Search

Preserving customers' expectations and understanding factors affecting their purchasing decisions are crucial in designing effective marketing and advertising strategies. However, constantly and swiftly changing the customers' interests and consumption behaviors make it inevitable to utilize sophisticated tools and approaches based on advanced technologies. Among them, by measuring the customers' physiological and neural signals, studying the customers' cognitive and affective responses to marketing stimuli, neuromarketing provides deep insight into the customers' motivations, preferences, and decisions. Recently, the internet of everything (IoE) has brought many new opportunities to the industry and has attracted the attention of many researchers in recent years. The main objective of this paper is to address how the IoE would empower neuromarketing techniques. Hence, an in-depth understanding of current research issues as well as emerging trends would help meet this goal. In this paper, a comprehensive review has been done by reviewing numerous journal and conference papers from various academic databases (sciencedirect.com, IEEE Xplore, Springer, Elsevier, Wiley, ACM digital library) based on applying various filtrations on specific keywords to identify and categorize the IoE devices in the neuromarketing field. In particular, we discussed the importance and the applications of IoE gadgets and devices, especially wearable medical technologies in eight neuromarketing techniques. Finally, we described critical existing challenges and limitations in neuromarketing and IoE gadgets that researchers have dealt with in this field of research. [ABSTRACT FROM AUTHOR]

This paper mainly studies the retrieval and location of violence in long time sequence video. Aiming at the low accuracy of violence detection in long time sequence video, a two-stage violence time sequence location method based on DC3D network model is proposed in this paper. In the video preprocessing stage, this paper adopts the method of generating video at multiple scales. In the first stage, we first use a large number of labeled datasets for pre-training to obtain a C3D network model for generating candidate videos and finally filter out meaningless background videos. In the second stage, we use the deconvolution method to identify the candidate video accurately to the frame level, so as to determine the specific time of violence. In the first stage, this method can improve the overall accuracy by generating candidate videos. DC3D network model can accurately locate the time of violence to the level of frame. The experimental results show that the proposed method can quickly retrieve and locate the violent fighting behavior in the surveillance video under the long time sequence video. The research results of this paper can provide convenience for the surveillance personnel to quickly retrieve and locate the target segment in a large amount of video data. [ABSTRACT FROM AUTHOR]

The number of research articles published on COVID-19 has dramatically increased since the outbreak of the pandemic in November 2019. This absurd rate of productivity in research articles leads to information overload. It has increasingly become urgent for researchers and medical associations to stay up to date on the latest COVID-19 studies. To address information overload in COVID-19 scientific literature, the study presents a novel hybrid model named CovSumm, an unsupervised graph-based hybrid approach for single-document summarization, that is evaluated on the CORD-19 dataset. We have tested the proposed methodology on the scientific papers in the database dated from January 1, 2021 to December 31, 2021, consisting of 840 documents in total. The proposed text summarization is a hybrid of two distinctive extractive approaches (1) GenCompareSum (transformer-based approach) and (2) TextRank (graph-based approach). The sum of scores generated by both methods is used to rank the sentences for generating the summary. On the CORD-19, the recall-oriented understudy for gisting evaluation (ROUGE) score metric is used to compare the performance of the CovSumm model with various state-of-the-art techniques. The proposed method achieved the highest scores of ROUGE-1: 40.14%, ROUGE-2: 13.25%, and ROUGE-L: 36.32%. The proposed hybrid approach shows improved performance on the CORD-19 dataset when compared to existing unsupervised text summarization methods. [ABSTRACT FROM AUTHOR]

In industry, robots are widely used to solve repetitive or dangerous actions in product production, so that product production can be more efficient. However, the problem that robots are often challenged is the convenience and the efficiency of introducing the production line. Therefore, the intuitive robot guidance method is an important issue; this paper will introduce the concept of human–robot interactions (HRI) and use deep learning methods on the machine vision system to complete the robot-guided assembly operation analysis, and the assembly operation analysis requires semantic segmentation as pre-processing. Therefore, we propose a novel semantic template correlation model architecture based on zero-shot learning (ZSL) to achieve the effect of rapid deployment. The semantic template correlation model is to search for the object area offline learning through the semantic template generated by the physics engine, and when inferring online, we can directly enter the semantic template to obtain the relevant object region. Finally, this paper verifies that the MIoU can be increased by more than 5% through the verification of the general database VOC2012. [ABSTRACT FROM AUTHOR]

The aim of this special issue is to introduce the burgeoning topic of Autonomous System. Finally, in order to check the conformity of this enforcement with defined multi-cloud SLA, the authors propose an approach for multi-cloud SLA reporting inspired by conformance checking techniques. I In i I the i I paper i I authored i I by i I Jeremy i I Mechouche i , the authors propose a hierarchical representation of multi-cloud SLAs: sub-SLAs associated with a system's components deployed on distinct cloud service providers and global-SLA associated with the whole system. [Extracted from the article]

Arithmetic logic unit (ALU), a core component of a processor, is one of the thrust areas of the current research. Presently, ALU is designed by transistor-based CMOS technique and its individual components are placed in different layers. The current design is affected by the limitations of Moore's law and design complexity. At present, 'Quantum cellular automata electro-spin (QCA-ES)' technology is widely accepted technology as an alternative of 'CMOS' to minimize the above discussed problems. In this research paper, the design of a novel multilayer portable, dynamic, fault-tolerant, power-efficient, thermally stable reversible ALU is proposed which is explored through QCA-ES. All the arithmetic and logical components of ALU are separately placed in different layers. Area density, delay, fault tolerance and thermal stability are investigated. A specific type of gate, known as reversible gate (modified 3:3 'TSG' gate), is used in this proposed design with QCA technology to get the optimized design ALU with low occupied area, complexity, delay and power dissipation. Investigation of a fault-free design and saturated amplitude level (of output) change with respect to temperature increment in the proposed device are also discussed in this paper. Not only the thermal stability (up to 6 k temperature) but also an investigation on cell complexity of the 100% fault-free (against multiple cell-omission, cell-displacement, cell-orientation change and extra cell deposition), multilayer nano-device is represented in this work. 'QCA-Designer' software is used in this research work to design and develop layout of the proposed components in quantum-field and find out the occupied area, delay and complexity of proposed design. 'QCA-Pro' software is used for getting the value of dissipated power. [ABSTRACT FROM AUTHOR]

Collecting and analyzing data from all devices to improve the efficiency of business processes is an important task of Industrial Internet of Things (IIoT). In the age of data explosion, extensive text data generated by the IIoT have given birth to a variety of text representation methods. The task of text representation is to convert the natural language to a form that computer can understand with retaining the original semantics. However, these methods are difficult to effectively extract the semantic features among words and distinguish polysemy in natural language. Combining the advantages of convolutional neural network (CNN) and variational autoencoder (VAE), this paper proposes an intelligent CNN-VAE text representation algorithm as an advanced learning method for social big data within next-generation IIoT, which help users identify the information collected by sensors and perform further processing. This method employs the convolution layer to capture the local features of the context and uses the variational technique to reconstruct feature space to make it conform to the normal distribution. In addition, the improved word2vec model based on topical word embedding (TWE) is utilized to add topical information to word vectors to distinguish polysemy. This paper takes the social big data as an example to illustrate the way of the proposed algorithm applied in the next-generation IIoT and utilizes Cnews dataset to verify the performance of proposed method with four evaluating metrics (i.e., recall, accuracy, precision, and F1-score). Experimental results indicate that the proposed method outperforms word2vec-avg and CNN-AE in K-nearest neighbor (KNN), random forest (RF), and support vector machine (SVM) classifiers and distinguishes polysemy effectively. [ABSTRACT FROM AUTHOR]

Whether it is an NLP (natural language processing) task or an NLU (natural language understanding) task, many methods are model oriented, ignoring the importance of data features. Such models did not perform well for many tasks based on feature loose, unbalanced tricky data including text classification tasks. In this regard, this paper proposes a classification method called LSTM-SN (long-short term memory RNN fusion social network) based on extremely complex datasets. The approach condenses the characteristics of the dataset. LSTM combines with social network methods derived from specific datasets to complete the classification task, and then use complex network structure evolution methods to discover dynamic social attributes. The experimental results show that this method can overcome the shortcomings of traditional methods and achieve better classification results. Finally, a method to calculate the accuracy of fusion model is proposed. The research ideas of this paper have far-reaching significance in the domain of social data analysis and relation extraction. [ABSTRACT FROM AUTHOR]

The Sunway TaihuLight is the first supercomputer built entirely with domestic processors in China. On Sunway Taihulight, the local data memory (LDM) of the slave core is limited, so data transmission with the main memory is frequent during calculation, and the memory access efficiency is low. On the other hand, for many scientific computing programs, how to solve the storage problem of irregular access data is the key to program optimization. Software cache (SWC) is one of the effective means to solve these problems. Based on the characteristics of Sunway TaihuLight structure and irregular access, this paper designs and implements a new software cache structure by using part of the space in LDM to simulate the cache function, which uses new cache address mapping and conflicts solution to solve high data access overhead and storage overhead in a traditional cache. At the same time, the SWC uses the register communication between the slave cores to share on the different slave core LDMs, increasing the capacity of the software cache and improving the hit rate. In addition, we adopt a double buffer strategy to access regular data in batches, which hides the communication overhead between the slave core and the main memory. The test results on the Sunway TaihuLight platform show that the software cache structure in this paper can effectively reduce the program running time, improve the software cache hit rate, and achieve a better optimization effect. [ABSTRACT FROM AUTHOR]

A. Kireeva et al. contributed another paper in the field of cellular automata: "Parallel simulation of drift-diffusion-recombination by cellular automata and global random walk algorithm". Parallel computing technologies enable the solution of large-scale numerical simulation and data processing problems in science and industry. The previously proposed contention-aware scheduling algorithm is formally described with this model, modifications of the algorithm are proposed: look-ahead technique, chunked data transfers, data caching, and peer-to-peer data transfers. [Extracted from the article]

Channel pruning has recently become a widely used model compression method. However, most existing channel pruning methods only prune to decrease the model size, such as the number of parameters or FLOPs, and hence the decrease in model size does not effectively lead to an improvement in inference performance. To address this problem, this paper proposes a latency-optimized channel pruning method for CNN inference acceleration on GPU platforms by latency stair-step discrimination, two-stage benefit assessment and latency-sharing channel pruning. Compared with recent state-of-the-art model compression methods, it can achieve significant improvements in inference performance with comparable compression rates and model accuracy. The contributions of this paper include the following: first, a three-point latency stair-step discrimination method is proposed for determining the candidate prunable coordinates with the best latency performance adapted to the current hardware. Then, a two-stage benefit assessment method based on interlayer dependencies is proposed for determining the optimal channel pruning rate of each layer in the network. Finally, a latency-sharing channel pruning framework is proposed to accelerate the model pruning adaptation process. The method proposed in this paper can significantly reduce the model inference latency on multiple types of GPU platforms. To verify the effectiveness, we use three general-purpose GPU platforms and two embedded GPU platforms to evaluate the algorithm performance. The experimental results show that for recent state-of-the-art CNNs, the proposed method can achieve a 22.0–6.6% latency reduction and a 1.3 –3.0 inference performance improvement as well as a 1.2–4.3 pruning adaptation speedup with high model accuracy. [ABSTRACT FROM AUTHOR]

Traditional deep learning models prefer large data sets, and in reality small data sets are easier to obtain. It is more practical to build models suitable for small data sets. Based on CNN, this paper proposes the concept of union convolution to build a deep learning model Union-net that is suitable for small data sets. The Union-net has small model size and superior performance. In this paper, the model is tested based on multiple commonly used data sets. The experimental results show that Union-net outperforms most models when dealing with small datasets, and Union-net outperforms other models when dealing with complex classification tasks or dealing with few-shot datasets. The codes for this paper have been uploaded to https://github.com/yeaso/union-net. [ABSTRACT FROM AUTHOR]

Inverse Synthetic Aperture Radar (ISAR) is widely used as an all-weather imaging radar system in military and civilian fields. However, for complex moving targets, the presence of their velocity, acceleration, and jerk leads to range migration and Doppler shift, which causes difficulties in imaging, especially in multi-target scenarios. In this paper, the adjacent cross-correlation function (ACCF) is used to reduce the migration order. Based on the idea of Lv transform and Hough transform (HT), an ISAR imaging algorithm for complex maneuvering multi-target is proposed. The algorithm proposed in this paper is simple and fast and can be accomplished by complex multiplication, fast Fourier transform (FFT) and inverse fast Fourier transform (IFFT) without the searching process. The simulation results show that the proposed algorithm can effectively compensate for the range migration (RM) and Doppler migration (DM) caused by the target motion parameters, and form a focused ISAR image. Compared with the existing methods, the algorithm has lower computational complexity and achieves ideal results in both anti-noise performance and imaging. [ABSTRACT FROM AUTHOR]

Actual trends in the real-time system field consists of migration towards complex central processing unit (CPU) architectures with enhanced execution predictability and rapid CPU contexts switch, thus obtaining high-performant control systems. The main objective of this paper is to present the results obtained following the implementation of real-time operating systems (RTOS) functions in hardware. Based on the CPU resource multiplication concept, actual researches has been focused on synthesizing in field-programmable gate array (FPGA) and implementing innovative solutions to improve RTOS performance. The results are materialized by validating an efficient hardware scheduler micro-architecture, from which a remarkable efficiency and a plus of performance and predictability are obtained. The experimental results, the FPGA resource requirements for implementation of the processor in different configurations, and the comparisons with other similar processor architectures are presented in order to verify theoretical aspects proposed through this paper. [ABSTRACT FROM AUTHOR]

Remaining useful life (RUL) prediction is critical for industrial equipment status detection, and the accurate prediction results provide decision-makers with actionable information. Preventive maintenance can be carried out to prevent the sudden failure of the equipment effectively based on the predicted results. However, as Industry 4.0 technologies develop, the amount of data collected by sensors is also rapidly increasing. The existing RUL prediction methods are gradually unable to cope with complex industrial equipment data, and deep learning methods gradually come to the fore. In this background, this paper proposes a transformer-based model with a multi-layer encoder–decoder structure to extract domain-invariant features. The decoder in the traditional transformer structure only obtains a single piece of information from the last layer of the encoder, and this paper uses an integrated layer-cross decoding strategy to compensate. Based on the encoder–decoder cross-connection, each decoder layer is provided with global view information from the final encoder layer simultaneously, improving the model's performance. The validity and superiorities of the proposed method are evaluated through several experiments on the publicly available C-MAPSS dataset provided by NASA. As can be seen from the results, the proposed method gets higher prediction accuracy than other network architectures and state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

With the gradual expansion of microservice architecture-based applications, the complexity of system operation and maintenance is also growing significantly. With the advent of AIOps, it is now possible to automatically detect the state of the system, allocate resources, warn, and detect anomalies using machine learning models. Given the dynamic nature of online workloads, the running state of a microservice system in production is constantly in flux. Therefore, it is necessary to continuously train, encapsulate, and deploy models based on the current system status for the AIOps model to dynamically adapt to the system environment. This paper proposes a model update and management pipeline framework for AIOps models in microservices systems in order to accomplish the aforementioned objectives and simplify the process. In addition, a prototype system based on Kubernetes and Gitlab is designed to provide preliminary framework implementation and validation. The system consists of three components: model training, model packaging, and model deploying. Parallelization and parameter search are incorporated into the model training procedure in order to facilitate rapid training of multiple models and automated model hyperparameter tuning. We automate the packaging and deployment process using technology for continuous integration. Experiments are conducted to validate the prototype system, and the results demonstrate the feasibility of the proposed framework. This work serves as a useful resource for constructing an integrated and streamlined AIOps model management system. [ABSTRACT FROM AUTHOR]

The target detection algorithm has a high accuracy rate for the detection of general objects. However, for specific kinds of objects, especially those with complex application scenarios and high requirements on recognition rates, the accuracy and real-time performance are still far behind the actual application requirements. Meanwhile, the high real-time performance of most current target detection algorithm depends on high-performance computers or Graphics Processing Units (GPUs), which limits their application in low-performance embedded system. In this paper, an embedded vehicle detection algorithm was proposed. Compared to the existing general detection algorithms, the network structure of the proposed algorithm has been improved in many aspects. Firstly, the overall network structure of YOLOv3-tiny has been improved, and the way of feature extraction has been modified to increase the network in terms of detection speed and recognition accuracy. Secondly, a spatial pyramid pooling structure is introduced into the network to enhance the learning capability, and the spatial pyramid pooling structure is also improved by increasing the number of pooling layers in the structure and cascading multiple groups of pooling layers of different sizes. Thirdly, to speed up target detection process, the original clustering algorithm is replaced by Kmean++ clustering algorithm, and the original IoU (The Intersection over Union) loss function is replaced by GIoU (Generalized IoU). Fourthly, according to the scale factor of batch normalization layer, the model is compressed and trimmed to remove redundant channels, and a small model more suitable for embedded system application is obtained. Finally, Raspberry Pi 4B and Inter NCS2 neural computing stick are used as the embedded system with the input image size of 416 × 416. Experimental tests on UA-DETRAC dataset obtain higher mAP than other advanced existing methods, indicating that the network model proposed in this paper could effectively improve the real time and accuracy of multi-vehicle target detection. [ABSTRACT FROM AUTHOR]

In a large-scale networking scenario with massive distribution of devices, data are independently generated and maintained by multiple domains. To solve the problem of isolated data islands in multi-domain, this paper designs a multi-domain data sharing model based on consortium chain. Aiming at the problems of low efficiency and lack of consideration of intra-domain consensus processes in traditional data sharing schemes, this paper optimizes the security and efficiency of data sharing from two parts: inter-domain chain and intra-domain chain. To ensure fine granularity, flexibility and security of the access control process in inter-domain chain, this model combines the Attribute-Based Access Control (ABAC) with smart contract, and a permission grading mechanism is designed to solve the problem of low retrieval efficiency in this access control model through reducing the retrieval range by grading and matching Policy set. Considering the problem of high delay of node consensus in the large-scale networking environment of intra-domain chain, a layered practical Byzantine fault tolerance optimization (LPBFT) algorithm with introducing a reputation model is designed, which greatly improves the efficiency and security of the intra-domain consensus. The consortium chain model is designed and implemented from the perspectives of inter-domain and intra-domain, which optimized the security and efficiency of the multi-domain data sharing model. Finally, a prototype system is implemented based on Hyperledger Fabric, and the validity of the model is verified from the aspects of inter-domain access control and intra-domain consensus algorithm. [ABSTRACT FROM AUTHOR]

In order to solve the problems of the basic mayfly optimization algorithm (MOA) in the field of robot path planning, such as slow convergence speed, poor accuracy, insufficient stability, and only applicable to static environment, a fusion algorithm combining improved mayfly optimization algorithm and dynamic window approach is proposed in this paper. Firstly, an improved mayfly optimization algorithm based on Q-learning (IMOA-QL) is proposed to solve robot global path planning problem. Q-learning is taken as the core of the improved mayfly optimization algorithm. For the basic MOA, the inertia weight and positive attraction coefficients are set to fixed values, which are not reasonable and will make the global search ability unbalanced, fall into local optimization easily, and also limit the iteration speed. In this paper, the parameters are adaptively adjusted based on Q-learning, and the appropriate parameters are selected according to the fitness of each mayfly. Meanwhile, the memory mechanism is introduced to speed up the convergence speed and implement the global path planning. Then, the global path nodes are extracted as the sub-target points, and the improved dynamic window approach is used to carry out the local path planning, which effectively improves the dynamic real-time avoidance ability. In order to verify the effectiveness of the proposed IMOA-QL algorithm in this paper, 20 random simulation experiments are carried out in the 100 × 100 static map environment and compared with the basic mayfly optimization algorithm (MOA) and the mayfly optimization algorithm based on linear adaptive inertia weight (MOA-LAIW). The results show that the average path length of the proposed IMOA-QL algorithm is reduced by 4.48% and 2.17% compared with MOA and MOA-LAIW in simple environment, and the average path length of the proposed IMOA-QL algorithm is reduced by 6.58% and 3.24% compared with MOA and MOA-LAIW in complex environment. In 20 experiments, the average variance of the proposed IMOA-QL algorithm in this paper is reduced by 74.15% and 57.67% compared with MOA and MOA-LAIW in simple environment, and the average variance of the proposed IMOA-QL algorithm is reduced by 51.22% and 38.67% in complex environment compared with MOA and MOA-LAIW. The simulation results show that the proposed IMOA-QL algorithm has significantly improved the accuracy and speed of solution. Moreover, dynamic obstacles are added in the static environment to carry out the simulation test of the fusion dynamic path planning algorithm. The results show that a fusion algorithm combining improved mayfly optimization algorithm and dynamic window approach in this paper can better complete the path planning task well in the complex dynamic environment. [ABSTRACT FROM AUTHOR]

Due to the rapid development of the wireless communication technology, the data volume of 5G mobile network continues to grow, which leads to the continuous reduction of signaling analysis and processing efficiency. To overcome the problem, the intelligent communication will be one of the mainstream directions of mobile communication development which combines with 5G and AI. In this paper, we introduce a method of using machine learning classification and signaling analysis technology. The proposed method in this paper is an improved signaling analysis algorithm based on naive Bayesian classification, which improves the signaling classification accuracy of the algorithm. In the signaling analysis process of the algorithm, the key messages of signaling data are selected as user characteristics before the association and synthesis of user signaling processes. Then, the supervised signaling feature classification model is trained according to the user ID in the signaling data, and the model is used for signaling classification. Subsequently, aiming at the zero probability problem in the algorithm, we use Laplace smoothing to correct it. Then, the algorithm continues to use orthogonal matrix to make orthogonal transformation on continuous attributes and discrete attributes after numerical marking, so as to enhance the independence between attributes. It closes to the assumption of naive Bayes, which improves the classification accuracy of the algorithm. The experimental results show that the improved algorithm model has high comprehensive performance index and good classification performance. The F1 score of this algorithm reaches 67.86%, and achieves the expected effect. The improved signaling analysis method proposed in this study is expected to be useful in 5G network optimization and testing. [ABSTRACT FROM AUTHOR]

This paper presents a prototype filter design using the orthant optimization technique to assist a filter bank multicarrier (FBMC) modulation scheme of a NextG smart e-healthcare network framework. Low latency and very high reliability are one of the main requirements of a real-time e-healthcare system. In recent times, FBMC modulation has gotten more attention due to its spectral efficiency. The characteristics of a filter bank are determined by t's, prototype filter. A prototype filter cannot be designed to achieve an arbitrary time localization (for low latency) and frequency localization (spectral efficiency), as time and frequency spreading are conflicting goals. Hence, an optimum design needed to be achieved. In this paper, a constraint for perfect or nearly perfect reconstruction is formulated for prototype filter design and an orthant-based enriched sparse ℓ1-optimization method is applied to achieve the optimum performance in terms of higher availability of subcarrier spacing for the given requirement of signal-to-interference ratio. Larger subcarrier spacing ensures lower latency and better performance in real-time applications. The proposed FBMC system, based on an optimum design of the prototype filter, also supports a higher data rate as compared to traditional FBMC and OFDM systems, which is another requirement of real-time communication. In this paper, the solution for the different technical issues of physical layer design is provided. The presented modulation scheme through the proposed prototype filter-based FBMC can suppress the side lobe energy of the constituted filters up to large extent without compromising the recovery of the signal at the receiver end. The proposed system provides very high spectral efficiency; it can sacrifice large guard band frequencies to increase the subcarrier spacing to provide low-latency communication to support the real-time e-healthcare network. [ABSTRACT FROM AUTHOR]

The back-projection algorithm can adapt to synthetic aperture radar imaging in any mode because of its time-domain imaging characteristics, so it has received increasing attention in scientific research and engineering practice. However, the large enormous of computation of the back-projection algorithm limits its application. Since the back-projection algorithm has a high degree of data parallelism, it is very suitable for GPU parallel processing. The data scale and hardware environment in different application scenarios will result in different bottlenecks which may prevent the back-projection algorithm from performing optimally. Thus, this paper proposes a series of strategies to aid the back-projection algorithm in achieving the best peak performance with any given problem scale on any given platform. A heuristic block strategy is proposed to optimize the peak performance of the back-projection algorithm on servers and miniaturized GPU devices, which can handle the differences in hardware platforms as well as the differences in data scales. To optimize the peak performance of the back-projection algorithm on miniaturized GPU devices, a memory management strategy is proposed by using unified memory and pinned host memory. The method proposed in this paper has achieved an acceleration ratio close to the device's peak performance on servers and miniaturized GPU devices. By using the proposed methods, any algorithm with data-independent chunking and any GPU device with a uniform memory architecture can achieve higher peak performance. [ABSTRACT FROM AUTHOR]

Due to the problem of high link load of edge cache and small storage space of edge server, a caching architecture by the collaborative of edge nodes and the cloud server is proposed. The content cache location is designed and optimized, which can be the content provider, cloud server (CS), and edge node (EN). In the proposed system, cloud servers collaborate with edge servers and the performance of content caching can be improved by coordinating caching on the cloud server or caching on the edge server. In this paper, a cloud-edge collaborative caching model based on the greedy algorithm is proposed, which includes the content caching model and collaborative caching model. Network architecture, file popularity estimation, link capacity, and other factors are considered in the model. Correspondingly, a cloud-edge collaborative cache algorithm based on a greedy algorithm is proposed. The related optimization problem is decomposed into the knapsack problem of cache layout in each layer, and then the greedy algorithm is used to solve the knapsack problem of cache placement and cooperative cache proposed in this paper. The affiliation between CS cache and EN caches in the layered architecture is improved and recognized. In the experimental results, the link load is reduced, the cache hit rate is improved by using the proposed method of edge caching, and it also has obvious advantages in the average end-to-end service delay. [ABSTRACT FROM AUTHOR]

To realize the great flexibility and cost savings for providing location-based service, data owners are incentivized to migrate their data to cloud servers. However, direct data outsourcing to untrusted servers may pose significant privacy risks. This paper proposes a practical privacy-preserving nearest neighbor searching method over encrypted spatial data. We simultaneously protect data and location privacy (access and pattern privacy) by encrypting data using asymmetric scalar-product-preserving encryption (ASPE) and performing computational private information retrieval (CPIR) on encrypted subspace datasets. To mitigate the performance degradation introduced by the combination of ASPE and CPIR, we propose a hierarchical index that enables users to safely obtain encrypted subspace datasets with configurable privacy, where different degrees of privacy can be traded off against query processing performance. Experiments demonstrate that our method outperforms the state-of-the-art method in efficiency while allowing for a flexible trade-off between performance and privacy. [ABSTRACT FROM AUTHOR]

Forest fire is natural disasters that are sudden, destructive and difficult to handle and rescue, with millions of hectares of forests burned every year all over the world, causing serious ecological damage, loss of life and property. Therefore, timely detection and treatment of early fire is of positive and important significance for forest fire early control. The fire detection method based on image processing is one of the most important means of preventing the occurrence of large-scale forest fires at present by extracting the flame and smoke features in the image and quickly determining the location of the fire. The current deep learning-based forest fire early detection methods have problems such as high false alarm rate due to small detection targets and complex environmental background, and large number of detection model parameters. In view of this, this paper proposes a lightweight early forest fire and smoke detection method based on GS-YOLOv5. Firstly, this paper proposes a novel Super-SPPF structure to replace the SPPF structure in YOLOv5, by which the output of the feature extraction network is used as input, and the input is divided into two branches to retain more semantic information, and the GhostConv operation is performed separately to reduce the number of model parameters. The Super-SPPF structure performs the serial MaxPooling operation on one of the branches, which improves the computation speed by choosing a smaller pooling kernel, and then fuses the outputs of the two branches to reduce the false alarm rate of the detection model. Secondly, C3Ghost is utilized instead of the C3 module in YOLOv5 to further reduce the number of detection model parameters. Finally, the coordinate attention (CA) module is introduced in backbone of YOLOv5 to obtain the relationship between channels and space, which enables the network to obtain the location information of interest more accurately and further improves the detection accuracy of early fires. In this paper, a self-constructed DL-Fire dataset is used to verify the performance of the GS-YOLOv5 detection model by collecting environmental interference samples and combining them with the D-Fire dataset. The experimental results show that the detection accuracy of GS-YOLOv5 is 95.9%, and the model size is 10.58 mb. Compared with YOLOv5, false alarm rate is reduced from 12% to 6% and computational complexity is reduced from 16.0 GFlops to 12.8 GFlops by GS-YOLOv5. [ABSTRACT FROM AUTHOR]

Reconfigurable hardware, extensively employed in mission-critical digital applications like space and military electronics due to its adaptability, encounters the issue of soft errors, especially in control path elements, which could result in functional failure. Various system-level fault tolerance methodologies exist, and this paper implements a bio-inspired fault tolerance technique called evolvable hardware (EHW). The preferred implementation of the EHW system involves hosting the evolutionary algorithm on the processor alongside the reconfigurable hardware. However, this approach encounters delays in the intercommunication of the evolved circuit between the reconfigurable hardware and the processor. To address this issue, the paper proposes a two-tier architecture to achieve absolute fault mitigation in the controller. In this architecture, Tier-1 involves the digital implementation of the genetic algorithm on the reconfigurable hardware to mitigate errors in the controller, while Tier-2 focuses on mitigating errors occurring in Tier-1. The aim is to establish an absolute and self-resilient controller hardware to mitigate faults. The study simulates faults at the target circuit and genetic module as a proof of concept. The proposed two-tier single event upset (SEU) mitigation is deployed on Microsemi's ProAsic3e FPGA (Field Programmable Gate Array), achieving an average efficiency of 91%. This efficiency is accompanied by ten times lesser resource utilization compared to traditional methodologies and a 30% accelerated speed when compared to hybrid evolvable systems. [ABSTRACT FROM AUTHOR]

To investigate how the petroleum industry evaluates and selects powerful cloud service providers, first, an evaluation index system including 25 indices such as scalability and private data protection is built. This index system can systematically examine the comprehensive strengths of cloud service providers. Aiming to solve the problems that the traditional expert evaluation method has high requirements on expert experience and is easily affected by subjective factors, a novel artificial intelligence evaluation model named TentISSA-BPNN is proposed. The objective evaluation ability of this model can be effectively used in evaluation research on cloud service providers for petroleum enterprises. In this model, the SSA algorithm is optimized by Tent chaotic mapping and adaptive inertia weight; an algorithm, TentISSA, that has good stability and fast convergence speed is designed and proposed; and the BPNN is improved with TentISSA to obtain more accurate evaluation results. To evaluate the performance of the TentISSA algorithm, nine unimodal and multimodal functions are selected in this paper to test the convergence accuracy. Then, seven models are selected as the control groups to validate the effectiveness and performance of the TentISSA-BPNN evaluation model proposed in this paper. Finally, the preprocessed data of the candidate cloud service providers are input into the trained neural network model proposed in this paper for evaluation. Based on the ranking of the evaluation scores, the comprehensive strengths of the cloud service providers are obtained to provide a decision-making reference for managers of petroleum enterprises in the process of choosing cloud service providers. [ABSTRACT FROM AUTHOR]

Wireless sensor networks (WSNs) are mostly used for monitoring the environment; however, they are usually powered by non-rechargeable batteries with limited energy. Solar energy harvesting is an attractive solution to the limit by charging the sensor nodes; however, the harvested solar energy is easily affected by weather conditions. Based on the characteristics of uncertainty and intermittency of solar energy, this paper proposes a plus-profile solar energy prediction algorithm. This algorithm makes the prediction of future available solar energy by finding the data in the dataset that is most similar to the data of the day and combining it with recent weather trend. According to the predicted result, the paper further proposes an adaptive energy management scheme to suit the harvested energy. In the scheme, sensor nodes can adaptively adjust task scheduling to achieve energy neutrality. The simulation results show that compared with other algorithms, the prediction accuracy of the proposed prediction algorithm is improved by 17.7 and 22.4%, respectively, and the proposed energy management scheme reduced energy loss by 6.2 and 46.8%, respectively. [ABSTRACT FROM AUTHOR]

In order to explore the application of neural networks in the intelligent protection of digital music copyrights, this paper initially investigates the key aspects of digital music-related supply chain management (SCM). Subsequently, this paper introduces a fuzzy, comprehensive evaluation-based method for authenticating digital music infringement. Furthermore, it employs a convolutional neural network to analyze digital music's source components. Within this context, an attention mechanism and an adaptive gate model are meticulously designed. Moreover, a selective adaptive cascade is implemented to optimize the structure of the multi-resolution coder–decoder. Through fuzzy authentication, it becomes feasible to accurately and reliably track and verify copyright information pertaining to digital music, thus providing an effective protective mechanism for the digital music industry. Experimental results reveal that in comparison with the SHN-4 series models, the network model embedded with the channel space (CS) attention module achieves improvements of up to 0.04 dB in terms of performance metrics for human voice and accompaniment sources. Furthermore, it outperforms these models by up to 0.19 dB in music source separation metrics. The incorporation of the adaptive attention mechanism profoundly enhances the model's performance in audio source separation tasks, benefiting both vocal and accompaniment sources. This improvement is evident in the form of a 0.44 SDR enhancement for vocal sources and a 0.18 SDR enhancement for accompaniment sources. Additionally, the proposed two-stage music separation-gate self-attention model surpasses the two-stage music separation-gate model in metrics associated with human voice and accompaniment sources. This outcome convincingly underscores the enhanced effectiveness achieved by integrating CS, self-attention, and adaptive connection in network performance, as well as the adaptive gate structure for music copyright protection. This approach heightens the model's autonomous selectivity, minimizes outliers, stabilizes waveform characteristics, and significantly contributes to the safeguarding of digital content. [ABSTRACT FROM AUTHOR]

In order to solve the problems of slow convergence and low accuracy of the traditional ant colony algorithm in solving the traveling salesman problem (TSP), this paper proposes a heterogeneous ant colony algorithm based on the selective evolution mechanism and game strategy (SGHACA). First, we propose a game mechanism. In each iteration, the best population and the worst population will play a game. At the end of the game, pheromone gains will be given to the participants according to the weight coefficients to motivate the populations to participate in the next round of the game, in order to promote cooperation among different populations and thus improve the diversity of the algorithm. Second, we introduce a selective evolution mechanism. The optimal paths of each iteration are compared with the historical optimal paths to evolve relatively better path fragments, thus improving the convergence speed of the algorithm. Finally, when the algorithm stagnates, we introduce a dynamic shaking mechanism, which is able to dynamically dither the pheromone on the current optimal path according to the number of iterations, increasing the probability that the algorithm chooses other paths, thus helping to jump out of the locally optimal solution. In order to verify the effectiveness of the improved algorithm in this paper, we experimentally verify through a large number of TSP instances that SGHACA is able to improve the convergence of the algorithm and the accuracy of the solution to a certain extent when solving large-scale TSP. Meanwhile, compared with the latest improved algorithms, the accuracy of the solution of SGHACA is better than most of the latest improved algorithms, which indicates that the improved algorithm of this paper has some competitiveness. [ABSTRACT FROM AUTHOR]

While traveling, people need to utilize smart tourism service systems based on the Internet of Things (IoT) to enhance their journeys' convenience. However, existing IoT-based smart tourism service systems still possess certain shortcomings. In light of this, this paper delves into the integration of IoT and machine learning (ML) technologies to improve the design of smart tourism service systems and establish a platform for intelligent tourism services. Simulated datasets are employed to evaluate the performance and functionality of the smart tourism service system. The primary contributions, as inferred from the test results, are as follows: In area 1, the model achieves a congestion recognition rate of 92.6% with an error rate of only 1.5% in congestion judgment. In area 2, congestion is accurately identified 93.5% of the time, with a misjudgment rate of just 1.2%. In area 3, the model achieves a congestion identification rate of 94.2% along with a 1.3% misjudgment rate, effectively identifying congestion situations in scenic spots. When comparing the research and design system with the system based on the support vector machine (SVM) algorithm, the results demonstrate that the smart tourism service system designed in this paper, leveraging IoT technology and ML algorithms, achieves higher prediction accuracy than the SVM-based system. These findings can serve as a valuable guide for designing future smart tourism service systems. [ABSTRACT FROM AUTHOR]

Existing biohashing authentication schemes usually store bioinformation in the cloud, whose storage method can face the problems of bioinformation leakage and tampering, and single point of failure. To address the above problems, a ciphertext speech biohashing authentication scheme based on chaotic system and improved public chain is proposed. First, the fusion features are extracted and generated with the inner product of 3D chaotic system to generate the biohashing sequence. Second, the original speech is encrypted and uploaded to the InterPlanetary file system and then, uploaded to the blockchain after the data are encrypted. Finally, the server matches the hash sequence queried on the chain with the to-be-authenticated hash at the user's end using the Hamming distance. The experimental results show that the biohashing algorithm in this paper has better distinguishability, robustness, security and time efficiency. The key space of the speech encryption algorithm is 2 256 , which can well resist the exhaustive attack and ensure the security during speech authentication. In this paper, the combination of on-chain and off-chain storage of speech information can prevent the problem of bioinformation leakage and tampering, as well as avoiding the problem of single point of failure. [ABSTRACT FROM AUTHOR]

In the stochastic and dynamic edge-cloud collaborative environment, the computing resources of the host are limited, and the resource requirements of computing tasks are random and changeable. Therefore, how to efficiently schedule dynamic tasks and improve system performance becomes challenging. The scheduling algorithm based on deep reinforcement learning optimizes the delay and energy consumption of the system by dynamically interacting with the environment, which can solve the problem of dynamic and changeable environment to a certain extent, but there are still problems such as poor model adaptability, low training efficiency, and unbalanced system load. In this paper, aiming at optimizing the average response time of task scheduling and the average energy consumption of the system, a multi-objective task scheduling model is designed, and a task scheduling policy optimization algorithm based on improved asynchronous advantage actor-critic (A3C) is proposed. The residual convolutional neural network (RCNN) improves the network structure of A3C, using asynchronous multi-threaded training methods to interact with the edge-cloud collaborative environment and capturing the random dynamic characteristics of resources required for computing tasks and heterogeneous edge-cloud hosts resource change characteristics to better adapt to random dynamic environments, and it can quickly updating network parameters, improving training speed, and make full use of host resources to solve system load imbalance problems. Simulation results show that the scheduling algorithm proposed in this paper can effectively reduce task response time and system energy consumption in an edge-cloud collaborative environment. [ABSTRACT FROM AUTHOR]

The reliable prediction of solar energy production and surplus is crucial for the stability of the electricity grid and effective energy distribution, particularly during peak consumption periods. Encouraging solar adoption by customers is also important despite fluctuations in solar energy generation. This paper addresses the need for an intelligent surplus solar energy prediction model, referred to as the Green Model, which aims to accurately predict surplus electrical energy that can be returned to the distribution grid. The Green Model utilizes a combination of a developed long short-term memory (LSTM) neural network and a novel optimization technique called deterministic selection network by gain sharing knowledge (DSN-GSK). The DSN-GSK optimizes the structure of LSTM by determining the optimal number of hidden layers, nodes in each layer, biases, weights, and activation functions based on the most important features that represent the relationship between generated solar energy and weather factors. The Green Model was evaluated and analysed, demonstrating high accuracy in the short-term prediction of surplus electrical energy with a minimal percentage of error. The model shows promising prospects for the development of returned electrical energy prediction. The main novelty of this paper lies in the development of the Green Model, which combines LSTM and DSN-GSK to enhance the accuracy of surplus solar energy prediction. The benefits of the Green Model include improved stability and efficiency of the electricity grid, effective utilization of renewable energy sources, and the reduction of negative environmental impacts. The contributions of this paper include advancing the field of intelligent energy prediction, optimizing the structure of LSTM, and providing valuable insights for the integration of renewable energy into the existing energy infrastructure. [ABSTRACT FROM AUTHOR]

The early and effective diagnosis of Alzheimer's disease (AD) and mild cognitive impairment (MCI) has received increasing attention in recent years. However, currently available deep learning methods often ignore the contextual spatial information contained in structural MRI images used for early diagnosis and classification of Alzheimer's disease. This may lead us to miss important structural details by failing to adequately capture the potential connections between each slice and its neighboring slices. This lack of contextual information may cause the accuracy of the network model to suffer, which in turn affects its generalization ability and application in real-life scenarios. To explore deeper the connection between spatial context slices, this research is designed to develop a new network model to effectively detect or predict AD by digging into the deeper spatial contextual structural information. In this paper, we design a spatial context network based on 3D convolutional neural network to learn the multi-level structural features of brain MRI images for AD classification. The experimental results show that the model has good stability, accuracy and generalization ability. Our experimental method had a classification accuracy of 92.6% in the AD/CN comparison, 74.9% in the AD/MCI comparison, and 76.3% in the MCI/CN comparison. In addition, this paper demonstrates the effectiveness of the proposed network model through ablation experiments. [ABSTRACT FROM AUTHOR]

This paper solves the task of unsupervised video object segmentation (UVOS) that segments the objects of interest through the entire videos without any annotation. In recent years, many unsupervised video object segmentation (UVOS) methods have been proposed. Although these methods perform well, they rely on networks with heavy weights, often leading to large model size. In order to reduce the model size while keeping a competitive performance, we propose a saliency-based dual-attention (SDA) method for UVOS in this paper. In our method, we take optical flow and video frames as inputs and extract the appearance information and motion information from optical flow and video frames. We design a two-branch network with appearance information and motion information. The information from these two branches is fused via a saliency-based dual-attention module to segment the primary object in one path. The saliency-based dual-attention module is composed of saliency attention and saliency-based reverse attention. To demonstrate the effectiveness of our network, we tested it on the DAVIS-2016 and SegtrackV2 datasets. Experimental results demonstrate that our method can achieve competitive results in terms of accuracy and model size. [ABSTRACT FROM AUTHOR]

Aiming at the problems of color compensation bias and loss of detail information in local areas of current underwater image enhancement methods, this paper proposes an underwater attenuation image enhancement method with adaptive color compensation and detail optimization. The method fully considers the attenuation level of each optical channel to guide the color correction based on the attenuation image, and introduces a brightness adjustment method to give the output image a good natural appearance. Gradient-oriented local contrast enhancement and multi-scale edge optimization methods are used to process the color- corrected image separately to obtain two clear images with balanced natural colors, high contrast and good preservation of detail information and then combine with the multi-scale fusion process Artifact-free image fusion is achieved. The experimental results on the UIEB dataset show that the method in this paper improves the UIQM and BRISQUE scores by 14.32% and 8.53%, respectively, and has fewer free parameters, which can effectively enhance attenuated image contrast, detail information and balance image color. In addition, the method in this paper has a complete feature point matching and target detection application, which can be used as a general framework for image preprocessing tasks. [ABSTRACT FROM AUTHOR]

Sign language synthesis technology is a bridge of communication between deaf and non-deaf people, and has important social value. Sign language recognition is the key to sign language synthesis technology, which is to acquire gesture data containing sign language information through hardware equipment and convert it into corresponding sound or text through relevant technology. The representation of sign language by computer virtual 3D mannequin is sign language synthesis. This paper reports on the construction of a video frame sequence based on interpolation transition video frame to improve the readability of the composition of sign language. Transfer of video frames is traditionally performed via the target tracking algorithm for gesture tracking arithmetic, but this can lead to several tracking errors because of the complexity of the hand background environment, the covering of the hand, multiple perspectives, the great similarity of different sign language words, and the interaction between the hand and other parts of the body. Therefore, in the present paper, in order to improve the accuracy of target tracking, a more accurate gesture tracking method is studied. Particle filter and optical flow algorithms are combined to build a more natural and smooth 3D transition frame image, which makes video-based sign language synthesis more realistic. [ABSTRACT FROM AUTHOR]

Cloud computing is seen as a more promising technology than any other traditional information technology computing paradigm in today's world. It essentially functions as an on-demand resource provisioning platform that requires no active user participation. The resource provisioning strategies necessitate proper load distribution management across the cloud network, without which the cloud would experience biased workload performance. Today virtualization is the cornerstone of cloud computing, allowing data dissemination and administration via deploying virtual machines. Modern applications contain data that need to be stored into a scheme called polyglot persistence (combining SQL and NoSQL data-stores). However, these services are tailored to specific storage requirements, necessitating aggregating them from several heterogeneous clouds or migrating data from one cloud to another. Data migration can be done offline where the database is independent of the application, or otherwise, the application has to be down for the migration period. This paper developed a middleware in.NET Core facilitating the live migration of persistent polyglot data in heterogeneous clouds. This paper presents the proof of concept for live migration of the database layer of an application hosted on any supported clouds to any implemented cloud's data-store. Our suggested technique performs better in migration time, energy usage, and throughput aspects as compared with the offline migration scenario. In our experimentation, we found that while migrating data in offline mode from SQL to mongo and vice versa there is a marginal increase of 29% and 11%, respectively, in latency time. This increase is acceptable and tolerable while considering the live data migration scenario. [ABSTRACT FROM AUTHOR]

With the rapid development of Internet technology, network attacks occur frequently and numerous hidden dangers appear in network security. Therefore, improving the performance of intrusion detection systems to detect and defend against attacks is the key to ensuring network security. However, in the face of complex and massive network data feature information, traditional machine learning methods suffer from data imbalance and feature redundancy, which results in low detection rates, high false alarm rates and poor real-time performance of intrusion detection systems. Therefore, to address these problems, this paper proposes a data imbalance-based Convolutional Neural Network Intrusion Detection Method (CNN-IDMDI). First, an oversampling method is used to solve the data imbalance problem by decomposing the increased number of samples for the few attacks with multiple sampling to form multiple sub-samples. Second, the gradient coordination mechanism and the improved loss function Focal Loss are combined to calculate the loss between the actual and expected values to detect network malicious attacks in high-dimensional and unbalanced data. Finally, the methods in this paper are compared with the current mainstream intrusion detection methods on the NSL-KDD dataset for binary and multi-classification detection. The experimental results show that the method in this paper can effectively improve the effectiveness of CNN intrusion detection and network anomaly. The average accuracy of the CNN intrusion detection method based on data imbalance for binary intrusion detection is 98.73% and the implementation time of the method is 1.42 s, which is 15.45%, 12.76%, and 2.91% higher than the average accuracy of the CNN, the CNN Long Short-Term Memory (CNN-LSTM) and the CNN Neural-induced Support Vector Machine (CNN-NSVM) methods, respectively, and the detection time is saved by 0.82 s, 0.72 s, and 0.61 s, respectively. The average accuracy of the CNN intrusion detection method based on data imbalance for multi-classification intrusion detection is 94.55% and the time required to complete the detection is 2.96 s. This improves the average accuracy by 16.09%, 12.71%, and 3.66% compared with the CNN, CNN-LSTM and CNN-NSVM methods, respectively. It is also quicker, as the time consumption of CNN is 8.84 s, CNN-LSTM is 8.31 s and CNN-NSVM is 6.43 s. Therefore, the CNN-IDMDI method for intrusion detection proposed in this paper has higher accuracy and faster speed. [ABSTRACT FROM AUTHOR]

This paper proposes field-programmable gate array (FPGA) acceleration on a scalable multi-layer perceptron (MLP) neural network for classifying handwritten digits. First, an investigation to the network architectures is conducted to find the optimal FPGA design corresponding to different classification rates. As a case study, then a specific single-hidden-layer MLP network is implemented with an eight-stage pipelined structure on Xilinx Ultrascale FPGA. It mainly contains a timing controller designed by Verilog Hardware Description Language (HDL) and sigmoid neurons integrated by Xilinx IPs. Finally, experimental results show a greater than × 10 speedup compared with prior implementations. The proposed FPGA architecture is expandable to other specifications on different accuracy (up to 95.82%) and hardware cost. [ABSTRACT FROM AUTHOR]

Cooperative computation offloading (CCO) is a technique to improve computation offloading performance in edge networks through collaboration between edge nodes. CCO can achieve better resource utilization, balance the computational load and further reduce delay to improve the service experience of user equipment (UE). In this paper, we investigate the problem of collaborative computing task offloading scheme and computing resource allocation in mobile edge computing and propose a data compression cooperation computing offloading (DCCO) scheme. To reduce the amount of data transmitted on the UE offloading link, we introduce a Data Compression into CCO and present the computational offloading strategy, collaborative offloading and computational resource allocation problems with the goal of minimizing the weighted sum of delay and energy consumption of the UE under the constraints of UE delay and energy consumption. And an improved genetic algorithm is proposed to solve the problem which is a non-convex mixed-integer problem with binary and continuous variables. The offloading strategy and computational resource allocation correspond to the genes in the genetic algorithm chromosome. The simulation results show that the DCCO scheme can reduce the offloading cost up to 11% compared with the existing schemes. It effectively improves the computation offloading performance of edge network computing. [ABSTRACT FROM AUTHOR]

With the advancements in mobile technology, a large amount of data is generated by end devices, which has created a renewed interest in developing new AI-based applications for gaining insights into these data. However, most of these distributed applications need data aggregated at a central server, which has posed severe bandwidth, latency, security, and privacy issues. This paper presents OSGAN (One-Shot distributed learning algorithm using Generative Adversarial Networks), a generic framework that trains a generative adversarial network (GAN) at each client and uses GAN's generative capabilities to create sample data at the server. The server aggregates these data from various clients, builds a deep learning model and sends its parameters back to the clients in one communication round, i.e. the exchange of information between the clients and the server happens only once. In this paper, we present the design and implementation of OSGAN and evaluate its performance by comparing it with the state-of-the-art federated learning (FL) and central training algorithms for both IID and non-IID distribution of data. Our experiments on multiple datasets show that our proposed approach achieves a similar accuracy when compared with both FL and central training algorithms. Specifically, the accuracy drop with OSGAN is maintained within 2% for multiple datasets and multiple numbers of clients. Our results show that the proposed approach reduces the amount of data transfer by almost 98% when compared with federated learning and close to 80% when compared with the central learning approach thereby providing substantial benefit in terms of saving the bandwidth. [ABSTRACT FROM AUTHOR]

Sentiment analysis involves extricating and interpreting people's views, feelings, beliefs, etc., about diverse actualities such as services, goods, and topics. People intend to investigate the users' opinions on the online platform to achieve better performance. Regardless, the high-dimensional feature set in an online review study affects the interpretation of classification. Several studies have implemented different feature selection techniques; however, getting a high accuracy with a very minimal number of features is yet to be accomplished. This paper develops an effective hybrid approach based on an enhanced genetic algorithm (GA) and analysis of variance (ANOVA) to achieve this purpose. To beat the local minima convergence problem, this paper uses a unique two-phase crossover and impressive selection approach, gaining high exploration and fast convergence of the model. The use of ANOVA drastically reduces the feature size to minimize the computational burden of the model. Experiments are performed to estimate the algorithm performance using different conventional classifiers and algorithms like GA, Particle Swarm Optimization (PSO), Recursive Feature Elimination (RFE), Random Forest, ExtraTree, AdaBoost, GradientBoost, and XGBoost. The proposed novel approach gives impressive results using the Amazon Review dataset with an accuracy of 78.60 %, F1 score of 79.38 %, and an average precision of 0.87, and the Restaurant Customer Review dataset with an accuracy of 77.70 %, F1 score of 78.24 %, and average precision of 0.89 as compared to other existing algorithms. The result shows that the proposed model outperforms other algorithms with nearly 45 and 42% fewer features for the Amazon Review and Restaurant Customer Review datasets. [ABSTRACT FROM AUTHOR]

Imperceptibility, robustness, and reliability are essential characteristics of image watermarking. In this paper, two image watermarking schemes, first Lifting wavelet transform (LWT)–Discrete cosine transform (DCT) with zigzag scanning-based embedding in Schur domain and second LWT-Schur decomposition-based embedding scheme in Schur domain itself, are proposed. In this paper, two schemes are proposed in which a pre-processed watermark image is generated using LWT and Schur decomposition, where the principle diagonal coefficients of the upper triangular matrix are embedded into the cover image in LWT-Schur domain. The remaining matrix, after removing diagonal components and other unitary matrix of Schur decomposition of LL sub-band of watermark image are used as keys which provide better security to the watermarking scheme. In the LWT-Schur-based scheme principle diagonal coefficients of upper triangular matrix obtained from the Schur decomposition of the LL sub-band of the cover image are modified with the pre-processed watermark in Schur domain. In the LWT–DCT-based scheme, the low and mid-band frequency coefficients are used to embed the watermark using zigzag scanning to avoid any artifacts like diagonal line problem present in the SVD-based watermarking scheme, especially for larger value of the embedding gain factor. The Schur decomposition has lesser computation complexity as compared to SVD and both the schemes are also free from False positive problem. The LWT–DCT with zigzag scanning-based scheme offers higher robustness against filtering and noise attacks as Normalized correlation coefficient (NCC) value is close to 1.0000. Both the proposed schemes provide better imperceptibility and higher robustness as NCC value of more than 0.9800 against various checkmark attacks. Performance analysis shows that these schemes perform better in terms of Peak signal-to-noise ratio, and NCC under most of the attacks and also have lesser computational time than existing schemes. [ABSTRACT FROM AUTHOR]

The positions of sensor nodes in wireless sensor networks (WSNs) plays important role in order to achieve optimum values of various parameters, such as coverage, localization, etc. WSNs are used in every field such as in agriculture, military etc. and to achieve optimum performance of WSNs in these fields the sensor nodes are required to deploy at optimized positions. In this paper the jaya-sine cosine optimization algorithm (Jaya-SCOA) is proposed to deploy the sensor nodes at optimum positions and to increase the coverage of sensor nodes. Further the sensor nodes are deployed in three different topologies such as C shape, outer boundary and random topology and coverage rate is computed for all topologies. The major goal of the paper is to find the best topology for deploying the sensor nodes in WSNs and to enhance the coverage of nodes. The simulation results of proposed Jaya-SCOA is compared with fruit fly optimization algorithm (FFOA) and bat optimization algorithm (BOA) for all topologies in terms of maximum and minimum coverage, standard deviation and mean coverage rate. The simulation results demonstrate that for all topologies the coverage rate of Jaya-SCOA is higher than that of FFOA and BOA and average coverage rate of Jaya-SCOA for random topology is increased by 14.55% and 27.04% as compared to outer boundary and C shape topology, respectively. Therefore, the random topology is the best topology as compared to other topologies and Jaya-SCOA is effective than BOA and FFOA for all topologies in terms of coverage and stability. [ABSTRACT FROM AUTHOR]

The constant growth of social media, unconventional web technologies, mobile applications, and Internet of Things (IoT) devices create challenges for cloud data systems in order to support huge datasets and very high request rates. NoSQL databases, such as Cassandra and HBase, and relational SQL databases with replication, such as Citus/PostgreSQL, have been used to increase horizontal scalability and high availability of data store systems. In this paper, we evaluated three distributed databases on a low-power low-cost cluster of commodity Single-Board Computers (SBC): relational Citus/PostgreSQL and NoSQL databases Cassandra and HBase. The cluster has 15 Raspberry Pi 3 nodes with Docker Swarm orchestration tool for service deployment and ingress load balancing over SBCs. We believe that a low-cost SBC cluster can support cloud serving goals such as scale-out, elasticity, and high availability. Experimental results clearly demonstrated that there is a trade-off between performance and replication, which provides availability and partition tolerance. Besides, both properties are essential in the context of distributed systems with low-power boards. Cassandra attained better results with its consistency levels specified by the client. Both Citus and HBase enable consistency but it penalizes performance as the number of replicas increases. [ABSTRACT FROM AUTHOR]

The performance growth in processors has been continuing toward increasing the number of processing cores on the chip and scaling the feature size of transistors. However, in the nanoera, side effects of the scaling, such as induced heterogeneities in the performance, power, and soft error rate of identically designed cores, prevent the potential performance from being fully utilized. In this paper, we harness the mentioned side effects in shared-memory multicore processors with unknown workloads by a dynamic heuristic scheduling algorithm called HDSAP. HDSAP aims to maximize performance, i.e., the average response time, under power and reliability constraints in presence of induced heterogeneities. In this regard, we use a mathematical model to quantify task to core assignments based on performance variation. We also consider the variation in power to change selected cores when the power constraint is missed. To meet the reliability constraint, we use N-modular redundancy while being aware of the variation in the soft error rate of cores to prevent under/over reliability estimation. To evaluate HDSAP, we run SPLASH benchmark suite on Sniper and MACPat simulators. As a result, the response time of HDSAP reduces by 6%, 8%, and 25% in comparison with similar algorithms under the same power and reliability constraints. [ABSTRACT FROM AUTHOR]

Electronic health records provide a vast amount of text health data written by physicians as patient clinical notes. The world health organization released the international classification of diseases version 10 (ICD-10) system to monitor and analyze clinical notes. ICD-10 is system physicians and other healthcare providers use to classify and code all diagnoses and symptom records in conjunction with hospital care. Therefore, the data can be easily stored, retrieved, and analyzed for decision-making. In order to address the problem, this paper introduces a system to classify the clinical notes to ICD-10 codes. This paper examines 7541 clinical notes collected from a health institute in Jordan and annotated by ICD-10's coders. In addition, the research uses another outsource dataset to augment the actual dataset. The research presented many approaches, such as the baseline and pipeline models. The Baseline model employed several methods like Word2vec embedding for representing the text. The model structure also involves long-short-term memory a convolutional neural network, and two fully-connected layers. The second Pipeline approach adopts the transformer model, such as Bidirectional Encoder Representations from Transformers (BERT), which is pre-trained on a similar health domain. The Pipeline model builds on two BERT models. The first model classifies the category codes representing the first three characters of ICD-10. The second BERT model uses the outputs from the general BERT model (first model) as input for the special BERT (second model) to classify the clinical notes into total codes of ICD-10. Moreover, Baseline and Pipeline models applied the Focal loss function to eliminate the imbalanced classes. However, The Pipeline model demonstrates a significant performance by evaluating it over the F1 score, recall, precision, and accuracy metric, which are 92.5%, 84.9%, 91.8%, and 84.97%, respectively. [ABSTRACT FROM AUTHOR]

With the migration of enterprise applications to microservices and containers, cloud service providers, starting with Amazon in 2014, announced a new computational model called function-as-a-service. In these platforms, developers create a set of fine-grained functions with shorter execution times instead of developing coarse-grained software. In addition, the management of system resources and servers is delegated to cloud service providers. This model has many benefits, such as reducing costs, improving resource utilization, and helping developers focus on the core logic of applications. But, still faces many challenges such as cost/usage balance, optimizing performance, programming models, using current development tools, containers' cold start problem, saving data in caches, security issues, privacy concerns, and scheduling challenges like execution time prediction. In this paper, we focus on scheduling and cold start problems. Reducing the cold start time can result in better response times and hence a better experience for end-users. Compromise occurs when keeping warm operating environments which can reduce cold start times while increasing resource usage costs. The cold start problem is wildly studied, but in this work, we propose a novel dynamic waiting-time adjustment approach in which the waiting time of a container will change dynamically according to the situation. In this study, four different types of decisions for changing the waiting time are introduced and investigated. We aim to create a trade-off by using these decisions at runtime using a heuristic method. Functions' invocation frequency, function dependency graph, maximum merging time, correct mergers, cost, and timeline are considered for the decision-making process. The performed evaluations demonstrate that the proposed approach results in a 32% improvement over the fixed-time method (i.e., the method used by Apache OpenWhisk). This comparison is made from the cumulative measurement viewpoint which is a combination of response time, turnaround time, cost, and utilization. [ABSTRACT FROM AUTHOR]