TSA-MAC) protocol which is based on Clustered Time-Division-Medium-Access approach, which permits Secondary-User’s to allocate opportunistic spectrum with the help of co-operative decisions by exchange control information. In this approach, we are dividing the channels as different slots on which Secondary-User’s can transfer control and data packets. The TSA-MAC protocol will enhance the throughput for the Secondary-User’s over the communication channel. And also this method will facilitate to decide and allocate free channels for Secondary-User’s without interfering with Primary User’s. [ABSTRACT FROM AUTHOR]

Recently, explosive growth of bandwidth demands has motivated many technological revolutions in the Wireless Local Area Networks (WLANs) such as the IEEE 802.11ax task group, which is established to enhance the throughput performance for the Next Generation WLANs (NGW) under high dense deployment scenarios. However, on the one hand, it is known that the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) has become a generally accepted access mechanism in the WLANs, which is shown to bring about serious collisions when the stations (STAs) are relatively crowded. In this case, the channel access efficiency is definitively decreased and thus some frequency channel resources are eventually wasted. On the other hand, due to the inherent fading effect of wireless channel, network throughput of the NGW (i.e., 802.11ax) is further degraded by the existence of Low-Rate-Links (LRLs), where the available data transmission rate is relatively low. To resolve the above two technical issues, a distributed multi-channel MAC protocol, called CRC-MMAC, is proposed for the NGW. In the proposed CRC-MMAC, the concept of reserved-cooperative-link (RCL) is proposed and initiated under multi-channel environment, to fully exploit the potential of both channel reservation and cooperative relay. Accordingly, collisions in the network are effectively decreased using channel reservation as well as the data transmission rate of LRLs is significantly improved with cooperative relay. Furthermore, an analysis of the upper bound of saturation throughput gain is derived, which is validated by extensive simulations. Compared with the ‘Baseline’ scheme, i.e., the existing Dynamic Channel Assignment (DCA) protocol [1] using TXOP (Transmission Opportunity), the experiments results show that the saturation throughput of CRC-MMAC exceeds about 140%, and the average packet delay is decreased by nearly 60%. [ABSTRACT FROM AUTHOR]

Recent rich applications for the Internet of Things are demanding large bandwidth for communication which can cause congestion within multi-hop wireless sensor and ad hoc networks (WSANs). The capacity of the WSANs can be enhanced by using dual radios that allow concurrent use of multiple available wireless channels. It is a desirable feature that the enhanced capacity can be shared in a max-min fair manner by all existent flows in such multi-channel WSANs. In this paper, we propose a distributed resource allocation solution that achieves max-min fairness among multiple flows in multi-channel WSANs based on hybrid channel assignment. We find that the existence of two different types of enhanced network constraints in hybrid channel assignment-based multi-channel wireless networks leads to a new multi-constraint max-min resource allocation problem. We model the new max-min problem in the network utility maximization framework, with a particular focus on how to deal with resource prices induced by multi-constraints and adjust flow rates in response to the prices in a max-min fair manner. We present extensive simulation results to demonstrate the performance of the proposed distributed solution. We also discuss the trade-off between network throughput and fairness that exist in multi-channel WSANs. [ABSTRACT FROM AUTHOR]

In cellular network, channels are scarce radio resources. These resources are to be utilized judiciously. Though efforts have been made in the past for maximum channel utilization, studies show that effective channel utilization is often not done because of various constraints. The cognitive radio concept suggests that opportunistic channel utilization is possible if the mobile services are classified into two classes. The usage of the channels are based on the cognition that the radio derives. This work proposes a new model that applies a meta-heuristic Genetic Algorithm and the concept of cognitive radio for better channel utilization. The concept of single channel and multi-channel lending is also introduced towards this. Priority is given to the handoff services in comparison to the new services. The performance study, of the model, suggest the effective channel utilization in terms of blocked and dropped services. [ABSTRACT FROM AUTHOR]

Recently, wireless mesh network (WMN) technology recently has been used in various industries. Especially, a number of multi-channel assignment schemes have been presented to improve the throughput of IEEE 802.11-based multi-hop WMN. However, performance of the conventional multi-channel assignment schemes is not enough to satisfy the industry requirements. We, thus, should study more about the multi-channel assignment scheme in order to enhance the performance. This paper proposes a novel channel assignment scheme that employs multi-channel and multi-Interface in the WMN. The proposed scheme obtains the traffic information of the network and does the efficient channel assignment without any message exchanges. We verify the efficiency of the proposed scheme through the mathematical modeling and the real-world experiments. The results show that the proposed scheme improves the throughput of the network compared with the conventional schemes. [ABSTRACT FROM AUTHOR]

A number of multichannel medium access control (MAC) protocols have been developed earlier with the objective of maximizing energy efficiency and guaranteeing QoS requirements in wireless sensor networks. However, the trade-off between energy efficiency and QoS requirements remains an open issue. In this paper, we propose an energy efficient multichannel MAC protocol, so that nodes dynamically switch their interface between the channels to achieve this goal. The proposed protocol works in two phases, the advertisement (ADV) phase and data phase, which run concurrently. During the ADV phase, the nodes having data try to contend the advertisement channel to transmit the ADV packet. The nodes after receiving the acknowledgment from the receiver switch to the available free channel for the communication of the data packet. We analyze the performance of our proposed protocol through simulation. The simulation results reveal the proposed protocol performs better than the existing MAC protocols in terms of the energy efficiency, data delivery ratio, and end-to-end delay. [ABSTRACT FROM AUTHOR]

In this paper, we propose Multi-channel EMBA ( M-EMBA), efficient multihop broadcast for asynchronous multi-channel wireless sensor networks. Our scheme employs two channel-quality-aware forwarding policies of improved forwarder's guidance and fast forwarding to improve multihop broadcast performance. The improved forwarder's guidance allows forwarders to transmit broadcast messages with guidance to their receivers through channels with good quality. The guidance indicates how each receiver should forward the broadcast message to its neighbor nodes. The improved forwarder's guidance tremendously reduces redundant transmissions and collisions. Fast forwarding allows adjacent forwarders to send their broadcast messages simultaneously through different channels that have good quality, which helps to reduce multihop broadcast latency and improve multi-channel broadcast utility. In this work, we evaluate the multihop broadcast performance of M-EMBA through theoretical analysis of the system design and empirical simulation-based analysis. We implement M-EMBA in ns-2 and compare it with the broadcast schemes of ARM, EM-MAC, and MuchMAC. The performance results show that M-EMBA outperforms these protocols in both light and heavy network traffic. M-EMBA reduces message cost in terms of goodput, total bytes transmitted, as well as broadcast redundancy and collision. M-EMBA also achieves a high broadcast success ratio and low multihop broadcast latency. Finally, M-EMBA significantly improves energy efficiency by reducing average duty cycle. [ABSTRACT FROM AUTHOR]

Cognitive radio (CR) as a promising way to solve the spectrum scarcity allows exploitation of the shared frequency bands while guaranteeing acceptable interference to incumbent users in satellite communication systems. An improved spread slotted ALOHA (SSA) based on multi-user, multi-channel CR model applying to the satellite communications is proposed in this paper. To make full use of the detection information of satellite earth stations, a novel joint collaborative sensing method is used in the sensing phase. Moreover, a better throughput is achieved by using the improved SSA strategy in the transmission phase comparing with the traditional slotted ALOHA (SA). Theoretical analysis shows that the system performs better when SSA is adopted. Theoretical analysis and simulation results indicate that the sensing method used in this model outperforms the traditional 'hard combining' strategy in the whole sensing process. [ABSTRACT FROM AUTHOR]

This paper proposes an effective method for visual saliency detection based on multi-scale and multi-channel mean. 2-D wavelet transform is used to decompose and reconstruct image. Bicubic interpolation algorithm is applied to narrow the filtered image in multi-scale. We take the distances between the narrowed images and the means of their channels as saliency values, and we only reserve part values which are not less than the mean saliency of the given image. Bicubic interpolation algorithm is applied again to amplify the images in multi-scale, and then the saliency map is calculated by adding the amplified images. Finally, linear normalization is employed to obtain the final saliency map. Experimental results show that the proposed method outperforms 9 state-of-the-art methods both on the definition and accuracy of salient detection. [ABSTRACT FROM AUTHOR]

Underwater acoustic channels pose significant difficulty for the development of high speed communication due to highly limited band-width as well as hostile multipath interference. Enlightened by rapid progress of multiple input multiple output (MIMO) technologies in wireless communication scenarios, MIMO systems offer a potential solution by enabling multiple spatially parallel communication channels to improve communication performance as well as capacity. For MIMO acoustic communications, deep sea channels offer substantial spatial diversity among multiple channels that can be exploited to address simultaneous multipath and co-channel interference. At the same time, there are increasing requirements for high speed underwater communication in very shallow water area (for example, a depth less than 10 m). In this paper, a space-time multichannel adaptive receiver consisting of multiple decision feedback equalizers (DFE) is adopted as the receiver for a very shallow water MIMO acoustic communication system. The performance of multichannel DFE receivers with relatively small number of receiving elements are analyzed and compared with that of the multichannel time reversal receiver to evaluate the impact of limited spatial diversity on multi-channel equalization and time reversal processing. The results of sea trials in a very shallow water channel are presented to demonstrate the feasibility of very shallow water MIMO acoustic communication. [ABSTRACT FROM AUTHOR]

Wireless mesh networks (WMNs) consist of static nodes that usually have one or more radios or media. Optimal channel assignment (CA) for nodes is a challenging problem in WMNs. CA aims to minimize interference in the overall network and thus increase the total capacity of the network. This paper proposes a new method for solving the CA problem that comparatively performs more efficient than existing methods. The link layer in the TCP/IP model is a descriptive realm of networking protocols that operates on the local network link in routers discovery and neighboring hosts. TCP/IP employs the link-layer protocol (LLP) that is included among the hybrid states in CA methods, and learning automata are used to complete the algorithm with an intelligent method for suitable CA. We call this algorithm LLLA, which are short for LLP and learning automata. Our simulation results show that LLLA performs more efficient than ad hoc on-demand distance vector (AODV) types with respect to parameters such as packet drop, end-to-end delay, average goodput, jitter in special applications, and energy usage. [ABSTRACT FROM AUTHOR]

A new link layer protocol, quality of service hybrid multi channel protocol (QoS-HMCP), that takes into account the need for QoS support in multi-channel multi-interface wireless mesh networks (MCMI-WMNs) is proposed in this paper. QoS-HMCP protocol retains most properties of enhanced distributed channel access (EDCA) mechanism, defined by the 802.11 standard, as well as the basic mechanisms of HMCP protocol. To achieve good network performances with QoS-HMCP protocol in MCMI network, it was necessary to precisely define the channel coordination and procedure for the medium competition of EDCA functions (EDCAFs) after the expiration of interval of time within which a particular EDCAF has the right to initiate data exchange sequences onto the wireless medium-transmission opportunity. Taking into consideration the previously defined conditions, four versions of QoS-HMCP protocol are proposed in this paper. These versions were compared with each other and with basic HMCP with regard to the throughput of user data and delay through the network. The four versions of the QoS-HMCP protocol, multi channel routing (MCR) protocol and chosen metric were implemented in Glomosim simulator. The results have shown that implementation of QoS-HMCP protocol in MCMI-WMN resulted in a very small delay for voice and video traffic, while low-priority traffic categories (best effort and background) were still able to access the media. [ABSTRACT FROM AUTHOR]

In a regular wireless ad hoc network, the Medium Access Control (MAC) protocol coordinates channel access among nodes, and the throughput of the network is limited by the bandwidth of a single channel. The multi-channel MAC protocols can exploit multiple channels to achieve high network throughput by enabling more concurrent transmissions. In this paper, we propose a hybrid and adaptive protocol, called H-MMAC, which utilizes multi-channel resources more efficiently than other multi-channel MAC protocols. The main idea is to adopt the IEEE 802.11 Power Saving Mechanism and to allow nodes to transmit data packets while other nodes try to negotiate the data channel during the Ad hoc Traffic Indication Message window based on the network traffic load. The analytical and simulation results show that the proposed H-MMAC protocol improves the network performance significantly in terms of the aggregate throughput, average delay, fairness and energy efficiency. [ABSTRACT FROM AUTHOR]

In wireless mesh networks, the number of gateway nodes are limited, when the nodes access to the internet by fixed gateway node, different requirements of nodes lead to the dataflow shows heterogeneity. Many new routing metrics and algorithms existing in traditional wired networks and the Ad Hoc network, can not be directly applied to wireless mesh networks, so how to design a routing metric and algorithm which can dynamically adapt to current networks topology and dataflow changes, avoid bottleneck node, and select the most stable and least congestion link to establish a route is very important. In this paper, we presented a new dynamic adaptive channel load-aware metric (LAM) to solve the link load imbalance caused by inter-flow and inner-flow interference, designed a self-adaptive dynamic load balancing on-demand routing algorithm through extending and improving AODV routing method with the LAM, to achieve flow balance, reduce the high packet loss ratio and latency because congestion and Packet retransmission, and can increase Network Throughput. [ABSTRACT FROM AUTHOR]