Integrated Sensing and Communication (ISAC) technology is pivotal in advancing multifunctional radio applications for nextgeneration wireless communications. One of the primary challenges in ISAC systems lies in understanding and interpreting wireless channels for both sensing and communication purposes. A clear relationship between these channels would allow ISAC systems to reduce channel estimation complexity. This paper introduces a comprehensive framework for reconstructing communication channels using monostatic sensing channels within an ISAC system. We first consider the frequency spacing of an ISAC system to geometrically illustrate the synergies and differences between the sensing and communication regions. Subsequently, we construct a communication propagation channel based on overlap determined by this geometric relationship and a simplified monostaticto-bistatic equivalence theorem. Lastly, we combine multipaths located in the same delay bin to construct channel state information (CSI), taking bandwidth limitation into account. Simulations demonstrate that the proposed framework efficiently generates communication channels across various configurations, with numerical results highlighting the impacts of environmental and system constraints on key channel characteristics.

This manuscript summarizes work on the Capsule Vision Challenge 2024 by MISAHUB. To address the multi-class disease classification task, which is challenging due to the complexity and imbalance in the Capsule Vision challenge dataset, this paper proposes CASCRNet (Capsule endoscopy-Aspp-SCR-Network), a parameter-efficient and novel model that uses Shared Channel Residual (SCR) blocks and Atrous Spatial Pyramid Pooling (ASPP) blocks. Further, the performance of the proposed model is compared with other well-known approaches. The experimental results yield that proposed model provides better disease classification results. The proposed model was successful in classifying diseases with an F1 Score of 78.5% and a Mean AUC of 98.3%, which is promising given its compact architecture.

The limitations of fifth generation (5G) networks in simultaneously addressing the growing demands of diverse applications, including ultra-reliable low-latency communication (URLLC), massive machine-type communication (mMTC), and high-throughput services, drive the exploration of 6G technologies. Given the multifaceted nature of 6G use cases and the proliferation of edge functions, such as computing nodes, charging nodes, and sensing targets, selecting an optimal multiple access (MA) technique becomes paramount. This article reviews the literature to identify key milestones and insights from the recent progression of MA technology, exploring robust research directions and optimization aspects of developing an intelligent, unified MA framework for 6G. The article examines promising access schemes such as rate-splitting multiple access (RSMA) and non-orthogonal multiple access (NOMA) and discusses the pivotal role of the latest paradigms, such as artificial intelligence (AI) and Open radio access network (RAN), in optimizing the MA framework.

Wireless sensor networks (WSNs) were originally envisioned to consist of a massive number of diminutive, lowcost, and long-life wireless sensors. However, this vision has faced significant obstacles, particularly in transferring sensing information to data centers. This article proposes a novel protocol, termed relaying-as-a-service (RaaS), to leverage public smart devices to relay sensing information to data centers. By offloading major communications and networking operations to these more powerful smart devices and well-developed fifth generation (5G) or sixth generation (6G) networks, the size and cost of sensors can be significantly reduced, and the need for extensive network infrastructure can be eliminated. This reduction in complexity can greatly streamline the development and implementation of WSNs and Internet of Things (IoT) solutions. Furthermore, using low-power sensors enables large-scale frequency-reuse, which can help alleviate the load on the already congested spectrum.

Quantum computing is poised to revolutionize computational performance and capabilities, offering unprecedented efficiency in solving complex problems that surpasses classical approaches. This potential is particularly evident in fields such as optimization and AI. This chapter delves into critical priority areas in quantum computing, including the development and application of quantum software tools. We place a strong emphasis on reimagining the use of quantum computing for modeling and simulation, sensing, and secure communication. Furthermore, we explore current trends like quantum communications for 6G networks, quantum cloud and serverless computing, scalable qubit arrays, and the pursuit of robust and reliable quantum systems. Lastly, we address emerging research areas and the open challenges ahead, encompassing advancements in foundational theory, education, ethical and societal considerations, and pathways to commercialization.

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that adversely affects the healthy development of children. The current clinical diagnosis of ASD relies on demographics and descriptions of clinical symptoms and assessments based on rating scales, which are both time-consuming and subjective. Previous research reported atypical visual patterns exhibited by ASD children, implying that eye-tracking could be a potential tool to assist in distinguishing ASD from others. However, most studies only examined a singular dimension of visual trajectory, neglecting the possible use of comprehensive information on eye scan paths. Given this, the present study introduced the FF-ASDNET - a framework designed for the automated screening of ASD individuals leveraging eye-tracking technologies in conjunction with deep representation learning and multi-scale feature fusion. Eye movement trajectory was first objectively quantified from multi-scale perspectives, including Gaze Movements (GM), Spatial Attention Distribution (SAD), and Temporal Visual Information (TVI). The residual network (ResNet) and the rectangular residual convolutional temporal network (RecResTCNN) were then proposed to learn the features of the three representations. Finally, the features of the three scales were fused using a convolutional neural network. The proposed design achieved an AUC of 85.96% and an F1 score of 0.85 on an independent dataset, showing that fusing multi-scale representations of eye scan trajectories could significantly improve the model’s performance, rendering the development of an automated ASD screening plausible. In summary, FF-ASDNET is a promising tool for screening ASD individuals, with potential applications in clinical diagnosis and intervention.

Ransomware continues to pose a significant threat to global cybersecurity, with increasingly sophisticated variants that can evade traditional detection methods. Dynamic Code Sequence Mapping (DCSM) offers a novel solution that addresses this challenge through the real-time monitoring of execution behavior, allowing for the detection of ransomware based on its dynamic characteristics rather than relying on static signatures. DCSM operates through a modular framework that analyzes code sequences, identifies patterns associated with ransomware activity, and triggers protective measures before the malware can execute its payload. Extensive testing has demonstrated the system's ability to maintain high detection accuracy, even in the presence of advanced evasion techniques such as code obfuscation and polymorphism, while minimizing false positives. Moreover, DCSM's adaptability and integration potential within existing cybersecurity infrastructures make it a versatile tool capable of responding to the continuously evolving ransomware landscape. The overall impact of DCSM lies in its ability to enhance ransomware detection capabilities, providing a more resilient and adaptive approach to mitigating one of the most persistent threats in modern cybersecurity.

Due to the inherent complexity arising from powertrain, aerodynamic, and unseen driving scenarios, efficient coordination of connected hybrid electric vehicles (C-HEVs) platooning encounters complex challenges. Existing studies primarily do not consider thermal load impacts. However, this overlooks the interconnection between the C-HEVs related to powertrain power and thermal loads, which can hinder effective optimization and collaboration across C-HEV platoon members. To fully address this gap, a multi-agent deep reinforcement learning (DRL)-based joint cooperative adaptive cruise control (CACC) with an integrated thermal and energy management (ITEM) framework is proposed. Additionally, to decrease the sample complexity and accelerate policy learning, the vehicle-level (string stability and time headway) and powertrain-level (energy consumption and health indicator) information are integrated into the reward function to design a safety-and physics-informed DRL framework. To meticulously assess the mutual impact of thermal load and management on the platooning optimization problem, air drag and airflow reductions related to inter-vehicle distance (IVD) are accounted for with a high-fidelity model provided by Autonomie. Numerical and hardware-in-the-loop (HIL) experiments demonstrate that the proposed CACC+ITEM algorithm outperforms the classic DRL version in stability, energy-saving, driving comfort, and battery longevity.

In an era of growing security concerns, with the context of Ad hoc wireless communication which is common in the domain of Internet of Things (IoT) or remote areas where Internet is not a viable option, establishing trust between two entities is challenging. Not only the protocol should be formidable to deny connection for untrusted entities but also cost-effective and practically viable. This paper demonstrates how zero-trust framework can be employed in ad hoc wireless communications through a surveillance robot with remote control and live video stream. We achieve zero-trust connection using mutual-Transport Layer Security (mTLS) protocol between the endpoints. This secure connection is then used to exchange control messages for robot movement and for the live video stream. The endpoints communicate over Ad hoc Wi-Fi and hence there is no need for router or Wi-Fi access points or internet. Since Wi-Fi has limited range, this work can be further extended over long-distance communication where a network of Ad Hoc peers employs a routing protocol for transfer of packets between two endpoints within their network.

The increasing complexity of linguistic tasks has driven a need for more adaptable and contextually aware language models, capable of producing accurate and coherent outputs even when faced with ambiguous or evolving input. Traditional token embedding methods, which rely on static representations, often fall short in capturing the complex relationships between tokens and their contextual dependencies. Introducing a novel approach, Dynamic Token Embedding Sampling (DTES), allows embeddings to be dynamically adjusted during inference based on the evolving input sequence, offering a significant improvement in predictive accuracy and reduction in hallucinations. Through seamless integration into transformer architectures, DTES enhances both the flexibility and efficiency of token processing, ensuring more precise and contextually relevant outputs across various natural language processing tasks. Experimental results demonstrate considerable improvements in key metrics, including accuracy, inference speed, and factual consistency, further highlighting the potential of DTES to optimize language models for real-world applications. The method's scalability and adaptability across diverse linguistic domains make it a compelling advancement in the ongoing effort to improve the performance and reliability of language models in increasingly complex environments.

Despite the frenetic pace of e-commerce growth, over 80% of grocery sales remain carried out in physical retail environments around the globe. This traditional retail structure causes a number of serious challenges: frenzied product searches, crowded aisles, and long lines at the tills-all of which severely affect customer satisfaction. On account of such deficiency mentioned above, here we are presenting TechCart: an intelligent shopping framework to help improve the in-store shopping experience of a customer who utilizes AI, deep learning, and wireless communication technologies. Integration of independent navigation, immediate item recognition, personalized recommendations, and budget management takes place via a high-level mobile application. The architectures of this system are: the intelligent shopping cart, the mobile application, and the central server that make easy intercommunication and information exchange between these parts happen. From the results of evaluation, TechCart could identify 5,351 objects with a Mean Average Precision of 84.59%. This, therefore, means that this application is also full of producing results with strong accuracy in object detection. The grocery items commonly detected have a realized accuracy of 95%. The average time saved in shopping was 30% as perceived by the users.

This paper presents a method to determine the impact of low-voltage direct current circuit breakers on the availability of the dc bus. Low-voltage direct current is gaining traction in industry, and protection is an important aspect. New circuit breaker technologies are extensively researched, but the impact of this speed of interruption on direct current bus availability has not yet been investigated. This paper demonstrates how to determine the fault clearance probability, which is used to set up the state transition diagrams for the feeder states. These state transition diagrams are solved using the semi-Markov process. The solutions of the semi-Markov process are used in a UGO method to determine the availability of the direct current bus. Furthermore, it is shown that the results depend on the length of the feeder and the capacitance of the direct current bus. For feeders with a fault clearance probability lower than 100percent, the number of feeders also influences the availability of the direct current bus.

Introduction: Natural Language Processing (NLP) has the potential to bring about a quantum shift to the field of mental health. NLP techniques engender the birth of strategies to analyze text which may bode significant mental health outcomes, individually as well as socially. It can help us anticipate and intercept fallouts associated with any kind of deterioration in mental health statistics, most commonly depression and suicide, leading to better social upshots. It also establishes pathbreaking approaches to preventing, detecting, diagnosing, and treating various mental health conditions. This book article provides an extensive overview of the applications of NLP approaches in mental health. It will focus on both well-established methods and emerging trends within this area.Scope and Objectives: The article aims to explore the wide range of NLP applications in mental health, inter alia, sentiment analysis, emotion recognition, and predictive modeling. It discusses the utility of these applications in mental health screening, diagnosis, treatment, and monitoring. The article examines the role of NLP in identifying specific mental health conditions such as depression, anxiety, post-traumatic stress disorder (PTSD), and schizophrenia, as well as its integration into therapy settings and self-help tools like accessible online chatbots that assist with behavioral and mental struggles.Key Applications of NLP in Mental Health: There is a plethora of applications of NLP that can be thought of in the field of mental health. A summary of some of these applications is hereby provided. Potential applications of sentiment analysis, a practice in NLP, include detection of mood fluctuations and mental states by analyzing user-generated content. Emotion recognition in text identifies specific feelings to assist in affect assessments. Depression and suicide risk detection, assessment, and prevention, including detection of suicidal ideation in text data, are also prospective applications. The text data fundamentally includes social media posts and online communication. Cognitive Behavioral Therapy (CBT) chatbots automate therapeutic interventions by acting as digital or virtual therapists. Automated analysis of therapy transcripts helps us discern corrective insights. NLP is also employed in mental health screening through surveys when text-based responses to online posts and publications are linguistically analyzed. The identification of PTSD symptomatology involves detecting signs of PTSD in text, especially text data generated by veterans, because they form a category most susceptible to PTSD. Predictive modeling for mental health crises uses NLP to predict potential emergencies, most commonly suicide, crimes like manslaughter, incidents involving bodily harm to oneself or to another person, and vengeful destruction of valuable property through arson and other means. Detection of substance abuse identifies discussions related to sale, purchase, use and addiction of drugs in online forums, telephonic conversations, and conversations incidentally overheard and recorded in public spaces. Monitoring relapse risks entails tracking recovery and relapse risks in patients. Among other applications of NLP in the domain mental healthcare, early detection of psychological conditions like bipolar disorder (BPD), which are enhanced by NLP techniques which brandish the facility to identify the occurrence of such ailments using text and speech.NLP methodologies: The article aims to throw light on a variety of NLP methodologies including text classification, sentiment analysis, emotion detection algorithms, statistical machine learning models, and deep learning architectures. Tools such as speech-to-text analytics, topic modeling, dependency parsing, and vectors and word embeddings are highlighted to illustrate their use in mental health applications. The article will further explore the integration of machine learning and deep learning methods, such as Recurrent Neural Networks (RNNs), Long Short-Term Memory networks (LSTMs) and transformers, in mental health domains like diagnostics.NLP includes several methods for analyzing applications in mental health. Sentiment analysis detects emotional tones in text. This helps in various aspects connected to mental health, notably identifying signs of anxiety. Topic modeling demystifies common themes in mental health discussions. Text classification categorizes textual data and helps discern mental health insights from it. The textual data includes social media posts, hospital-obtained data like patient notes and therapy transcripts, suicide letters and voice notes of people just before they attempted suicide and so on. Named entity recognition supplements and enhances the amount and relevance of information obtained from clinical records. Chatbots based on advanced language models, especially large language models (LLMs) like Mistral, simulate therapeutic conversations. They are also used in automating medical diagnostics, and to provide self-help advice to caregivers of patients undergoing mental health treatments. Speech and audio analysis detect emotional states through changes in tone and pitch. Longitudinal analysis tracks language use over time to monitor mental health progression and intervention effectiveness, while cross-sectional analysis investigates the similarities and differences in mental health outcomes over various regions, populations, languages, cultures, ethnicities, and other determinants of social identity.Challenges: The application of NLP within mental health encompasses several significant challenges. Data privacy and security are paramount challenges. Most data analyzed involves personal and confidential information of potential mental health patients and sometimes also of medical staff and others. Stringent measures must be put in place to protect sensitive information and comply with regulations such as Health Insurance Portability and Accountability Act (HIPAA) in the United States and General Data Protection Regulation (GDPR) in Europe. Issues associated with data quality and completeness arise from gathering such data which is noisy, incomplete, and inconsistent. These issues necessitate robust data cleaning and preprocessing within the NLP pipeline. Fairness concerns are critical, as NLP models can possibly inherit biases from training data (and, in case of transfer learning, from parent models). Unresolved biases in training data may cause class imbalances. Inherent class imbalances are handled using tools like Synthetic Minority Oversampling Technique (SMOTE). Biases negatively influence the accuracy and fairness of predictions, whereby the models beg regular auditing and updating. The context and nuance of mental health expressions can be challenging for NLP models to interpret accurately. It highlights the requirement of generating advanced contextual embeddings and vectors. Ambiguity and polysemy in language, where terms can have multiple meanings, pose difficulties in text analysis. This needs sophisticated semantic analysis to resolve. Integration with clinical practice involves complexities in adapting to novel technologies and affirming their utility. It necessitates user-friendly interfaces and seamless Emergency Help Response (EHR) system integration. Ethical concerns, such as potential misuse of data, emphasize the demand for establishing ethical guidelines. Generalization and adaptation of models to miscellaneous populations is another challenge. This involves continuous adaptation and retraining of the model using various datasets. Real-time processing stipulates efficiency and scalability to confront computational challenges. The interpretation of results from NLP models can be arduous, particularly with complex language patterns. This necessitates the use of explainable AI (XAI) methods like Shapley Additive Explanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME). Cultural and linguistic variability not only affects the prevalence of mental health issues but also the effectiveness of NLP tools across different demographics. It calls for customization and validation of architectures and algorithms across varied populations. Reliability of NLP tools through rigorous testing must be ensured to validate their clinical relevance.Impact and Future Directions: The article emphasizes how NLP tools provide actionable inferences in mental health. They assist in enhancing early detection of mental health conditions, in automating therapy support, in providing new opportunities for large-scale mental health monitoring, and so on. Current cutting-edge advancements and future applications of NLP, including the integration of multimodal data (e.g., combining text, audio, video, music, images etc.), the use of more sophisticated machine learning models (e.g., Transformers, Generative Adversarial Networks or GANs, Graph Neural Networks or GNNs), and ethical considerations surrounding data privacy and mental health, are discussed. The article lays weight to the fact that Artificial Intelligence (AI) must be used responsibly in the domains of mental health, especially in interventions to prevent incidence of depression and suicide. Conclusion: This article is an endeavor to provide a comprehensive exploration of the growing intersection of NLP and mental health. It attempts to provide a general assessment of the potential applications, benefits, challenges, and future directions of NLP in mental health care. Key areas of application are elaborated. This article contributes to the existing body of knowledge as well as to the ongoing research and development that goes on in this rapidly evolving field.

A novel, compact multiple-input-multiple-output (MIMO) antenna design, occupying a mere 25 × 39 mm², is proposed for ultra-wideband (UWB) applications. This innovative antenna comprises dual square-shaped radiating elements and a modified T-shaped stub, incorporating a vertical slot to mitigate mutual coupling. Key performance metrics, including impedance matching, inter-port mutual coupling, voltage standing wave ratio (VSWR) and correlation coefficient, are scrutinized. Simulation outcomes reveal an impressive impedance bandwidth spanning 2.5 to 11.8 GHz (S₁₁ ≤-10 dB), accompanied by an exceptionally low envelope correlation coefficient (<0.002) across the entire frequency spectrum. Notably, inter-element isolation exceeds-23 dB throughout the operational band. The proposed UWB MIMO antenna efficiently covers prominent frequency bands, encompassing WCDMA, WLAN, WiMax and UWB, making it an attractive solution for multi-standard wireless communication systems.

6G communications networks will be highly decentralized and distributed in nature. Existing security policies, procedures, and technologies, however, are mostly based on centralized control. Such centralized security will have serious consequences from two perspectives in 6G. First, the security systems will not be able to meet the strict requirements, such as latency, of emerging services, for example industrial and vehicular communications. Second, centralized security will lead to high consumption of resources, such as spectrum, in securing distributed systems in 6G. Therefore, in this paper we discuss intriguing challenges arising from security centralization that lead to another pressing challenge of the time, i.e., sustainability, as a consequence. Furthermore, we evaluate centralized and decentralized artificial intelligence (AI)-based security to demonstrate the resource and efficiency trade-off. Ultimately, this paper presents important research insights and visions that can pave the way toward distributed security in 6G.

Amorphous magnetic alloys are now widely used in electronics, electrical engineering, robotics, sensors, etc. They are mainly used in the form of wires, tapes and foils. The technology of manufacturing products from amorphous magnetic alloys is constantly improving and new methods are being implemented. In our work, microspirals (microsprings) made from amorphous microwires are studied. Measurements of hysteresis loops for different types of microspirals have been carried out and the relationship between the hysteresis anisotropy and the parameters of the spirals is discussed. Special attention is given to the study of the magnetic impedance. It is found that the field dependence of the impedance of microspirals is asymmetric, and the sign of the asymmetry is related to the chirality of the spirals. The possibility of using amorphous magnetic microspirals for the development of sensor and microrobotic elements is discussed.

Determining the electric field distribution using electromagnetic solvers requires significant computational effort, a high time consumption, and operator training. In this paper, a deep learning (DL)-based method is specifically developed to reconstruct the electric field distribution generated by antennas, based on the dielectric properties of the breast. ResNet and UNet were employed to predict the electric field within complex breast structures. Nine distinct breast dielectric models were utilized to create the initial dataset. A novel data augmentation technique was applied and the size was increased by three to address the challenge of limited dataset size. The electric field distribution data were generated using commercially available finite element method-based electromagnetic simulations. DL architectures were customized to fit the study and trained over 4374 breast slices. ResNet demonstrated an average Signal-to-Noise Ratio (SNR) of 23.7±6.7 dB and an average Normalized Mean Squared Error (NMSE) of 0.0029±0.0039 across 504 test breast slices, while UNet averaged 24.55±6.64 dB and 0.0024±0.0031 respectively. Additionally, a masked loss calculation was incorporated to enhance the learning accuracy of the implemented models. This approach yielded an increase of 0.9 dB in average SNR for ResNet and a 0.8 dB SNR increase for the UNet architecture on average. These results indicate that the proposed DL models effectively reconstruct electric field distributions with substantial accuracy, highlighting its potential for practical applications in real-time microwave medical imaging and treatment planning. Further refinement and validation with a more diverse set of breast models could enhance the model's applicability and performance.

Hypoplastic Left Heart Syndrome (HLHS) is a severe and rare congenital heart disease characterized by the underdevelopment of the left side of the heart, which is responsible for pumping oxygenated blood throughout the body. This underdevelopment results in an insufficient supply of oxygenated blood to vital organs and tissues. Current treatments for HLHS face significant challenges and typically involve complex surgical interventions that can only extend patient survival to a limited degree. Heart transplantation is another potential option, but it is constrained by the scarcity of donors and the risk of immune rejection. Induced pluripotent stem cells (iPSCs), derived from somatic cells, possess the remarkable ability to differentiate into any cell type within the three primary germ layers, making them a promising tool for tissue regeneration, disease modeling, and repair. However, despite their potential, the clinical application of iPSCs remains limited, primarily due to the risk of tumorigenesis. This review explores the role of iPSCs in advancing the understanding and treatment of HLHS, examining their cellular and molecular mechanisms, historical development, and their application in cardiovascular research. Additionally, we consider the major challenges and limitations associated with the use of iPSCs in the context of HLHS, underscoring the critical need for further research to enhance their therapeutic efficacy and safety.