Search

Abstract In recent years, research on opinion mining from X (formerly Twitter) has rapidly advanced, focusing on processing tweets to determine user sentiments about events. Many researchers prefer using machine and deep learning techniques for this analysis. This work proposes a novel approach integrating the C4.5 procedure, fuzzy rule patterns, and convolutional neural networks. The approach involves six steps: pre-processing to remove noisy data, vectorizing tweets with word embedding, extracting sentiment and contextual features using convolutional neural networks, fuzzifying outputs with a Gaussian fuzzifier to handle ambiguity, constructing a fuzzy tree and rule base using a fuzzy version of C4.5, and classifying tweets with fuzzy General Reasoning. This method combines the benefits of convolutional neural networks and C4.5 while addressing imprecise data with fuzzy logic. Implemented on a Hadoop framework-based cluster with five computing units, the approach was extensively tested. The results showed that the model performs exceptionally well on the COVID-19_Sentiments dataset, surpassing other classification algorithms with a precision rate of 94.56%, false-negative rate of 5.28%, classification rate of 95.15%, F1-score of 94.63%, kappa statistic of 95.12%, execution time of 11.81 s, false-positive rate of 4.26%, error rate of 4.26%, specificity of 95.74%, recall of 94.72%, stability with a mean deviation standard of 0.09%, convergence starting around the 75th round, and significantly reduced complexity in terms of time and space.

Abstract The generative Artificial Intelligence (AI) tools based on Large Language Models (LLMs) use billions of parameters to extensively analyse large datasets and extract critical information such as context, specific details, identifying information, use this information in the training process, and generate responses for the requested queries. The extracted data also contain sensitive information, seriously threatening user privacy and reluctance to use such tools. This article proposes the conceptual model called PrivChatGPT, a privacy‐preserving model for LLMs consisting of two main components, that is, preserving user privacy during the data curation/pre‐processing and preserving private context and the private training process for large‐scale data. To demonstrate the applicability of PrivChatGPT, it is shown how a private mechanism could be integrated into the existing model for training LLMs to protect user privacy; specifically, differential privacy and private training using Reinforcement Learning (RL) were employed. The privacy level probabilities are associated with the document contents, including the private contextual information, and with metadata, which is used to evaluate the disclosure probability loss for an individual's private information. The privacy loss is measured and the measure of uncertainty or randomness is evaluated using entropy once differential privacy is applied. It recursively evaluates the level of privacy guarantees and the uncertainty of public databases and resources during each update when new information is added for training purposes. To critically evaluate the use of differential privacy for private LLMs, other mechanisms were hypothetically compared such as Blockchain, private information retrieval, randomisation, obfuscation, anonymisation, and the use of Tor for various performance measures such as the model performance and accuracy, computational complexity, privacy vs. utility, training latency, vulnerability to attacks, and resource consumption. It is concluded that differential privacy, randomisation, and obfuscation can impact the training models' utility and performance; conversely, using Tor, Blockchain, and Private Information Retrieval (PIR) may introduce additional computational complexity and high training latency. It is believed that the proposed model could be used as a benchmark for privacy‐preserving LLMs for generative AI tools.

Blockchain technology enables users to control and record their cryptocurrency transactions through the use of digital wallets. As the use of blockchain technology and cryptocurrency wallets continues to grow in popularity, the potential for attacks on these wallets increases, as attackers seek to gain access to the large sums of cryptocurrency they contain. To mitigate these risks, it is important to conduct thorough security evaluations of wallets and implement strong protective measures. In recent years, there have been several incidents involving significant losses of cryptocurrency in crypto-wallets, and in this research, a comprehensive evaluation of seed phrase and password attack methods found in the published literature was conducted, and the topic was advanced by addressing the question of whether seed phrases are hackable. The research aims to use the elliptic-curve cryptography (ECC) encryption algorithm for storing the seed phrase online by encrypting the seed phrase and using the splitting technique to store the crypto wallet seed phrase. It was concluded that it is only possible to hack a seed phrase if a significant portion of it is already known, but even this would require a significant amount of time and computational power.

In the modern landscape of information and communication technologies, the current healthcare industry confronts significant challenges. These include a shortage of experienced medical professionals, disparities in access to healthcare services that persist across different regions around the globe, and an increased need for detailed, real-time monitoring of patients in both urban and remote regions. This article delves into the potential of the Internet of Audio Things for Healthcare (IoAuT4H), which lies at the intersection of Internet of Audio Things and the Internet of Medical Things, as a solution to these pressing issues. By seamlessly merging cutting-edge audio technology, networking, and the advanced deep learning techniques, the IoAuT4H emerges as a promising solution. It has the potential to reshape hospital and clinical infrastructure, streamline early medical interventions, and facilitate rapid emergency responses. Additionally, this study underscores the pivotal role of the IoAuT4H in strengthening overall health practices, refining care methods for the elderly, and rejuvenating paediatric health approaches. While the benefits of the IoAuT4H are numerous, this article also critically examines the challenges in its widespread adoption. These include ethical considerations, ensuring the accuracy of audio data, and integrating it effectively with existing healthcare systems. In conclusion, this article seeks to provide a comprehensive understanding of the IoAuT4H, positioning it as a bridge between current healthcare challenges and technological advancements.

The metaverse is currently undergoing a profound transformation, fundamentally reshaping our perception of reality. It has transcended its origins to become an expansion of human consciousness, seamlessly blending the physical and virtual worlds. Amidst this transformative evolution, numerous applications are striving to mould the metaverse into a digital counterpart capable of delivering immersive human-like experiences. These applications envisage a future where users effortlessly traverse between physical and digital dimensions. Taking a step forward, affective computing technologies can be utilised to identify users' emotional cues and convey authentic emotions, enhancing genuine, meaningful, and context-aware interactions in the digital world. In this paper, we explore how integrating emotional intelligence can enhance the traditional metaverse, birthing an emotionally intelligent metaverse (EIM). Our work illuminates the multifaceted potential of EIM across diverse sectors, including healthcare, education, gaming, automotive, customer service, human resources, marketing, and urban metaverse cyberspace. Through our examination of these sectors, we uncover how infusing emotional intelligence enriches user interactions and experiences within the metaverse. Nonetheless, this transformative journey is riddled with challenges, and we address the obstacles hindering the realisation of EIM's full potential. By doing so, we lay the groundwork for future research endeavours aimed at further enhancing and refining the captivating journey into the world of EIM.

Despite the recent progress in emotion recognition, state-of-the-art systems are unable to achieve improved performance in cross-language settings. In this article we propose a Multimodal Dual Attention Transformer (MDAT) model to improve cross-language multimodal emotion recognition. Our model utilises pre-trained models for multimodal feature extraction and is equipped with dual attention mechanisms including graph attention and co-attention to capture complex dependencies across different modalities and languages to achieve improved cross-language multimodal emotion recognition. In addition, our model also exploits a transformer encoder layer for high-level feature representation to improve emotion classification accuracy. This novel construct preserves modality-specific emotional information while enhancing cross-modality and cross-language feature generalisation, resulting in improved performance with minimal target language data. We assess our model's performance on four publicly available emotion recognition datasets and establish its superior effectiveness compared to recent approaches and baseline models.

Abstract Urban activities, particularly vehicle traffic, are contributing significantly to environmental pollution with detrimental effects on public health. The ability to anticipate air quality in advance is critical for public authorities and the general public to plan and manage these activities, which ultimately help in minimizing the adverse impact on the environment and public health effectively. Thanks to recent advancements in Artificial Intelligence and sensor technology, forecasting air quality is possible through the consideration of various environmental factors. This paper presents our novel solution for air quality prediction and its correlation with different environmental factors and urban activities, such as traffic density. To this aim, we propose a multi-modal framework by integrating real-time data from different environmental sensors and traffic density extracted from Closed Circuit Television footage. The framework effectively addresses data inconsistencies arising from sensor and camera malfunctions within a streaming dataset. The dataset exhibits real-world complexities, including abrupt camera or station activations/deactivations, noise interference, and outliers. The proposed system tackles the challenge of predicting air quality at locations having no sensors or experiencing sensor failures by training a joint model on the data obtained from nearby stations/sensors using a Particle Swarm Optimization (PSO)-based merit fusion of the sensor data. The proposed methodology is evaluated using various variants of the LSTM model including Bi-directional LSTM, CNN-LSTM, and Convolutions LSTM (ConvLSTM) obtaining an improvement of 48%, 67%, and 173% for short-term, medium-term, and long-term periods, respectively, over the ARIMA model.

Over the past six decades, the computing systems field has experienced significant transformations, profoundly impacting society with transformational developments, such as the Internet and the commodification of computing. Underpinned by technological advancements, computer systems, far from being static, have been continuously evolving and adapting to cover multifaceted societal niches. This has led to new paradigms such as cloud, fog, edge computing, and the Internet of Things (IoT), which offer fresh economic and creative opportunities. Nevertheless, this rapid change poses complex research challenges, especially in maximizing potential and enhancing functionality. As such, to maintain an economical level of performance that meets ever-tighter requirements, one must understand the drivers of new model emergence and expansion, and how contemporary challenges differ from past ones. To that end, this article investigates and assesses the factors influencing the evolution of computing systems, covering established systems and architectures as well as newer developments, such as serverless computing, quantum computing, and on-device AI on edge devices. Trends emerge when one traces technological trajectory, which includes the rapid obsolescence of frameworks due to business and technical constraints, a move towards specialized systems and models, and varying approaches to centralized and decentralized control. This comprehensive review of modern computing systems looks ahead to the future of research in the field, highlighting key challenges and emerging trends, and underscoring their importance in cost-effectively driving technological progress.

Interviewing skills play a pivotal role in the job application and search, professional development to prepare for interviewing is a neglected area of research. Professional training methods are available but are often prohibitively expensive, limiting opportunities primarily to privileged individuals. To bridge this accessibility gap and democratize access to job opportunities, there is a need to develop automated interview simulation platforms. The advent of Generative AI (GenAI) technology, in particular Large Language Models (LLMs), makes this is viable proposition but progress is hindered by the absence of open-source implementations for reproducibility and comparison, as well as the lack of suitable evaluation benchmarks and experimental setups. In particular, we do not yet know how robust such systems are and if they will be bias-free, factors that will contribute to their acceptability and use. To this end, we propose Interview Training and Education Module (ITEM), a job interview training module that combines Virtual Reality-based metaverse technology with LLM-based GenAI models. Our module creates realistic interview experiences for skill enhancement, complete with personalized feedback and improvement guidelines based on user responses. In this paper, we present an experimental evaluation of the module to ascertain its robustness, including a bias analysis. Firstly, we establish an experimental setup to gauge platform robustness by examining question similarity across varied prompts using Bidirectional and Auto-Regressive Transformers (BART) and topic modeling. Subsequently, we explore biases in three categories—country of origin, religion, and gender—by analyzing ITEM's evaluation scores while manipulating candidate backgrounds, all while keeping their responses unchanged. Our findings indicate potential biases replicated by ITEM, highlighting the need for caution in its application for personal development and training. This pioneering initiative introduces the first open-source module for job interview training within a virtual metaverse, leveraging LLM-based Generative AI, designed for extension and testing by the scientific community, thereby enhancing insights into the limitations and ethical considerations of AI-driven interview simulation platforms.

Abstract Automakers from Honda to Lamborghini are incorporating voice interaction technology into their vehicles to improve the user experience and offer value-added services. Speech recognition systems are a key component of smart cars, enhancing convenience and safety for drivers and passengers. In the future, safety-critical features may rely on speech recognition, but this raises concerns about children accessing such services. To address this issue, the LimitAccess system is proposed, which uses TinyML for age classification and helps parents limit children’s access to critical speech recognition services. This study employs a lite convolutional neural network (CNN) model for two different reasons: First, CNN showed superior accuracy compared to other audio classification models for age classification problems. Second, the lite model will be integrated into a microcontroller to meet its limited resource requirements. To train and evaluate our model, we created a dataset that included child and adult voices of the keyword “open”. The system approach categorizes voices into age groups (child, adult) and then utilizes that categorization to grant access to a car. The robustness of the model was enhanced by adding a new class (recordings) to the dataset, which enabled our system to detect replay and synthetic voice attacks. If an adult voice is detected, access to start the car will be granted. However, if a child’s voice or a recording is detected, the system will display a warning message that educates the child about the dangers and consequences of the improper use of a car. Arduino Nano 33 BLE sensing was our embedded device of choice for integrating our trained, optimized model. Our system achieved an overall F1 score of 87.7% and 85.89% accuracy. LimitAccess detected replay and synthetic voice attacks with an 88% F1 score.

Abstract Early diagnosis of dental caries progression can prevent invasive treatment and enable preventive treatment. In this regard, dental radiography is a widely used tool to capture dental visuals that are used for the detection and diagnosis of caries. Different deep learning (DL) techniques have been used to automatically analyse dental images for caries detection. However, most of these techniques require large-scale annotated data to train DL models. On the other hand, in clinical settings, such medical images are scarcely available and annotations are costly and time-consuming. To this end, we present an efficient self-training-based method for caries detection and segmentation that leverages a small set of labelled images for training the teacher model and a large collection of unlabelled images for training the student model. We also propose to use centroid cropped images of the caries region and different augmentation techniques for the training of self-supervised models that provide computational and performance gains as compared to fully supervised learning and standard self-supervised learning methods. We present a fully labelled dental radiographic dataset of 141 images that are used for the evaluation of baseline and proposed models. Our proposed self-supervised learning strategy has provided performance improvement of approximately 6% and 3% in terms of average pixel accuracy and mean intersection over union, respectively as compared to standard self-supervised learning. Data and code will be made available to facilitate future research.

Owing to the geographically scattered end devices (EDs) in long-range wide area networks (LoRaWAN), devices that combat challenging cyber threats and attacks are of critical significance. In this perspective, LoRa Alliance® is continuously evolving the security of LoRaWAN and recently introduced a new version i.e., LoRaWAN 1.1x that is featured with security improvement. However, the wireless nature of LoRaWAN implementation still leaves it vulnerable to security breaches that compromise its integrity. Several problems have been pinpointed in the newer version such as one issue with key distribution in LoRaWAN 1.1 is that the keys are often pre-installed on the devices at the time of manufacturing. It can introduce security risks if the keys are not adequately protected or if the devices are compromised before they are deployed. In other words, the pre-installed keys may not be updated regularly, which can also introduce security risks. Thus, the keys need to be handled securely to maintain the security of the network and the over-the-air firmware updates feature could introduce new security challenges for the key distribution. This paper presents a key generation and distribution (KGD) mechanism that securely exchanges the root key between the ED and the application server (AS). The KGD protocol provides authentication by integrating Advanced Encryption Standard (AES-128) in addition to a secure hash function known as Argon2. The proposed protocol utilizes Elliptic-Curve Diffie-Hellman (ECDH) key exchange method that makes the protocol resilient to cyber threats. The ECDH algorithm exchanges the keys on the insecure channels and is, therefore, vulnerable to Man-in-the-Middle (MITM) attacks in the network. Therefore, to validate the key agreement and avoid adversaries, the KGD protocol considers the Elliptic Curve Digital Signature Algorithm (ECDSA) that authenticates and allows legitimate instances in the network. In last, a formal security analysis using the Scyther tool validates the security enhancement of the KGD protocol.

Abstract The twenty-first century technological advances driven by exponential rise of artificial intelligence (AI) technology have ushered in a new era that offers many of us hitherto unimagined luxuries and facilities. However, under the guise of this progressive discourse, particularly in the backdrop of current neo-liberal late-capitalist postmodern world, AI development also has prompted an increasingly uncertain ethical tomorrow. This paper aims to probe the question of ethics by exploring the true ramifications of AI and interrogating its various ethical dimensions. It questions the essential goodness that is attributed to unstinted AI development before elucidating the ethical repercussions of AI advancements and the aptness of the current market logics and business models that govern the tech-industry. The paper next positions a holistic Islamic virtue-based AI ethics framework grounded in the context of Islamic objectives (maqāṣid) as an alternative ethical system for AI governance. We argue that this distinctive Islamic virtue-based ethical approach, which can be used to explore AI-related ethical problems more holistically due to its ontological base and rich tradition while keeping in check undue influence from the current socio-politico-economic climate, can be a valuable addition to the global discourse on AI ethics.

Despite significant improvements over the last few years, cloud-based healthcare applications continue to suffer from poor adoption due to their limitations in meeting stringent security, privacy, and quality of service requirements (such as low latency). The edge computing trend, along with techniques for distributed machine learning such as federated learning, has gained popularity as a viable solution in such settings. In this paper, we leverage the capabilities of edge computing in medicine by evaluating the potential of intelligent processing of clinical data at the edge. We utilized the emerging concept of clustered federated learning (CFL) for an automatic COVID-19 diagnosis. We evaluate the performance of the proposed framework under different experimental setups on two benchmark datasets. Promising results are obtained on both datasets resulting in comparable results against the central baseline where the specialized models (i.e., each on a specific image modality) are trained with central data, and improvements of 16% and 11% in overall F1-Scores have been achieved over the trained model trained (using multi-modal COVID-19 data) in the CFL setup on X-ray and Ultrasound datasets, respectively. We also discussed the associated challenges, technologies, and techniques available for deploying ML at the edge in such privacy and delay-sensitive applications.

At present, with the growing number of Web 2.0 platforms such as Instagram, Facebook, and Twitter, users honestly communicate their opinions and ideas about events, services, and products. Owing to this rise in the number of social platforms and their extensive use by people, enormous amounts of data are produced hourly. However, sentiment analysis or opinion mining is considered as a useful tool that aims to extract the emotion and attitude from the user-posted data on social media platforms by using different computational methods to linguistic terms and various Natural Language Processing (NLP). Therefore, enhancing text sentiment classification accuracy has become feasible, and an interesting research area for many community researchers. In this study, a new Fuzzy Deep Learning Classifier (FDLC) is suggested for improving the performance of data-sentiment classification. Our proposed FDLC integrates Convolutional Neural Network (CNN) to build an effective automatic process for extracting the features from collected unstructured data and Feedforward Neural Network (FFNN) to compute both positive and negative sentimental scores. Then, we used the Mamdani Fuzzy System (MFS) as a fuzzy classifier to classify the outcomes of the two used deep (CNN+FFNN) learning models in three classes, which are: Neutral, Negative, and Positive. Also, to prevent the long execution time taking by our hybrid proposed FDLC, we have implemented our proposal under the Hadoop cluster. An experimental comparative study between our FDLC and some other suggestions from the literature is performed to demonstrate our offered classifier's effectiveness. The empirical result proved that our FDLC performs better than other classifiers in terms of true positive rate, true negative rate, false positive rate, false negative rate, error rate, precision, classification rate, kappa statistic, F1-score and time consumption, complexity, convergence, and stability.

Opinion Mining (OM) is a field of Natural Language Processing (NLP) that aims to capture human sentiment in the given text. With the ever-spreading of online purchasing websites, micro-blogging sites, and social media platforms, OM in online social media platforms has picked the interest of thousands of scientific researchers. Because the reviews, tweets and blogs acquired from these social media networks, act as a significant source for enhancing the decision making process. The obtained textual data (reviews, tweets, or blogs) are classified into three different class labels which are negative, neutral and positive for analyzing and extracting relevant information from the given dataset. In this contribution, we introduce an innovative MapReduce improved weighted ID3 decision tree classification approach for OM, which consists mainly of three aspects: Firstly We have used several feature extractors to efficiently detect and capture the relevant data from the given tweets, including N-grams or character-level, Bag-Of-Words, word embedding (GloVe, Word2Vec), FastText, and TF-IDF. Secondly, we have applied a multiple feature selector to reduce the high feature’s dimensionality, including Chi-square, Gain Ratio, Information Gain, and Gini Index. Finally, we have employed the obtained features to carry out the classification task using an improved ID3 decision tree classifier, which aims to calculate the weighted information gain instead of information gain used in traditional ID3. In other words, to measure the weighted information gain for the current conditioned feature, we follow two steps: First, we compute the weighted correlation function of the current conditioned feature. Second, we multiply the obtained weighted correlation function by the information gain of this current conditioned feature. This work is implemented in a distributed environment using the Hadoop framework, with its programming framework MapReduce and its distributed file system HDFS. Its primary goal is to enhance the performance of a well-known ID3 classifier in terms of accuracy, execution time, and ability to handle the massive datasets. We have carried out several experiences that aims to assess the effectiveness of our suggested classifier compared to some other contributions chosen from the literature. The experimental results demonstrated that our ID3 classifier works better on COVID-19_Sentiments dataset than other classifiers in terms of Recall (85.72 %), specificity (86.51 %), error rate (11.18 %), false-positive rate (13.49 %), execution time (15.95s), kappa statistic (87.69 %), F1-score (85.54 %), classification rate (88.82 %), false-negative rate (14.28 %), precision rate (86.67 %), convergence (it convergent towards the iteration 90), stability (it is more stable with mean deviation standard equal to 0.12 %), and complexity (it requires much lower time and space computational complexity).

With the hyperconnectivity and ubiquity of the Internet, the fake news problem now presents a greater threat than ever before. One promising solution for countering this threat is to leverage deep learning (DL)-based text classification methods for fake-news detection. However, since such methods have been shown to be vulnerable to adversarial attacks, the integrity and security of DL-based fake news classifiers are under question. Although many works study text classification under the adversarial threat, to the best of our knowledge, we do not find any work in literature that specifically analyzes the performance of DL-based fake-news detectors under adversarial settings. We bridge this gap by evaluating the performance of fake-news detectors under various configurations under black-box settings. In particular, we investigate the robustness of four different DL architectural choices—multilayer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN) and a recently proposed Hybrid CNN-RNN trained on three different state-of-the-art datasets—under different adversarial attacks (Text Bugger, Text Fooler, PWWS, and Deep Word Bug) implemented using the state-of-the-art NLP attack library, Text-Attack. Additionally, we explore how changing the detector complexity, the input sequence length, and the training loss affect the robustness of the learned model. Our experiments suggest that RNNs are robust as compared to other architectures. Further, we show that increasing the input sequence length generally increases the detector’s robustness. Our evaluations provide key insights to robustify fake-news detectors against adversarial attacks.

BackgroundContact tracing has been globally adopted in the fight to control the infection rate of COVID-19. To this aim, several mobile apps have been developed. However, there are ever-growing concerns over the working mechanism and performance of these applications. The literature already provides some interesting exploratory studies on the community’s response to the applications by analyzing information from different sources, such as news and users’ reviews of the applications. However, to the best of our knowledge, there is no existing solution that automatically analyzes users’ reviews and extracts the evoked sentiments. We believe such solutions combined with a user-friendly interface can be used as a rapid surveillance tool to monitor how effective an application is and to make immediate changes without going through an intense participatory design method. ObjectiveIn this paper, we aim to analyze the efficacy of AI and NLP techniques for automatically extracting and classifying the polarity of users’ sentiments by proposing a sentiment analysis framework to automatically analyze users’ reviews on COVID-19 contact tracing mobile apps. We also aim to provide a large-scale annotated benchmark data set to facilitate future research in the domain. As a proof of concept, we also developed a web application based on the proposed solutions, which is expected to help the community quickly analyze the potential of an application in the domain. MethodsWe propose a pipeline starting from manual annotation via a crowd-sourcing study and concluding with the development and training of artificial intelligence (AI) models for automatic sentiment analysis of users’ reviews. In detail, we collected and annotated a large-scale data set of user reviews on COVID-19 contact tracing applications. We used both classical and deep learning methods for classification experiments. ResultsWe used 8 different methods on 3 different tasks, achieving up to an average F1 score of 94.8%, indicating the feasibility of the proposed solution. The crowd-sourcing activity resulted in a large-scale benchmark data set composed of 34,534 manually annotated reviews. ConclusionsThe existing literature mostly relies on the manual or exploratory analysis of users’ reviews on applications, which is tedious and time-consuming. In existing studies, generally, data from fewer applications are analyzed. In this work, we showed that AI and natural language processing techniques provide good results for analyzing and classifying users’ sentiments’ polarity and that automatic sentiment analysis can help to analyze users’ responses more accurately and quickly. We also provided a large-scale benchmark data set. We believe the presented analysis, data set, and proposed solutions combined with a user-friendly interface can be used as a rapid surveillance tool to analyze and monitor mobile apps deployed in emergency situations leading to rapid changes in the applications without going through an intense participatory design method.

Brain tumors are among the deadliest forms of cancer, characterized by abnormal proliferation of brain cells. While early identification of brain tumors can greatly aid in their therapy, the process of manual segmentation performed by expert doctors, which is often time-consuming, tedious, and prone to human error, can act as a bottleneck in the diagnostic process. This motivates the development of automated algorithms for brain tumor segmentation. However, accurately segmenting the enhanced and core tumor regions is complicated due to high levels of inter- and intra-tumor heterogeneity in terms of texture, morphology, and shape. This study proposes a fully automatic method called the selective deeply supervised multi-scale attention network (SDS-MSA-Net) for segmenting brain tumor regions using a multi-scale attention network with novel selective deep supervision (SDS) mechanisms for training. The method utilizes a 3D input composed of five consecutive slices, in addition to a 2D slice, to maintain sequential information. The proposed multi-scale architecture includes two encoding units to extract meaningful global and local features from the 3D and 2D inputs, respectively. These coarse features are then passed through attention units to filter out redundant information by assigning lower weights. The refined features are fed into a decoder block, which upscales the features at various levels while learning patterns relevant to all tumor regions. The SDS block is introduced to immediately upscale features from intermediate layers of the decoder, with the aim of producing segmentations of the whole, enhanced, and core tumor regions. The proposed framework was evaluated on the BraTS2020 dataset and showed improved performance in brain tumor region segmentation, particularly in the segmentation of the core and enhancing tumor regions, demonstrating the effectiveness of the proposed approach. Our code is publicly available.

In recent years, the interdisciplinary field of quantum computing has rapidly developed and garnered substantial interest from both academia and industry due to its ability to process information in fundamentally different ways, leading to hitherto unattainable computational capabilities. However, despite its potential, the full extent of quantum computing’s impact on healthcare remains largely unexplored. This survey paper presents the first systematic analysis of the various capabilities of quantum computing in enhancing healthcare systems, with a focus on its potential to revolutionize compute-intensive healthcare tasks such as drug discovery, personalized medicine, DNA sequencing, medical imaging, and operational optimization. Through a comprehensive analysis of existing literature, we have developed taxonomies across different dimensions, including background and enabling technologies, applications, requirements, architectures, security, open issues, and future research directions, providing a panoramic view of the quantum computing paradigm for healthcare. Our survey aims to aid both new and experienced researchers in quantum computing and healthcare by helping them understand the current research landscape, identifying potential opportunities and challenges, and making informed decisions when designing new architectures and applications for quantum computing in healthcare.

The coronavirus disease 2019 (COVID-19) is a severe global pandemic that has claimed millions of lives and continues to overwhelm public health systems in many countries. The spread of COVID-19 pandemic has negatively impacted the human mobility patterns such as daily transportation-related behavior of the public. There is a requirement to understand the disease spread patterns and its routes among neighboring individuals for the timely implementation of corrective measures at the required placement. To increase the effectiveness of contact tracing, countries across the globe are leveraging advancements in mobile technology and Internet of Things (IoT) to aid traditional manual contact tracing to track individuals who have come in close contact with identified COVID-19 patients. Even as the first administration of vaccines begins in 2021, the COVID-19 management strategy will continue to be multi-pronged for the foreseeable future with digital contact tracing being a vital component of the response along with the use of preventive measures such as social distancing and the use of face masks. After some months of deployment of digital contact tracing technology, deeper insights into the merits of various approaches and the usability, privacy, and ethical trade-offs involved are emerging. In this paper, we provide a comprehensive analysis of digital contact tracing solutions in terms of their methodologies and technologies in the light of the new data emerging about international experiences of deployments of digital contact tracing technology. We also provide a discussion on open challenges such as scalability, privacy, adaptability and highlight promising directions for future work.

In this digital age, ports face stiff competition in global supply chain. Smart ports, as high performing ports, utilize information and communications technology (ICT) to provide a wide range of smart applications, resulting in vastly improved vessels and container management among others, which subsequently improve the competitiveness and sustainability of the national economy. While various novel solutions, such as information system and locating system, have been proposed to improve smart port activities, there are several key issues pertaining to ports and port operations that warrant specific attention, particularly greenhouse gases emission, which has accelerated to an alarming level. The urgent need to address such issues is lacking. This article aims to offer a review of the research literature on smart ports, including Internet of Things platform, greenhouse gases emission reduction, energy efficiency enhancement, and so on. The objective is to establish a foundation of existing research conducted on smart ports in order to motivate new research interests in this area. Open issues are also presented to foster new research initiatives on smart ports.

In underwater wireless sensor networks (UWSNs), protocols with efficient energy and reliable communication are challenging, due to the unpredictable aqueous environment. The sensor nodes deployed in the specific region can not last for a long time communicating with each other because of limited energy. Also, the low speed of the acoustic waves and the small available bandwidth produce high latency as well as high transmission loss, which affects the network reliability. To address such problems, several protocols exist in literature. However, these protocols lose energy efficiency and reliability, as they calculate the geographical coordinates of the node or they do not avoid unfavorable channel conditions. To tackle these challenges, this article presents the two novel routing protocol for UWSNs. The first one energy path and channel aware (EPACA) protocol transmits data from a bottom of the water to the surface sink by taking node's residual energy (Re), packet history (Hp), distance (d) and bit error rate (BER). In EPACA protocol, a source node computes a function value for every neighbor node. The most prior node in terms of calculated function is considered as the target destination. However, the EPACA protocol may not always guarantee packet reliability, as it delivers packets over a single path. To maintain the packet reliability in the network, the cooperative-energy path and channel aware (CoEPACA) routing scheme is added which uses relay nodes in packet advancement. In the CoEPACA protocol, the destination node receives various copies from the source and relay(s). The received data at the destination from multiple routes make the network more reliable due to avoiding the erroneous data. The MATLAB simulations results validated the performance of the proposed algorithms. The EPACA protocol consumed 29.01% and the CoEPACA protocol 19.04% less energy than the counterpart scheme. In addition, the overall 12.40% improvement is achieved in the packet's reliability. Also, the EPACA protocol outperforms for packets' latency and network lifetime.

Owning to the vital resources in a harsh and unforeseeable aqueous environment, the network stability and reliability in underwater acoustic wireless sensor networks (UAWSNs) have paramount significance. Stability guarantees the consistent performance of the network node's energy consumption, avoids data loss, packets reception time and network lifetime. The reliability of packet ensures the selection of favorable channel and avoid adverse channel effects, and the vital information is easily obtained from data packets. This paper introduces two new routing schemes for UAWSNs; stable and reliable short-path routing (RSPR) scheme, and cooperative reliable short-path routing (CoRSPR). In RSPR routing, the destination node is selected by considering the weighting function parameters of the highest residual energy, highest SNR, lowest euclidean distance, and least number of neighbor nodes. The scheme reduces the energy consumption due to less number of nodes contribution in the packet advancement process. The RSPR protocol is a non-cooperative technique, where the packets are delivered using a single-path link, which may not be consistently reliable. To cope with this issue, the CoRSPR protocol is proposed, which takes cooperative routing into account, for stable and reliable data delivery. In cooperative routing, the reception of more than one copy of the data packet is involved by the destination node. This reduces the unfavorable channel effects during data delivery. The simulation results show that the proposed schemes achieve better performance in terms of dead nodes, energy left in the battery, packet acceptance ratio, successful receiving of packets at the sink and E-2-E delay.

The feasibility of Federated Learning (FL) is highly dependent on the training and inference capabilities of local models, which are subject to the availability of meaningful and annotated data. The availability of such data is in turn contingent on the tedious and time-consuming annotation job that typically requires the manual analysis of training samples. Active Learning (AL) provides an alternative solution allowing a Machine Learning (ML) model to automatically choose and label the data from which it learns without involving manual inspection of each training sample. In this work, we explore how FL can benefit from unlabelled data available at each participating client using AL. To this aim, we propose an AL-based FL framework by employing and evaluating several AL methods in two different application domains. Through an extensive experimentation setup, we show that AL is equally useful in federated and centralized learning by achieving comparable results with manually labeled data using fewer samples without involving human annotators in collecting training data. We also demonstrated that the proposed method is dataset/application independent by evaluating the proposed method in two interesting applications, namely natural disaster analysis and waste classification, having different properties and challenges. Promising results are obtained on both applications resulting in comparable results against the best-case scenario where each sample is manually analyzed and annotated (Baseline 1), and improvement of 3.1% and 4% with best methods respectively over the training sets with irrelevant images on natural disaster and waste classification datasets (Baseline 2).

The realm of cloud computing has revolutionized access to cloud resources and their utilization and applications over the Internet. However, deploying cloud computing for delay critical applications and reducing the delay in access to the resources are challenging. The Mobile Edge Computing (MEC) paradigm is one of the effective solutions, which brings the cloud computing services to the proximity of the edge network and leverages the available resources. This paper presents a survey of the latest and state-of-the-art algorithms, techniques, and concepts of MEC. The proposed work is unique in that the most novel algorithms are considered, which are not considered by the existing surveys. Moreover, the chosen novel literature of the existing researchers is classified in terms of performance metrics by describing the realms of promising performance and the regions where the margin of improvement exists for future investigation for the future researchers. This also eases the choice of a particular algorithm for a particular application. As compared to the existing surveys, the bibliometric overview is provided, which is further helpful for the researchers, engineers, and scientists for a thorough insight, application selection, and future consideration for improvement. In addition, applications related to the MEC platform are presented. Open research challenges, future directions, and lessons learned in area of the MEC are provided for further future investigation.

Smart health presents an ever-expanding attack surface due to the continuous adoption of a broad variety of Internet of Medical Things (IoMT) devices and applications. IoMT is a common approach to smart city solutions that deliver long-term benefits to critical infrastructures, such as smart healthcare. Many of the IoMT devices in smart cities use Bluetooth technology for short-range communication due to its flexibility, low resource consumption, and flexibility. As smart healthcare applications rely on distributed control optimization, artificial intelligence (AI) and deep learning (DL) offer effective approaches to mitigate cyber-attacks. This paper presents a decentralized, predictive, DL-based process to autonomously detect and block malicious traffic and provide an end-to-end defense against network attacks in IoMT devices. Furthermore, we provide the BlueTack dataset for Bluetooth-based attacks against IoMT networks. To the best of our knowledge, this is the first intrusion detection dataset for Bluetooth classic and Bluetooth low energy (BLE). Using the BlueTack dataset, we devised a multi-layer intrusion detection method that uses deep-learning techniques. We propose a decentralized architecture for deploying this intrusion detection system on the edge nodes of a smart healthcare system that may be deployed in a smart city. The presented multi-layer intrusion detection models achieve performances in the range of 97–99.5% based on the F1 scores.

Modern cities are complex adaptive systems in which there is a lot of dependency and interaction between the various stakeholders, components, and subsystems. The use of digital Information and Communications Technology (ICT) has opened up the vision of smart cities in which the city dwellers can have a better quality of life and the city can be better organized and managed. The deployment of ICT solutions, however, does not automatically or invariably improve the quality of living of the citizens. Analyzing cities as complex systems with various interacting sub-systems can help us understand urban dynamics and the fate of smart cities. We will be able to analyze various policy interventions and ascertain their effectiveness and anticipate potential unintended consequences. In this paper, we discuss how smart cities can be viewed through the lens of systems thinking and complex systems and provide a comprehensive review of related techniques and methods. Along with highlighting the science of cities in light of historic urban modeling and urban dynamics, we focus on shedding light on the smart city complex systems. Finally, we will describe the various challenges of smart cities, discuss the limitations of existing models, and identify promising future directions of work.

With the advances in machine learning (ML) and deep learning (DL) techniques, and the potency of cloud computing in offering services efficiently and cost-effectively, Machine Learning as a Service (MLaaS) cloud platforms have become popular. In addition, there is increasing adoption of third-party cloud services for outsourcing training of DL models, which requires substantial costly computational resources (e.g., high-performance graphics processing units (GPUs)). Such widespread usage of cloud-hosted ML/DL services opens a wide range of attack surfaces for adversaries to exploit the ML/DL system to achieve malicious goals. In this article, we conduct a systematic evaluation of literature of cloud-hosted ML/DL models along both the important dimensions—attacks and defenses—related to their security. Our systematic review identified a total of 31 related articles out of which 19 focused on attack, six focused on defense, and six focused on both attack and defense. Our evaluation reveals that there is an increasing interest from the research community on the perspective of attacking and defending different attacks on Machine Learning as a Service platforms. In addition, we identify the limitations and pitfalls of the analyzed articles and highlight open research issues that require further investigation.

Future buildings will offer new convenience, comfort, and efficiency possibilities to their residents. Changes will occur to the way people live as technology involves people's lives and information processing is fully integrated into their daily living activities and objects. The future expectation of smart buildings includes making the residents' experience as easy and comfortable as possible. The massive streaming data generated and captured by smart building appliances and devices contain valuable information that needs to be mined to facilitate timely actions and better decision making. Machine learning and big data analytics will undoubtedly play a critical role to enable the delivery of such smart services. In this paper, we survey the area of smart building with a special focus on the role of techniques from machine learning and big data analytics. This survey also reviews the current trends and challenges faced in the development of smart building services.

Cyber-physical systems (CPSs) were envisaged as a way to manipulate the objects in the physical world through computer intelligence. This is usually done by providing a communication bridge between actuation and computing elements. This sought after control is hampered not only by the unavoidable certainty found in the physical world but also by the limitations of contemporary communication networks. These limitations hamper fine-grained control of elements that may be separated by large-scale distances. In this regard, soft computing is an emerging paradigm that can help to manage the unreliability of CPS by using techniques, including fuzzy systems, neural networks, evolutionary computation, probabilistic reasoning, and rough sets. We present a comprehensive contemporary review of soft computing techniques for CPS dependability modeling, analysis, and improvement. This paper provides an overview of CPS applications, explores the foundations of dependability engineering, and highlights the potential role of soft computing techniques for CPS dependability with various case studies while also identifying common pitfalls and future directions. In addition, this paper provides a comprehensive survey of the use of various soft computing techniques for making CPS dependable. This paper is timely due to the increasingly central role that CPSs are beginning to play in modern societies and the need to leverage all the relevant methodologies and tools (such as those provided by soft computing) for the development of highly dependable CPS.

While machine learning and artificial intelligence have long been applied in networking research, the bulk of such works has focused on supervised learning. Recently, there has been a rising trend of employing unsupervised machine learning using unstructured raw network data to improve network performance and provide services, such as traffic engineering, anomaly detection, Internet traffic classification, and quality of service optimization. The growing interest in applying unsupervised learning techniques in networking stems from their great success in other fields, such as computer vision, natural language processing, speech recognition, and optimal control (e.g., for developing autonomous self-driving cars). In addition, unsupervised learning can unconstrain us from the need for labeled data and manual handcrafted feature engineering, thereby facilitating flexible, general, and automated methods of machine learning. The focus of this survey paper is to provide an overview of applications of unsupervised learning in the domain of networking. We provide a comprehensive survey highlighting recent advancements in unsupervised learning techniques, and describe their applications in various learning tasks, in the context of networking. We also provide a discussion on future directions and open research issues, while identifying potential pitfalls. While a few survey papers focusing on applications of machine learning in networking have previously been published, a survey of similar scope and breadth is missing in the literature. Through this timely review, we aim to advance the current state of knowledge, by carefully synthesizing insights from previous survey papers, while providing contemporary coverage of the recent advances and innovations.

With the rapid increase in the popularity of social networks, the propagation of rumors is also increasing. Rumors can spread among thousands of users immediately without verification and can cause serious damages. Recently, several research studies have been investigated to control online rumors automatically by mining rich text available on the open network with deep learning techniques. In this paper, we conducted a systematic literature review for rumor detection using deep neural network approaches. A total of 108 studies were retrieved using manual research from five databases (IEEE Explore, Springer Link, Science Direct, ACM Digital Library, and Google Scholar). The considered studies are then examined in our systematic review to answer the seven research questions that we have formulated to deeply understand the overall trends in the use of deep learning methods for rumor detection. Apart from this, our systematic review also presents the challenges and issues that are faced by the researchers in this area and suggests promising future research directions. Our review will be beneficial for researchers in this domain as it will facilitate researchers' comparison with the existing works due to the availability of a complete description of the used performance matrices, dataset characteristics, and the deep learning model used per each work. Our review will also assist researchers in finding the available annotated datasets that can be used as benchmarks for comparing their new proposed approaches with the existing state-of-the-art works.

Maritime networks establish wireless multi-hop networks to provide wireless broadband service at sea, connecting various kinds of ships, maritime buoys, and beacons. The maritime networks possess two distinguishing characteristics highly affected by maneuver at sea-dynamic link quality and bandwidth constrained, and dynamic network topology-that warrant specific attention. Unlike land vehicles, maneuver at sea is affected by sea surface movement and wave occlusions, which can cause unstable environment with a high rate of link breakages caused by low link stability, as well as low and highly variable bandwidth. In spite of the need to achieve performance close to high-speed terrestrial wireless broadband service on land, there is only a perfunctory effort to investigate maritime networks. There is an urgent need to refresh the interest to investigate, as well as to further enhance, maritime networks. This paper presents a review of the limited research works of this topic, which revolve around the networking issues in the link, network, and upper layers, in the literature. The objective is to establish a foundation in order to motivate a new research interest in maritime networks. Open issues are also presented to foster new research initiatives in this burgeoning and exciting area.

The Internet of Things (IoT) has disrupted the IT landscape drastically, and Long Range Wide Area Network (LoRaWAN) is one specification that enables these IoT devices to have access to the Internet. Former security analyses have suggested that the gateways in LoRaWAN in their current state are susceptible to a wide variety of malicious attacks, which can be notoriously difficult to mitigate since gateways are seen as obedient relays by design. These attacks, if not addressed, can cause malfunctions and loss of efficiency in the network traffic. As a solution to this unique problem, this paper presents a novel certificate authentication technique that enhances the cyber security of gateways in the LoRaWAN network. The proposed technique considers a public key infrastructure (PKI) solution that considers a two-tier certificate authority (CA) setup, such as a root-CA and intermediate-CA. This solution is promising, as the simulation results validate that about 66.67% of the packets that are arriving from an illegitimate gateway (GW) are discarded in our implemented secure and reliable solution.

Disaster situations require a prompt emergency response. The failure of the legacy (wired) emergency response systems to efficiently cope in such time-critical environments has led to a great interest in wireless-based emergency response systems. In this regard, a number of wireless technologies and systems - each with its peculiar characteristics and pros and cons - have been proposed for use in various emergency response situations. This paper presents, to the best of our knowledge, the first comprehensive survey of the research on this practically important topic. We motivate the use of wireless technology for emergency response and present a comparative analysis of the available wireless technologies. After outlining the scope and requirements of emergency response systems, we provide an overview of the architectures and features of wireless-based emergency response systems and then compare different wireless technologies for different emergency settings. We provide not only details about individual emergency response systems and technologies but also expend efforts to show the forest for the trees to emergency response practitioners - in particular, we provide general high-level guidelines that can help in deciding the right technology for a particular situation, and discuss potential pitfalls from the large-scale emergency-network deployment experience of one of the authors. The emergency response operation is split into five functionalities: we present general guidelines that would help public safety agencies in choosing suitable wireless technologies for each one of them. Finally, we also highlight opportunities and identify the potential areas for further investigation.

Ultra-dense small cell networks represent a key future network solution that can help meet the exponentially rising traffic requirements of modern wireless networks. Backhauling these small cells are an emerging challenge to the extent that various cells are likely to have different backhaul constraints. The user-centric backhaul scheme has been proposed in the literature to jointly exploit the diversity in users' requirement and backhaul constraints. In this paper, we propose a novel scheme, termed the memory-based hybrid scheme, which additionally also exploits the predictability in a user's mobility. We compare the novel scheme to two variants of memory-less user-centric backhaul implementations and show significant gains in convergence time (15%), user-centric KPIs (51% and 82%) at the negligible cost 2% loss in cumulative throughput. The novel scheme requires additional memory in user-devices to store learned values, which is nonetheless well justified in view of the considerable gains achieved.

Cognitive radio (CR) enables unlicensed users to sense for and access underutilized licensed channels (or white spaces) owned by the licensed users in an opportunistic manner. Clustering segregates nodes in a network into logical groups called clusters. In CR networks (CRNs), larger cluster size improves network scalability thereby contributing to reduced routing overhead; however, it reduces cluster stability as the number of available common channels in a cluster reduces resulting in increased number of re-clusterings and clustering overhead. This paper presents our proposed first-of-its-kind cluster size adjustment scheme based on an artificial intelligence approach called reinforcement learning. The proposed scheme adapts the cluster size with the amount of white spaces as time goes by in order to improve network scalability and cluster stability in CRNs. Due to the lack of progress in the investigation of cluster size adjustment schemes in the literature, this paper also analyzes their attributes, and then presents such schemes investigated in various kinds of distributed wireless networks. Simulation results show that our proposed scheme improves network scalability by creating larger clusters, and improves cluster stability by reducing the number of re-clusterings (i.e., the number of cluster splits) and clustering overhead, while reducing interference between licensed and unlicensed users in CRNs.

This paper considers a line-of-sight underwater acoustic (UWA) sensor network consisting of M underwater sensor nodes randomly deployed according to uniform distribution within a vertical half-disc (the so-called trusted zone). The sensor nodes report their sensed data to a sink node on water surface on a shared UWA reporting channel in a time-division multiple-access fashion, while an active-yet-invisible adversary (so-called Eve) is present in the close vicinity who aims to inject malicious data into the system by impersonating some Alice node. To this end, this paper first considers an additive white Gaussian noise (AWGN) UWA channel, and proposes a novel, multiple-feature-based, two-step method at the sink node to thwart the potential impersonation attack by Eve. Specifically, the sink node exploits the noisy estimates of the distance, the angle of arrival, and the location of the transmit node as device fingerprints to carry out a number of binary hypothesis tests (for impersonation detection) as well as a number of maximum-likelihood (ML) hypothesis tests (for transmitter identification when no impersonation is detected). We provide closed-form expressions for the error probabilities (i.e., the performance) of most of the hypothesis tests. We then consider the case of a UWA with colored noise and frequency-dependent pathloss, and derive a ML distance estimator as well as the corresponding Cramer-Rao bound. We then invoke the proposed two-step, impersonation detection framework by utilizing distance as the sole feature. Finally, we provide detailed simulation results for both AWGN UWA channel and the UWA channel with colored noise. Simulation results verify that the proposed scheme is indeed effective for a UWA channel with the colored noise and frequency-dependent pathloss.

Despite the evolution of cellular networks, spectrum scarcity and the lack of intelligent and autonomous capabilities remain a cause for concern. These problems have resulted in low network capacity, high signaling overhead, inefficient data forwarding, and low scalability, which are expected to persist as the stumbling blocks to deploy, support and scale next-generation applications, including smart city and virtual reality. Fifth-generation (5G) cellular networking, along with its salient operational characteristics - including the cognitive and cooperative capabilities, network virtualization, and traffic offload - can address these limitations to cater to future scenarios characterized by highly heterogeneous, ultra-dense, and highly variable environments. Cognitive radio (CR) and cognition cycle (CC) are key enabling technologies for 5G. CR enables nodes to explore and use underutilized licensed channels; while CC has been embedded in CR nodes to learn new knowledge and adapt to network dynamics. CR and CC have brought advantages to a cognition-inspired 5G cellular network, including addressing the spectrum scarcity problem, promoting interoperation among heterogeneous entities, and providing intelligence and autonomous capabilities to support 5G core operations, such as smart beamforming. In this paper, we present the attributes of 5G and existing state of the art focusing on how CR and CC have been adopted in 5G to provide spectral efficiency, energy efficiency, improved quality of service and experience, and cost efficiency. This main contribution of this paper is to complement recent work by focusing on the networking aspect of CR and CC applied to 5G due to the urgent need to investigate, as well as to further enhance, CR and CC as core mechanisms to support 5G. This paper is aspired to establish a foundation and to spark new research interest in this topic. Open research opportunities and platform implementation are also presented to stimulate new research initiatives in this exciting area.

Technology is an extremely potent tool that can be leveraged for human development and social good. Owing to the great importance of environment and human psychology in driving human behaviour, and the ubiquity of technology in modern life, there is a need to leverage the insights and capabilities of both fields together for nudging people towards a behaviour that is optimal in some sense (personal or social). In this regard, the field of persuasive technology, which proposes to infuse technology with appropriate design and incentives using insights from psychology, behavioural economics, and human-computer interaction holds a lot of promise. Whilst persuasive technology is already being developed and is at play in many commercial applications, it can have the great social impact in the field of Information and Communication Technology for Development (ICTD) which uses Information and Communication Technology (ICT) for human developmental ends such as education and health. In this paper we will explore what persuasive technology is and how it can be used for the ends of human development. To develop the ideas in a concrete setting, we present a case study outlining how persuasive technology can be used for human development in Pakistan, a developing South Asian country, that suffers from many of the problems that plague typical developing country.

The mHealth trend, which uses mobile devices and associated technology for health interventions, offers unprecedented opportunity to transform the health services available to people across the globe. In particular, the mHealth transformation can be most disruptive in the developing countries, which is often characterized by a dysfunctional public health system. Despite this opportunity, the growth of mHealth in developing countries is rather slow and no existing studies have conducted an in-depth search to identify the reasons. We present a comprehensive report about the factors hindering the growth of mHealth in developing countries. Most importantly, we outline future strategies for making mHealth even more effective. We are also the first to conduct a case study on the public health system of Pakistan showing that mHealth can offer tremendous opportunities for a developing country with a severe scarcity of health infrastructure and resources. The findings of this paper will guide the development of policies and strategies for the sustainable adoption of mHealth not only in Pakistan but also for any developing country in general.

We live in a world with growing disparity in the quality of life available to people in the developed and developing countries. Healthcare in the developing world is fraught with numerous problems such as the lack of health infrastructure, and human resources, which results in very limited health coverage. The field of health informatics has made great strides in recent years towards improving public health systems in the developing world by augmenting them with state-of-the-art information and communication technologies (ICT). Through real-world deployment of these technologies, there is real hope that the health industry in the developing world will progress from its current, largely dysfunctional state to one that is more effective, personalized, and cost effective. Health informatics can usher a new era of personalized health analytics, with the potential to transform healthcare in the developing world. In conjunction with mHealth and eHealth, many other important health informatics trends—such as artificial intelligence (AI), machine learning (ML), big data, crowdsourcing, cloud computing—are also emerging. Exponentially growing heterogeneous data, with the help of big data analytics, has the potential to provide descriptive, predictive, and prescriptive health insights as well as enable new applications such as telemedicine and remote diagnostics and surgery. Such systems could enhance the overall process of monitoring, diagnosis, and prognosis of diseases.

With the world becoming flat with fluid boundaries, engineers have to be global in their outlook and their pedigree. Due to the need for international acceptance of engineering qualification, the incorporation of Outcome-Based Education (OBE) has become common and global accreditation treaties such as the Washington Accord have been ratified. Further, it becomes important, especially for an engineering university with a global outlook preparing its students for global markets, to ensure that its graduates attain the planned outcomes. Additionally, the higher education institutions need to make sure that all the stakeholders, including students, parents, employers, and community at large, are getting a quality educational service, where quality is categorized as (1) product-based ensuring that the graduate attained the planned outcomes and skills, and (2) process-based keeping an eye on whether the process is simple, integrated, and efficient. The development of quality movements, such as Total Quality Movement (TQM), Six Sigma, etc., along with quality standards such as ISO 9001 has been instrumental in improving the quality and efficiency in the fields of management and services. Critical to the successful deployment of a quality culture is the institutionalization of an integrated Quality Management System (QMS) in which formally documented processes work according to the Vision and Mission of an institute. At the same time, commitment to Continuous Quality Improvement (CQI) to close the loop through effective feedback, would ensure that the planned outcomes are attained to the satisfaction of all the stakeholders, and that the process overall is improving consistently and continuously. The successful adoption of quality culture requires buy-in from all the stakeholders (and in particular, the senior leadership) and a rigorous training program. In this paper, we provide a review of how a QMS may work for the provision of quality higher education in a 21st-century university.

In this paper, we strive to provide a primer for students on how to thrive and learn effectively in engineering education in the volatile, uncertain, complex, and ambiguous (VUCA) times following the onset of the COVID-19 global pandemic, which has disrupted the educational enterprise massively with universities physically closing in many parts of the world and students and faculty transitioning to remote learning. Although the immediate audience assumed in this paper comprises engineering students (such as those enrolled in electrical, electronics, or computer engineering programs) studying in an outcome-based education (OBE) environment—the global educational paradigm mandated by the Washington Accord that aims to standardize engineering competencies in terms of the attained student learning outcomes—the presented ideas are more general and broadly useful for learners of all types. We will describe seven evidence-based steps that the students can adopt to thrive in OBE settings in these challenging times. The main contribution of this paper is practical: we present a synthesis of the vast research literature on effective student learning in normal, online, and disrupted settings to present practical insights that students can leverage to substantially improve their learning. At the end of the paper, we also present a discussion of important issues related to remote teaching and online education such as ensuring equity and the handling of lab work for engineers in such settings (e.g., through simulators and virtual labs).

Cognitive radio (CR) enables unlicensed users to explore and exploit underutilized licensed channels (or white spaces). While multi-hop CR network has drawn significant research interest in recent years, majority work has been validated through simulation. A key challenge in multi-hop CR network is to select a route with high quality of service (QoS) and lesser number of route breakages. In this paper, we propose three route selection schemes to enhance the network performance of CR networks, and investigate them using a real testbed environment, which consists of universal software radio peripheral and GNU radio units. Two schemes are based on reinforcement learning (RL), while a scheme is based on spectrum leasing (SL). RL is an artificial intelligence technique, whereas SL is a new paradigm that allows communication between licensed and unlicensed users in CR networks. We compare the route selection schemes with an existing route selection scheme in the literature, called highest-channel (HC), in a multi-hop CR network. With respect to the QoS parameters (i.e., throughput, packet delivery ratio, and the number of route breakages), the experimental results show that RL approaches achieve a better performance in comparison with the HC approach, and also achieve close to the performance achieved by the SL approach.

Customer retention is a major issue for various service-based organizations particularly telecom industry, wherein predictive models for observing the behavior of customers are one of the great instruments in customer retention process and inferring the future behavior of the customers. However, the performances of predictive models are greatly affected when the real-world data set is highly imbalanced. A data set is called imbalanced if the samples size from one class is very much smaller or larger than the other classes. The most commonly used technique is over/under sampling for handling the class-imbalance problem (CIP) in various domains. In this paper, we survey six well-known sampling techniques and compare the performances of these key techniques, i.e., mega-trend diffusion function (MTDF), synthetic minority oversampling technique, adaptive synthetic sampling approach, couples top-N reverse k-nearest neighbor, majority weighted minority oversampling technique, and immune centroids oversampling technique. Moreover, this paper also reveals the evaluation of four rules-generation algorithms (the learning from example module, version 2 (LEM2), covering, exhaustive, and genetic algorithms) using publicly available data sets. The empirical results demonstrate that the overall predictive performance of MTDF and rules-generation based on genetic algorithms performed the best as compared with the rest of the evaluated oversampling methods and rule-generation algorithms.

With today’s computer networks becoming increasingly dynamic, heterogeneous, and complex, there is great interest in deploying artificial intelligence (AI) based techniques for optimization and management of computer networks. AI techniques—that subsume multidisciplinary techniques from machine learning, optimization theory, game theory, control theory, and meta-heuristics—have long been applied to optimize computer networks in many diverse settings. Such an approach is gaining increased traction with the emergence of novel networking paradigms that promise to simplify network management (e.g., cloud computing, network functions virtualization, and software-defined networking) and provide intelligent services (e.g., future 5G mobile networks). Looking ahead, greater integration of AI into networking architectures can help develop a future vision of cognitive networks that will show network-wide intelligent behavior to solve problems of network heterogeneity, performance, and quality of service (QoS).