No abstract available.

Deep learning and big data analytics can be attributed to recent trends and opportunities in many research activities and areas such as bioinformatics, beyond 5G and 6G communications, healthcare, internet of things (IoT), manufacturing business and social networks. Big data analytics and deep learning are sought-after and fastest-growing techniques for the enhancement of information and communication technology (ICT), and recent approaches are providing unexpected solutions that once seemed unachievable. Applications of big data analytics and deep learning in 5G and 6G are able to facilitate many new features in network management and operations, and 5G and beyond communication systems are expected to provide services with massive connectivity, ultra-low latency, extremely high security, extremely low energy consumption, and ultra-high data-rate. The main focus of this special issue is on deep learning and big data analytics to process and analyze data in 5G and 6G applications. The special issue includes novel studies about big data, IoT, Industry 4.0 applications, machine learning, deep learning solutions, 5th generation, and 6th generation technologies. There were in total nineteen papers accepted for publication in this special issue through careful peer review and revisions, and all are covered under the overarching theme of deep learning-based intelligent communication systems. The summary of every topic is given below. However, it is strongly encouraged to read the full paper if interested. Saeed et al., in their paper 'A comprehensive review on the users' identity privacy for 5G networks', aim to shed light on the survey about user privacy for 5G networks, which continues the identity and location privacy. Also, it discusses most of the studies which handle the user identifications in authentication, paging, and location update. The paper discusses various privacy issues in 5G network which use IMSI in clear text or join the temporary identities: TMSI and C-RNTI with IMSI to disclose the privacy of user indentity. After that the paper studies many proposed solutions which discuss user privacy (identity and location) and concludes that each of these studies has advantages and disadvantages for its proposed solutions. Elfatih et al., in their paper 'Internet of vehicle's resource management in 5G networks using AI technologies: Current status and trends', discuss and provide a comprehensive detail for resource allocation and management for IoV over 5G RAN network utilizing AI techniques. In addition, an extensive discussion of AI technologies that promise to be adopted and contributed to IoV and V2X applications is presented. The presented reviews in these areas have not taken into account the significance of integrating the multi-layers of vehicular network architecture for each AI strategy and how to be tailored for rapid and dynamic topology problems. Hence, this paper adresses these problems by describing how sophisticated and deep vehicle network architecture can be enhanced by AI techniques for layer-by-layer resources management and allocation problems. Hasan et al., in their paper 'A review on security threats, vulnerabilities, and counter measures of 5G enabled internet-of-medical-things', review the applications of the internet of medical things (IoMT) that has gained major attention as an ecosystem of connected clinical systems, computing systems, and medical sensors geared towards improving the quality of healthcare services. The 5G based AI technology can revolute the perception of healthcare and lifestyle. In light of the importance of IoT platforms and 5G networks, the purpose of this proposed research work is to identify threats that could undermine the integrity, privacy, and security of IoMT systems. Also, the novel blockchain-based approaches can help in improving the confidentiality of the IoMT network. It has been discovered that IoMT is vulnerable to various types of attacks, including denial of service (DoS), malware, and eavesdropping attack. In addition, IoMT is exposed to various vulnerabilities, such as security, privacy, and confidentiality. Le in his paper 'A comprehensive survey of imbalanced learning methods for bankruptcy prediction', gives a review about imbalanced learning methods. This study first reviews several state-of-the-art approaches for handling this problem in bankruptcy prediction, including an over sampling based (OSB) framework, a cost-sensitive method (the C Boost algorithm), a combination of resampling techniques and a cost-sensitive framework, and an ensemble-based model (the XGBS algorithm). The author also conducts empirical experiments to evaluate the methods surveyed here in terms of two performance metrics; the area under the ROC curve and the geometric mean. The results show that the ensemble-based model outperforms other methods in terms of bankruptcy prediction on the KB dataset. Poongodi et al., in their paper '5G Based blockchain network for authentic and ethical keyword search engine', carry out a proposed 5G-based blockchain network architecture for an encrypted keyword search engine. The suggested model permits to play out all connections amongst different users and mini-base stations through the use of diverse access nodes points and network brokers. It also helps to comprehend the complete application of blockchain technology, wherein the distribution of numerous digital ledgers and smart contracts were acknowledged between each network entity. Moreover, complete utilization of cryptocurrency is realized at essential points of the network layer to lessen the effect of interference rate and streamline the spectrum sharing when requested by the user. Alshammari et al., in their paper 'Technology-driven 5G enabled e-healthcare system during COVID-19 pandemic', reveal that most people receive information from social networking sites, health professionals, and television without facing any challenges. The analysis shows that, during the COVID-19 pandemic, about 42% of respondents felt tense always or most of the time on a daily basis. Only 28.6% of respondents felt tense sometimes, whereas the remainder (about 30%) did not feel tense in relation to the COVID-19 crisis. Satisfaction with COVID-19-related information is also positively correlated with COVID-19-related information literacy (r = 0.53, p < 0.01) that is also positively correlated with depression or emotion, anxiety, and stress (r = -0.15, p < 0.05). The long-term pandemic is creating several psychological symptoms including anxiety, stress, and depression, irrespective of age. Natarajan et al., in their paper 'An IoT and machine learning-based routing protocol for reconfigurable engineering application', present an upgradable cross-layer routing protocol based on CR-IoT to improve routing efficiency and optimize data transmission in a reconfigurable network. In this context, the system is developing a distributed controller which is designed with multiple activities, including load balancing, neighbourhood sensing and machine-learning path construction. The proposed approach is based on network traffic and load and various other network metrics including energy efficiency, network capacity and interference, on average of 2 bps/Hz/W. The trials are carried out with conventional models, demonstrating the residual energy and resource scalability and robustness of the reconfigurable CR-IoT Pandey et al., in their paper 'Lyapunov optimization machine learning resource allocation approach for uplink underlaid D2D communication in 5G networks', formulate the maximization of uplink and overall system capacity with resource management, which guarantees the signal to interference noise ratio for the D2D users. The optimization is a mixed-integer non-linear problem that uses the Lyapunov optimization method to optimize the BER value and an iterative algorithm to optimize the power value with different constraints. After attaining the optimized value, SVM (support vector machine) technique is utilized to ensure the spectral efficiency of the overall system in autonomous mode. Simulation results show that the proposed method provides higher reliability and power efficiency with higher system capacity in comparison to prevailing technologies. Liang et al., in their paper 'A new model path for the development of smart leisure sports tourism industry based on 5G technology', adopt the literature method to learn the theoretical basis of 5G technology and smart tourism in depth, establish a multi-dimensional resource allocation model for the smart leisure sports tourism industry, and conduct research on the influencing factors, information sources, channel factors and other aspects of the tourism industry. The general public's search for tourism strategies and attractions, food and specialty products, the use of online search information channels accounted for 70.3% and 69.3%, which further shows that the development of 5G technology has promoted the transformation and development of the sports tourism industry. Khan et al., in their paper '3D convolutional neural networks based automatic modulation classification in the presence of channel noise', consider the problem of multiclass (eight classes) classification of modulated signals (binary phase shift keying, quadrature phase shift keying, 16 and 64 quadrature amplitude modulation corrupted by additive white Gaussian noise, Rician and Rayleigh fading channels) using architectures in both frequency and spatial domains while deploying three approaches for data augmentation, such as random zoomed in/out, random shift and random weak Gaussian blurring augmentation techniques with a cross-validation (CV) based hyperparameter selection statistical approach. Simulation results testify the performance of 10-fold CV without augmentation in the spatial domain to be the best while the worst performing method happens to be 10-fold CV without augmentation in the frequency domain and learning in the spatial domain to be better than learning in the frequency domain. Chen et al., in their paper 'Resource electronic database for measuring regional cultural influence based on machine learning big data', aim to build a resource electronic database for measuring regional cultural influence through the current hot big data technology, and to provide some reference suggestions and data resources for the harmonious development of regional culture. In this article, the authors investigate the current cultural development in various regions of China and its impact on the development of Chinese culture and the culture of the world through literary research. Considering the current state of cultural development in the region, this article determines the key functional requirements for building an electronic database of cultural impact measurement resources in the region. In the specific process of designing the database, big data mining algorithms and machine learning classification and prediction algorithms are used to collect, categorize and process the data resources of the regional cultural influence measurement database. In the analysis of the measurement of regional cultural influence, this paper uses the regional cultural pattern index to evaluate and predict the distribution, concentration, prosperity and influence of regional culture. Duggal et al., in their paper 'A sequential roadmap to Industry 6.0: Exploring future manufacturing trends', scroll through patent pathways and intellectual developments throughout industrial revolutions listing significant products and services that landmarked each revolution up to Industry 4.0. The research pools of Industry 4.0 are classified and explored. A lack of human–machine workforce synergy in Industry 4.0 and the nascent 'customized manufacturing' concept is addressed in subsequent sections. The paper classifies two expected phases of Industry 5.0, highlighting the subdomains touted to be its focal areas. Gourisaria et al., in their paper 'Data science appositeness in diabetes mellitus diagnosis for healthcare systems of developing nations' use various machine learning, deep learning, and data dimensionality reduction techniques to detect diabetes mellitus. The research is principally conducted on two datasets, first from the Frankfurt hospital, Germany, second from the UCI repository. Models such as support vector machines, naïve Bayes, and random forests are implemented to classify diabetic patients from non-diabetic ones. Subsequently, after hyperparameter tuning, a comparative study on the results is done and the most prominent model promoted. This process is repeated for the datasets with reduced dimensionality using linear discriminant analysis (LDA) and principal component analysis (PCA). For the Frankfurt, Germany dataset, k-nearest neighbours showed the best accuracy of 98.2%, and the random forest classifier for the UCI repository showed 99.2%. Abbasi et al., in their paper 'An intelligent method for reducing the overhead of analyzing big data flows in OpenFlow switch', focus on developing a dynamic replacement method. This intelligent method utilizes the statistical features of the traffic flows in the table to select a table for replacement and makes use of the popularity of flows in the flow table for replacing entries and updating the flow table. The method aims to evaluate the existing entries according to the history of the activities of the flow, which was neglected in previous studies. For this purpose, the author uses the 'importance' feature which has been introduced in OpenFlow 1.4. Mohanty et al., in their paper 'Identification and evaluation of the effective criteria for detection of congestion in a smart city', propose a novel congestion detection system based on the combination of k-means clustering and analytical hierarchy process. A transport network is created in the simulation of urban mobility (SUMO) simulator. After receiving the parameters of vehicles from the simulator in a congested junction area, the key parameters are extracted by using the k-means clustering technique and mathematical mean algorithm. This key parameter is utilized in analytical hierarchy process to detect the highest priorities parameter. Based on that parameter the congestion is detected in a particular lane. Yadav et al., in their paper 'A secure data transmission and efficient data balancing approach for 5G based IOT data using UUDIS-ECC and LSRHS-CNN algorithms', propose a technique that contains authentication, destination selection, validation, secure DT, and also LB phases. The user and also the device are permitted to send the data towards the destination if they are authenticated. The UDDIS-ECC is employed for secure DT. For improving the SL, the SiP hash function is utilized and the 5G IoT data is balanced by employing the LSRHS-CNN algorithm. By deeming the input data's tasks, the LB is managed. Afterward, the performance analysis is conducted. Analysis for secure DT and also LB are the two parts wherein the analysis is handled. Centred upon the ET, DT, along with SL, the proposed UUDIS-ECC is analogized with the existent ECC, RSA, DES, and ECDSA in the secure DT analysis. Moorthy et al., in their paper 'Reduction of satellite images size in 5G networks using machine learning algorithms', propose a method which is implemented with a combination of intra-coding and machine learning algorithms. The standard compression technique does not give better results due to degradation of pixels, lack of spatial and spectral information. This paper enriches progressive results by reducing satellite images for transmission of data in IoT and 5G wireless networks, which qualitative results are compared by standard compression technique with suitable parameters. Li in his paper 'SWOT analysis of e-commerce development of rural tourism farmers' professional cooperatives in the era of big data', analyzes the e-commerce development strategy of China's rural tourism cooperatives in detail, and uses the analytic hierarchy process to analyze the establishment of the green development of e-commerce tourism business, affecting external opportunities and threats. This makes it possible to explore the sustainable development path for the follow-up development of e-commerce tourism business, which is conducive to the sustainable development path of rural tourism e-commerce tourism, and achieves multi-win business, environmental and social benefits. The experimental results of this paper show that through the calculation of the quadrangle of my country's tourism e-commerce enterprise development strategy, M1 = 0.0089, M2 = 0.0029, M3 = 0.0012, M4 = 0.0038, and M1 > M4 > M2 > M3 can be obtained. Singh et al., in their paper 'LoRa based intelligent soil and weather condition monitoring with internet of things for precision agriculture in smart cities', present the design of an intelligent irrigation system based on soil and weather conditions. The soil and weather parameters are selected through various research articles in Agriculture 4.0 and ML. The paper also juxtaposes the designed weather station with various patents developed. The system developed in this paper provides a cost-effective and state-of-the-art solution to local weather monitoring. Rohit Sharma is an associate professor in the Department of Electronics and Communication Engineering, SRM Institute of Science and Technology, India. He is an active member of ISTE, IEEE, ICS, IAENG, and IACSIT. He is an editorial board member and reviewer for more than 12 international journals and conferences, including IEEE Access and IEEE Internet of Things Journal. He has served as a book editor for seven different titles to be published by CRC Press, Taylor & Francis Group, USA and Apple Academic Press, CRC Press, Taylor & Francis Group, USA, and Springer. He has received the Young Researcher Award at the 2nd Global Outreach Research and Education Summit & Awards 2019 hosted by the Global Outreach Research & Education Association (GOREA). He has served as a guest editor in the SCI journal of Elsevier. He has actively organized various international conferences. He has served as an editor and organizing chair to the 3rd Springer International Conference on Microelectronics and Telecommunication (2019), IEEE International Conference on Microelectronics and Telecommunication (2018), IEEE International Conference on Microelectronics and Telecommunication (ICMETE-2016), and technical committee member of CSMA2017, EEWC 2017, IWMSE2017, ICG2016, and ICCEIS2016. Qin Xin received his PhD from the Department of Computer Science at the University of Liverpool, UK in December 2004. Currently, he is working as a professor of Computer Science and Faculty Research Leader in the Faculty of Science and Technology at the University of the Faroe Islands (UoFI), Faroe Islands. Prior to joining UoFI, he had held various research positions in world-leading universities and research laboratories including a Senior Research Fellowship at Universite Catholique de Louvain, Belgium, Research Scientist/Postdoctoral Research Fellowship at Simula Research Laboratory, Norway and Postdoctoral Research Fellowship at the University of Bergen, Norway. His main research focus is on design and analysis of sequential, parallel and distributed algorithms for various communication and optimization problems in wireless communication networks, as well as cryptography and digital currencies including quantum money. Moreover, he also investigates the combinatorial optimization problems with applications in Bioinformatics, Data Mining and Space Research. Currently, Prof. Dr. Xin is serving on the management committee board of Denmark for several EU ICT projects. Prof. Dr. Xin has produced more than 111 peer reviewed scientific papers. His works have been published in leading international conferences and journals, such as ICALP, ACM PODC, SWAT, IEEE MASS, ISAAC, SIROCCO, IEEE ICC, Algorithmica, Theoretical Computer Science, Distributed Computing, IEEE Transactions on Computers, Journal of Parallel and Distributed Computing, IEEE Transactions on Dielectrics and Electrical Insulation, IEEE Transactions on Sustainable Computing, ACM Transactions on Internet Technology, IEEE Transactions on Network Science and Engineering, ACM Transactions on Asian and Low-Resource Language Information Processing, and Advances in Space Research. He has been very actively involved in the services for the community in terms of acting (or acted) on various positions (e.g., Session Chair, Member of Technical Program Committee, Symposium Organizer and Local Organization Co-chair) for numerous international leading conferences in the fields of distributed computing, wireless communications and ubiquitous intelligence and computing, including IEEE MASS, IEEE LCN, ACM SAC, IEEE ICC, IEEE Globecom, IEEE WCNC, IEEE VTC, IFIP NPC, IEEE Sarnoff and so on. He is the organizing committee chair for the 17th and 18th Scandinavian Symposium and Workshops on Algorithm Theory (SWAT 2020 and SWAT 2022, Torshavn, Faroe Islands). Currently, he also serves on the editorial board for more than ten international journals. Patrick Siarry received the Ph.D. degree in computer science and optimization from University Paris VI, France, in 1986, and the Doctorate of Sciences (Habilitation) degree in computer science and optimization from University Paris XI, France in 1994. He was first involved in the development of analog and digital models of nuclear power plants with Electricité de France, Paris. Since 1995, he has been a Professor of Automatics and Informatics with Université Paris-Est Créteil, France. His main research interests include computer-aided design of electronic circuits, cognitive intelligence, and the applications of new stochastic global optimization heuristics to various engineering fields, also including the fitting of process models to experimental data, the learning of fuzzy rule bases, and of neural networks. Wei-Chiang Hong is a professor in the Department of Information Management at the Oriental Institute of Technology, Taiwan. His research interests mainly include computational intelligence (neural networks and evolutionary computation) and applications of forecasting technology (ARIMA, support vectorregression, and In his paper was as by In he was to be as Professor by the Science and Technology In the he was the Young Researcher by in the of and and In he was in the of In 2017, he was in the of Deep Communication Big Data

Abstract Human faces contain useful information that can be used in the identification of age, gender, weight etc. Among these biometrics, body mass index (BMI) and body weight are good indicators of a healthy person. Motivated by the recent health science studies, this work investigates ways to identify malnutrition affected people and obese people by analyzing body weight and BMI from facial images by proposing a regression method based on the 50‐layers Residual network architecture. For face detection, Multi‐task Cascaded Convolutional Neural Networks have been employed. A system is created to evaluate BMI along with age and gender from human facial real‐time images. Malnutrition and obesity are commonly determined with the help of BMI. In the previous works, height, weight, and BMI estimation through automatic means have predominantly focused on full‐body images and videos of humans. The usage of facial images for estimating such traits have been given less importance. In order to facilitate the analysis, the dataset is cleaned along with metadata containing information about the persons height, weight, age, and gender. Gender‐based analysis is performed for the prediction of BMI. Finally, an email containing the persons picture along with their details is sent to the concerned health officer.

The fifth industrial revolution (Industry 5.0) integrates humans and machines to satisfy the increasing customization demands of the manufacturing complexity using an optimized robotized manufacturing process. Industry 5.0 make use of collaborative robots (cobots) for optimizing productivity and ensuring safety. At the same time, unmanned aerial vehicles (UAVs) are predicted to be the main part of industry 5.0 in the forthcoming days. Regardless of high mobility and energy-limited UAVs for wireless communication as significant advantages, different issues are also existing in the UAV networks, such as security, reliability, etc. Several research works have focused on resolving security issues in UAV communication to support safety-critical applications. With this motivation, this article presents an artificial intelligence-based UAV-borne secure communication with classification (AIUAV-SCC) framework for industry 5.0 environment. The proposed AIUAV-SCC model involves two major phases namely image steganography-based secure communication and deep learning (DL)-based classification. At the initial stage, a new image steganography technique with multilevel discrete wavelet transformation, quantum bacterial colony optimization based optimal pixel selection, and encryption processes take place. Next, in the second stage, the Bayesian optimization (BO)-based SqueezeNet model is applied for the classification of securely received UAV images where the parameters in the SqueezeNet method are optimally tuned by the utilize of the BO technique. To validate the performance of the presented model, extensive simulations are applied using the UC Merced dataset (UCM) aerial dataset and the outcomes are investigated under several dimensions. The outcomes make sure the goodness of the presented model on test UCM aerial dataset over the compared methods.

introduction Share on Guest Editorial Introduction for the Special Section on Deep Learning Algorithms and Systems for Enhancing Security in Cloud Services Editors: Gunasekaran Manogaran Howard University, Washington D.C., USA Howard University, Washington D.C., USAView Profile , Hassan Qudrat-Ullah York University, Toronto, Canada York University, Toronto, CanadaView Profile , Qin Xin University of the Faroe Islands, Faroe Islands University of the Faroe Islands, Faroe IslandsView Profile , Latifur Khan The University of Texas at Dallas, Texas, USA The University of Texas at Dallas, Texas, USAView Profile Authors Info & Claims ACM Transactions on Internet TechnologyVolume 22Issue 2May 2022 Article No.: 39epp 1–5https://doi.org/10.1145/3516806Online:14 May 2022Publication History 0citation49DownloadsMetricsTotal Citations0Total Downloads49Last 12 Months49Last 6 weeks6 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

The Internet of Vehicles (IoV) communication platform provides seamless information exchange facilities in a dynamic mobile city environment. Heterogeneous communication is a common medium for information exchange through autonomous resources distributed and accessed using infrastructure units. Cyber-security is a primary concern in accessing autonomous information from the distributed resources due to anonymity and different types of targeted adversaries. This article proposes token-based authorization and authentication (TAA) for securing IoV communications. The proposed method relies on blockchain technology and random forest learning for authorization and key management for authentication, respectively. In this process, frequent change in tokens and key update features are restricted in a view to maximize the seamlessness in information exchange. Authentication is preceded by knowledge of the data classification without errors to prevent additional overhead. Blockchain-based authorization helps to update specific fields of the tokens to retain the communication ratio by reducing vehicle-to-vehicle losses. The performance of the proposed method is assessed using appropriate simulations for these metrics by varying vehicle density, error rate, and classification sets.

Linguistic resources for commonly used languages such as English and Mandarin Chinese are available in abundance, hence the existing research in these languages. However, there are languages for which linguistic resources are scarcely available. One of these languages is the Hindi language. Hindi, being the fourth-most popular language, still lacks in richly populated linguistic resources, owing to the challenges involved in dealing with the Hindi language. This article first explores the machine learning-based approaches—Naïve Bayes, Support Vector Machine, Decision Tree, and Logistic Regression—to analyze the sentiment contained in Hindi language text derived from Twitter. Further, the article presents lexicon-based approaches (Hindi Senti-WordNet, NRC Emotion Lexicon) for sentiment analysis in Hindi while also proposing a Domain-specific Sentiment Dictionary. Finally, an integrated convolutional neural network (CNN)—Recurrent Neural Network and Long Short-term Memory—is proposed to analyze sentiment from Hindi language tweets, a total of 23,767 tweets classified into positive, negative, and neutral. The proposed CNN approach gives an accuracy of 85%.

No abstract available.

research-article Share on Introduction to the Special Issue on Deep Structured Learning for Natural Language Processing Authors: Gunasekaran Manogaran Big Data Scientist, University of California, Davis, USA Big Data Scientist, University of California, Davis, USAView Profile , Hassan Qudrat-Ullah Professor of Decision Sciences, School of Administrative Studies, York University, Toronto, Canada Professor of Decision Sciences, School of Administrative Studies, York University, Toronto, CanadaView Profile , Qin Xin Full Professor of Computer Science, Faculty of Science and Technology, University of the Faroe Islands, Faroe Islands. Full Professor of Computer Science, Faculty of Science and Technology, University of the Faroe Islands, Faroe Islands.View Profile Authors Info & Claims ACM Transactions on Asian and Low-Resource Language Information ProcessingVolume 20Issue 3May 2021 Article No.: 37epp 1–3https://doi.org/10.1145/3474087Online:01 September 2021Publication History 0citation38DownloadsMetricsTotal Citations0Total Downloads38Last 12 Months38Last 6 weeks2 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

introduction Introduction to the Special Issue on Deep Structured Learning for Natural Language Processing, Part 3 Share on Editors: Gunasekaran Manogaran View Profile , Hassan Qudrat-Ullah View Profile , Qin Xin View Profile Authors Info & Claims ACM Transactions on Asian and Low-Resource Language Information ProcessingVolume 20Issue 5September 2021 Article No.: 72epp 1–3https://doi.org/10.1145/3476464Published:31 August 2021 0citation11DownloadsMetricsTotal Citations0Total Downloads11Last 12 Months11Last 6 weeks8 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

Cloud data centers (CDCs) have become increasingly popular and widespread in recent years with the growing popularity of cloud computing and high-performance computing. Due to the multi-step computation of data streams and heterogeneous task dependencies, task failure frequently occurs, resulting in poor user experience and additional energy consumption. To reduce task execution failure as well as energy consumption, we propose a novel AI-driven energy-aware proactive fault-tolerant scheduling scheme for CDCs in this paper. First, a prediction model based on the machine learning approach is trained to classify the arriving tasks into “failure-prone tasks” and “non-failure-prone tasks” according to the predicted failure rate. Then, two efficient scheduling mechanisms are proposed to allocate two types of tasks to the most appropriate hosts in a CDC. The vector reconstruction method is developed to construct super tasks from failure-prone tasks and separately schedule these super tasks and non-failure-prone tasks to the most suitable physical host. All the tasks are scheduled in an earliest-deadline-first manner. Our evaluation results show that the proposed scheme can intelligently predict task failure and achieves better fault tolerance and reduces total energy consumption better than the existing schemes.

Processing skyline queries in incomplete data is challenging, particularly, for a database with dynamic contents in which the database is frequently updated. These update operations not only affect the skyline computation, but also influence the skyline results. Furthermore, the incompleteness of data raises the issue of losing transitivity property of skyline technique, which leads to the problem of cyclic dominance. It is undesirable to process skyline queries on a dynamic and incomplete database by directly applying skyline process over the entire updated database due to the prohibitive cost. Thus, this paper proposes a framework called D-SKY for processing skyline queries in a dynamic and incomplete database. D-SKY aims at avoiding scanning the whole database after an update operation is performed to identify the new skylines. In this paper, we consider the case of an insert operation, in which database is updated by adding new data items. D-SKY framework exploits the existing skylines to identify the newly added dominated data items before applying the skyline process. Therefore, a large amount of dominated data items is pruned, which reduces the number of domination tests to be conducted and helps in avoiding scanning the whole data after an update operation is performed. Experiment result conducted on both real and synthetic datasets demonstrates that our solution outperforms the existing solutions in terms of reducing the number of pairwise comparisons and processing time.

With the popularity of wireless devices and the increasing demand of network applications, it is emergent to develop more effective communications paradigm to enable new and powerful pervasive applications, and to allow services to be accessed anywhere, at anytime. However, it is extremely challenging to construct efficient and reliable networks to connect wireless devices due to the increasing communications need and the dynamic nature of wireless communications. In order to improve transmission throughput, many efforts have been made in recent years to reduce traffic and hence transmission collisions by constructing backbone networks with the minimum size. However, many other important issues need to be considered. Instead of simply minimizing the number of backbone nodes or supporting some isolated network features, in this work, we exploit the use of algebraic connectivity to control backbone network topology design for concurrent improvement of backbone network robustness, capacity, stability and routing efficiency. In order to capture other network features, we provide a general cost function and introduce a new metric, connectivity efficiency, to trade off algebraic connectivity and cost for backbone construction. We formally prove the problem of formulating a backbone network with the maximum connectivity efficiency that is NP-hard, and design both centralized and distributed algorithms to build more robust and efficient backbone infrastructure to better support the application needs. We have made extensive simulations to evaluate the performance of our work. Compared to literature studies on constructing wireless backbone networks, the incorporation of algebraic connectivity into the network performance metric could achieve much higher throughput and delivery ratio, and much lower end-to-end delay and routing distances under all test scenarios. We hope our work could stimulate more future research in designing more reliable and efficient networks. Our performance studies demonstrate that, compared to peer work, the incorporation of algebraic connectivity into network performance metric could achieve much higher throughput and delivery ratio, and much lower end-to-end delay and routing distances under all test scenarios. We hope our work could stimulate more future research in designing more reliable and efficient networks.

In recent years, cognitive radio technology (CR) has been proposed to allow unlicensed secondary users (SUs) to opportunistically access the channels unused by primary users. As a result, there is a lot of recent interests on studying cognitive radio cellular networks (CogCells) that can support both PUs and SUs. Due to the limited transmission range of SUs, in this work we consider supporting Multi-hop infrastructure-based secondary systems (SSs), where SUs can communicate with the BS over multiple hops. The use of SSs improves the reliability and coverage compared to its single-hop counterpart. In addition, SUs are allowed to access multiple channels, which helps to increase transmission reliability and coverage and relieve interference at PUs. To enable multi-hop secondary transmissions, it is also important to support efficient routing. In CogCells, efficient admission control, channel assignment and routing is crucial for the coverage optimization of SSs and to ensure the QoS requirements in CogCells. In this paper, we mathematically formulate the problem of joint admission control, channel assignment and QoS routing to maximize the coverage of SUs in a CogCell system that supports multi-hop secondary transmissions, taking into account the interference constraints and QoS requirements from the PUs and admitted SUs. To our best knowledge, this is the first study that attempts to optimize the coverage of SUs in multi-hop CogCells with the concurrent support of the above three important procedures. We show that the problem is NP-hard and propose three different algorithms to solve the coverage optimization problem and give the theoretical analyses of its performances in terms of approximation ratio to the optimum. Our solutions include a greedy heuristic approximation scheme, an algorithm that can provide exact solution, and a new approximation solution with a poly-logarithmic approximation ratio guarantee, e.g., the performance of our algorithm is within a poly-logarithmic factor of that of any optimal algorithm for the problem. Our preliminary simulation results indicate that our new approximation algorithms can effectively exploit the increased number of SUs and channels, and performs much better than the theoretical worst case bound.

Abstract Wireless mesh networks (WMNs) is an emerging communication paradigm to enable resilient, cost‐efficient and reliable services for the future‐generation wireless networks. In this paper, we study the broadcasting (one‐to‐all communication) in WMNs with known topology, i.e. where for each primitive the schedule of transmissions is pre‐computed based on full knowledge about the size and the topology of the network. We show that broadcasting can complete in D + O(logn) time units in the WMN with sizenand diameterD. Moreover, we also propose an optimal O(D)‐time deterministic energy efficient broadcasting scheduling, under which each node in the WMN is only allowed to transmit at most once. Furthermore, we explore the fault‐tolerant broadcasting in the WMN. We show an O(n)‐time deterministic broadcasting schedule with large number of link failures. This is an optimal schedule in the sense that there exists a network topology in which the broadcasting cannot complete in less than Ω(n) units of time. Copyright © 2009 John Wiley & Sons, Ltd.

Wireless Mesh Networking (WMN) is an emerging communication paradigm to enable resilient, cost-efficient and reliable services for the future-generation wireless networks. We study the minimum-latency communication primitive of gossiping (all-to-all communication) in known topology WMNs under physical interference model, i.e., where the schedule of transmissions is pre-computed in advance based on full knowledge about the size and the topology of the Wireless Mesh Network (WMN). Each mesh node in the WMN is initially given a message and the objective is to design a minimum-latency schedule such that each mesh node distributes its message to all other mesh nodes. The problem of computing a minimum-latency gossiping schedule for a given WMN is NP-hard, hence it is only possible to get a polynomial approximation algorithm. In this paper, we show a deterministic O(log n)-approximation algorithm in which the proposed scheme can complete gossiping task in time at most O(log n) factor far from the optimum (e.g., the minimum-latency schedule) and it can be computed in polynomial time in terms of the size of the WMN. From our best knowledge, it is the first time to investigate gossiping problem in WMNs under physical interference model.

Wireless Mesh Network (WMN) is an emerging communication paradigm to enable resilient, cost-efficient and reliable services for the future-generation wireless networks. In this paper, we study the problem of multipoint-to-multipoint (M2M) multicasting in a WMN which aims to use the minimum number of time slots (minimum latency) to exchange messages among a group of k mesh nodes in a WMN with n mesh nodes under physical interference model. We study the M2M multicasting problem in a distributed environment where each participant only knows that there are k participants and it does not know who are other k - 1 participants among n mesh nodes. It is well known that the computation of an optimal M2M multicasting schedule is NP-hard. We present a fully distributed deterministic algorithm for such an M2M multicasting problem and analyze its time complexity. We show that if the maximum hop distance between any two out of the k participants is d, then the studied M2M multicasting problem can be solved in time O(d log n + k) with a polynomial-time computation in unit disk graphs, which is an almost optimal scheme in the sense that there exists a WMN topology, e.g., a line, in which the M2M multicasting cannot be completed in less than Ω(d + k) units of time. From our best knowledge, it is the first time to investigate M2M multicasting problem in WMNs under physical interference model.

Wireless mesh networks (WMNs) is an emerging communication paradigm to enable resilient, cost-efficient and reliable services for the future-generation wireless networks. We study the minimum-latency communication primitive of gossiping (all-to-all communication) in multi-hop ad-hoc WMNs. Each mesh node in the WMN is initially given a message and the objective is to design a minimum-latency schedule such that each mesh node distributes its message to all other mesh nodes. Minimum-latency gossiping problem is known to be NP-hard even for the scenario in which the topology of the WMN is known to all mesh nodes in advance. We show an approximation scheme that can complete gossiping task in O(n log3/2n) time units with high probability at least 1 - 1/n in any ad-hoc WMN of size n. Our algorithm allows the labels (identifiers) of the mesh nodes to be polynomially large in n. To the best of our knowledge, this is the first time that randomized algorithm has been considered in ad-hoc WMNs with large labels. Moreover, our gossiping scheme also significantly improved all current gossiping algorithms in terms of approximation ratio. Our work has approximation ratio at most O(log3/2n) which is a great improvement of the current best known state-of-the-art algorithm with approximation ratio O(log2n) but for linearly large node labels by Czumaj and Rytter [FOCS'03].

Wireless Mesh Network (WMN) is an emerging communication paradigm to enable resilient, cost-efficient and reliable services for the future-generation wireless networks. In this paper, we study the problem of multipoint-to-multipoint (M2M) multicasting in a WMN which aims to use the minimum number of time slots to exchange messages among a group of k mesh nodes in a multi-hop WMN with n mesh nodes. We study the M2M multicasting problem in a distributed environment where each participant only knows that there are k participants and it does not know who are other k -1 participants among n mesh nodes. It is known that the computation of an optimal M2M multicasting schedule is NP-hard. We present a fully distributed deterministic algorithm for such an M2M multicasting problem and analyze its time complexity. We show that if the maximum hop distance between any two out of the k participants is d, then the studied M2M multicasting problem can be solved in time O(d log2n+k log3n/log k) with a polynomial-time computation, which is an almost optimal scheme due to the lower bound Omega(d+ k log n/log k) given in [5]. Our algorithm also improves the currently best known result with running time O(d log2n + k log4n) in [13]. In this paper, we also propose a distributed deterministic algorithm which accomplishes the M2M multicasting in time O(d+k) with a polynomial-time computation in unit disk graphs. This is an asymptotically optimal algorithm in the sense that there exists a WMN topology, e.g., a line, a ring, a star or a complete graph, in which the M2M multicasting cannot be completed in less than Omega(d+k) units of time.

In cognitive radio cellular networks (CogCells), primary users (PUs) rarely utilize all the assigned frequency bands at a certain time and a location. The spectral inefficiency caused by the spectrum holes motivated cognitive radio technology (CR) that presents unlicensed secondary users (SUs) an opportunity for using spectrum holes. CR makes the SUs to find and use the spectrum holes without interrupting the operation of PUs. The SUs are allowed to access the channel licensed to the PUs which consist of primary transmitters (PTs) and primary receivers (PRs) when the interference to the PRs is less than acceptable value (i.e., predefined system threshold), and the quality of service (QoS) required by PTs are also guaranteed. According to different levels of QoS required by SUs, the network operator can achieve different secondary revenues by providing different QoS levels to SUs. Due to the high density, the mobility of SUs, the interference limitation at PRs and the QoS requirements from PTs, not all SUs can be supported. The problem we investigated in this paper is to select the maximum subset of SUs to maximize the total secondary revenue of the CogCell, meanwhile the QoS requirements from both PTs and admitted SUs must be guaranteed. Moreover, the interference caused by the admitted SUs and the PTs at the PRs (due to access the same channel) has to be less than the predefined system threshold. In this paper, we formulate such a joint QoS-aware admission control, channel assignment, and power allocation scheme as a non-linear NP-hard optimization problem. This is a very challenging problem and the NP-hardness has been shown in the literature even for the single-channel scenario. In this paper, we propose a new polynomial-time joint QoS-aware admission control, channel assignment and power allocation scheme which has a O(1/log npr+ log nw) approximation guarantee, e.g., the total secondary revenue achieved by our algorithm is at least OmegaO(1/log npr+ log nw)of the optimum, where nrpis the number of PRs and nwis the number of available channels in the CogCell. Note that Our algorithm also significantly improves the current best known solution with a O(1/npr) approximation guarantee for the single-channel scenario [11]. In this paper, we also propose a greedy heuristic approximation algorithm and an exact solution. The simulation results show that the approximation algorithms we proposed can achieve significantly higher secondary revenue than the currently best known approximation approach for this problem, an extension of the minimal SINR removal algorithm in [15]. Indeed, quite surprisingly, the simulation results also demonstrate that the secondary revenue achieved by our approximation approaches is very close to the optimum in practice, specially for the approximation O(1/log npr+ log nw) algorithm.

In this paper, we study the problem of gateway selection for throughput optimization in multi-radio multi-channel wireless mesh networks. In contrast to the various methodologies in the literature, we explicitly model the delay overhead that is incurred during channel switching, and consider this delay issue in the design of our mechanisms. From our best knowledge, it is the first time to take account switching overhead into the scenario of gateway placement in multi-radio multi-channel wireless networks not only in wireless mesh networks. Given the number of gateways to be deployed and the interference model employed for the network system, we study how to select the mesh nodes to be equipped with gateway functionality in the network such that the total network throughput is maximized while it also ensures a certain fairness among all mesh nodes. We propose a new gateway deployment method using a crosslayer throughput optimization, and prove that the performance of our scheme in term of the achieved throughput is only a constant factor far to the one of optimum. Simulation results demonstrate that our mechanism can effectively exploit the available resources and achieve much better performance on network throughput than random, fixed deployment and grid-based methods in the literature.

This paper concerns the communication primitives of broadcasting (one-to-all communication) and gossiping (all-to-all communication) in radio networks with known topology, i.e., where for each primitive the schedule of transmissions is precomputed based on full knowledge about the size and the topology of the network.The first part of the paper examines the two communication primitives in general graphs. In particular, it proposes a new (efficiently computable) deterministic schedule that uses O(D+Δ log n) time units to complete the gossiping task in any radio network with size n, diameter D and max-degree Δ. Our new schedule improves and simplifies the currently best known gossiping schedule, requiring time O(D+√[i+2]DΔ logi+1 n), for any network with the diameter D=Ω(logi+4n), where i is an arbitrary integer constant i ≥ 0, see [17]. For the broadcast task we deliver two new results: a deterministic efficient algorithm for computing a radio schedule of length D+O(log3 n), and a randomized algorithm for computing a radio schedule of length D+O(log2 n). These results improve on the best currently known D+O(log4 n) time schedule due to Elkin and Kortsarz [12].The second part of the paper focuses on radio communication in planar graphs, devising a new broadcasting schedule using fewer than 3D time slots. This result improves, for small values of D, on currently best known D+O(log3n) time schedule proposed by Elkin and Kortsarz in [12]. Our new algorithm should be also seen as the separation result between the planar and the general graphs with a small diameter due to the polylogarithmic inapproximability result in general graphs due to Elkin and Kortsarz, see [11].