The escalation of ransomware attacks has become a critical threat to digital security, driven by increasingly adaptive and evasive malware that bypasses conventional detection methods. Addressing the limitations of signature-dependent and behavior-based approaches, this paper introduces a novel, autonomous framework known as Temporal Signature Decomposition Analysis (TSDA), designed to detect ransomware through the decomposition of unique temporal event patterns. TSDA operates independently of predefined threat signatures, leveraging timeseries analysis and statistical pattern recognition to identify ransomware behaviors based on time-dependent anomalies, thus achieving a high detection rate with minimal false positives. Through extensive experimental evaluation, TSDA demonstrates both high detection accuracy and low latency, enabling it to function effectively in environments requiring rapid and reliable ransomware identification. Moreover, the method's scalability and efficient resource utilization establish it as a viable solution for large-scale ransomware defense across diverse operational domains, reducing the need for human oversight while ensuring resilience against sophisticated attack tactics. The study's findings demonstrate TSDA's potential as a transformative tool in advancing ransomware detection, providing a sophisticated, adaptable, and robust addition to modern cybersecurity frameworks.

The recent surge in demand for contextually coherent and semantically accurate language model outputs demonstrates the persistent challenge of maintaining relevance across extended sequences, a task that traditional architectures struggle to accomplish effectively. Dynamic Semantic Drift Encoding (DSDE) introduces a novel framework designed to enhance contextual retention by dynamically adjusting semantic representations as text progresses, addressing limitations that often cause context degradation over long spans. Implementing DSDE within a state-of-the-art Large Language Model, this study demonstrates substantial improvements in coherence, relevance, and semantic alignment, achieving a level of consistency that surpasses baseline models. Experimental results reveal that DSDE not only effectively mitigates issues related to semantic drift and recency bias but also facilitates more natural long-range dependency modeling, proving beneficial for applications requiring a high degree of contextual accuracy. Moreover, findings indicate that DSDE contributes to a notable expansion in lexical diversity and adaptability across varied domains, confirming its versatility and practical significance in enhancing language model performance across diverse linguistic tasks. These results demonstrate DSDE's potential as an innovative and scalable approach to overcoming fundamental limitations in long-sequence contextual retention, marking a meaningful advancement in language model technology.

Apache Spark has revolutionized the landscape of big data processing by harnessing the power of distributed computing to handle massive datasets. However, as Spark applications increase in size and complexity, effective performance tuning becomes essential. Optimizing Spark jobs is crucial for maximizing resource utilization, accelerating job completion, and minimizing operational costs. This article explores the architecture overview of Hadoop, Apache Spark and critical aspects of performance tuning in Apache Spark, focusing on techniques and strategies for enhancing data processing, resource allocation, and job execution. By leveraging Spark's features and optimization tactics, users can significantly improve the performance of their applications, leading to more efficient and cost-effective big data solutions.

Math anxiety is a significant factor contributing to the gender gap in science, technology, engineering, and mathematics (STEM), particularly among female students, as it undermines their confidence and discourages them from pursuing careers in these fields. This anxiety, characterized by fear or discomfort when engaging with math-related tasks, often becomes a barrier for girls at a young age, leading to a disconnect from STEM subjects as they progress through their education. While societal and cultural factors, such as stereotypes and gender expectations, also play a role, addressing math anxiety presents a direct opportunity to reduce this gap. One promising solution is increasing the visibility of female role models in STEM to help mitigate the negative effects of math anxiety. By creating programs that intentionally connect young female students with successful women in STEM industries, this approach aims to provide inspiration and practical guidance. Hearing stories from women who have excelled in male-dominated fields helps demystify the path to STEM careers, offering mentorship and advice on overcoming challenges, including math anxiety. This early intervention can reshape how female students perceive their own capabilities in math, fostering confidence and a more positive attitude toward STEM. Through these structured connections, the study highlights the potential for reducing the gender gap by empowering girls at a younger age, ensuring they see STEM as a viable and accessible career path.

In this brief we prove that both forms of Middlebrook's formula for the transfer impedance T of a two-port can be seen as particular instances of the more general expression T = T* (p_nq_3 + q_np_3)/(p_d q_3 + q_d p_3) , which involves a generic reference immittance, not necessarily an open-circuit or a short-circuit. The p-and q-parameters arising above owe to the use of a homogeneous description of the extra element and of the double null and the single injection immittances. The formula holds under minimal regularity assumptions.

Current and upcoming data-intensive Mission Critical (MC) applications rely on high Quality of Service (QoS) requirements related to connectivity, latency and network reliability. Beyond 5G networks shall accommodate MC services that enable voice, data and video transfer in extreme circumstances, for instance in occurrence of network overloads or infrastructure failures. In this work, we describe the specifications of the architectural framework that enables the roll-out of MC services over 5G networks and beyond, considering recent technological advancements of cloud-native functionalities, network slicing and edge deployments. The network architecture and the deployment process is described in three practical scenarios, including a capacity increase in the service load that necessitates the scaling of the computational resources, the deployment of a dedicated network slice for accommodating the stringent requirement of a MC application and a service migration scenario at the edge to cope with critical failures and QoS degradation. Furthermore, we illustrate the implementation of a Machine Learning (ML) algorithm that is used for overload prediction, validating its ability to predict the capacity increase and notify the components responsible to trigger the appropriate actions, based on a real dataset. To this end, we mathematically define the overload detection problem, as well as generalized prediction tasks in emergency situations and examine the key parameters (proactiveness ability, loockback window, etc.) of the ML model, also comparing its predictions abilities (∼ 93% accuracy in overload detection) against multiple baseline classifiers. Finally, we demonstrate the flexibility of the ML model to achieve reliable predictions in scenarios with diverse requirements.

The energy consumed by buildings is expected to significantly rise in the upcoming years, necessitating intelligent Home Energy Management Systems (HEMS) that create comfortable conditions for their inhabitants, while also offering sustainable and cost-effective solutions. The building environment, however, includes multiple time-varying parameters that cannot be controlled, such as the output of renewable energy sources, the market-dependent electricity prices, the outdoor temperature, as well as the occupants' energy habits. To overcome these barriers, we propose a hybrid Machine Learning (ML) algorithm for smart HEMS control, leveraging the properties of a decision-making deep deterministic policy gradient model, enhanced by the predictive capabilities of long short-term memory networks. Hence, the proposed algorithm aims to achieve an optimal balance between energy cost and occupant comfort by continuously adjusting the energy provided to the heating, ventilation, and air conditioning system, as well as controlling the energy storage system of the smart home. The proposed hybrid method is validated with simulations using real-world data and compared against baseline approaches, showcasing its effectiveness to achieve an optimal trade-off between the indoor temperature deviation and the average energy cost.

To extend the flight duration of unmanned aerial vehicles (UAV), laser-powering systems have been proposed in recent times that can charge a flying UAV with laser beams from the ground. In this study, we analyze the potential of using a ground-based laser station to charge a fixed-wing UAV flying in a turbulent atmosphere. This presents a new design challenge in optimizing the UAV's trajectory by jointly considering atmospheric turbulence and UAV energy consumption. To overcome this, we derive theoretical models which capture the UAV velocity and position evolution along a circular trajectory in a turbulent atmosphere. Based on the derived models, we first compute the received energy loss at the UAV that depends on the UAV velocity (variance) as well as the trajectory radius. Then, based on the received energy loss and the energy consumption models of the UAV, the UAV trajectory is optimized with respect to velocity and trajectory radius to maximize the net laser power harvested by the UAV. Our results show that the optimal velocity and trajectory radius depend strongly on the levels of turbulence in the atmosphere.

This paper presents an optimization methodology for the design of square and rectangular substation grounding grids with equal conductor spacing, ensuring compliance with IEEE Std. 80-2013 safety criteria while minimizing construction costs to minimum possible limits. The proposed algorithm systematically performed parameters sweeps on all the design variables as stipulated in the standard, including grid burial depth, surface layer thickness, conductor material, conductor size, conductor spacing, and the number and length of ground rods. For optimization, two different cost functions were used: one based solely on the material costs of conductors, rods, while the other incorporated additional factors such as installation, and excavation costs. The algorithm’s performance is evaluated using eight different test grid systems published in the literature, and results were validated with the optimization results obtained from ETAP software. The results demonstrate significant cost reductions while maintaining safe step and touch potentials, offering a practical tool for designing safe, reliable, and cost-effective grounding grids for electrical substations.

Cloud-Edge Computing Continuum (CEC) system, where edge and cloud nodes are seamlessly connected, is dedicated to handle substantial computational loads offloaded by end-users. These tasks can suffer from delays or be dropped entirely when deadlines are missed, particularly under fluctuating network conditions and resource limitations. The CEC is coupled with the need for hybrid task offloading, where the task placement decisions concern whether the tasks are processed locally, offloaded vertically to the cloud, or horizontally to interconnected edge servers. In this paper, we present a distributed hybrid task offloading scheme (HOODIE) designed to jointly optimize the tasks latency and drop rate, under dynamic CEC traffic. HOODIE employs a model-free deep reinforcement learning (DRL) framework, where distributed DRL agents at each edge server autonomously determine offloading decisions without global task distribution awareness. To further enhance the system pro-activity and learning stability, we incorporate techniques such as Long Short-term Memory (LSTM), Dueling deep Q-networks (DQN), and double-DQN. Extensive simulation results demonstrate that HOODIE effectively reduces task drop rates and average task processing delays, outperforming several baseline methods under changing CEC settings and dynamic conditions.

The increasing prevalence and sophistication of cyber threats have introduced unique challenges in protecting critical systems and sensitive data from evolving ransomware attacks that adapt rapidly to evade conventional detection strategies. Addressing the need for more adaptive and resilient defenses, the Dynamic Threat Signal Clustering (DTSC) model offers an innovative methodology for detecting ransomware through the clustering of threat signals based on distinctive ransomware behaviors. By employing a real-time, high-dimensional clustering algorithm, DTSC can dynamically identify ransomware patterns amidst vast network traffic, demonstrating improved detection accuracy while significantly reducing false positive rates compared to traditional signature-based approaches. DTSC not only enhances operational efficiency through its refined clustering mechanism, which isolates malicious signals even within complex network environments, but also optimizes resource utilization, making it adaptable to both high-traffic enterprise systems and more resource-limited infrastructures. Experimental results indicate that DTSC's ability to continuously recalibrate in response to emerging threats enables robust adaptability, positioning it as a vital addition to the modern cybersecurity landscape. The implications of DTSC's success extend across cybersecurity practices, offering a scalable and highly effective tool for ransomware detection that is capable of safeguarding critical infrastructures against one of the most persistent and rapidly evolving cyber threats.

The expanding capabilities of language models have intensified the need for adaptive mechanisms that can manage dynamic contextual shifts in extended interactions. Current approaches primarily rely on static embeddings, which limit the model's responsiveness to real-time contextual variations, thereby hindering coherence in complex dialogues. The proposed Dynamic Language Interaction Mapping (DLIM) framework addresses this gap by introducing a novel methodology for continuous context-sensitive adaptation, enabling more fluid, relevant responses. Through a series of quantitative analyses, the DLIM-enhanced model exhibited substantial improvements across metrics such as contextual relevance, coherence over longer sequences, response diversity, and robustness to input noise, significantly outperforming baseline models. The architecture of DLIM not only advances the adaptability of language models but also broadens their functional applications, providing a scalable solution suitable for deployment in multiuser environments. These findings demonstrate the potential of DLIM to transform contextually adaptive language technologies, supporting a range of applications that require both high accuracy and flexibility in response generation.

Urban Air Mobility (UAM) is an emerging transport mode, offering on-demand and automated air transit of passengers or cargo within urban ecosystems. Unrestricted by terrestrial road networks, UAM provides a timely and efficient way of transportation. Due to its profound significance, industries, regulatory authorities, and academic researchers have made extensive efforts toward its development and improvement. This paper contributes to these ongoing endeavors by providing a comprehensive and systematic review of existing literature on UAM from the perspective of airspace and traffic management. Specifically, this paper chronologically encapsulates the development history of UAM, aiming to foster a holistic understanding of contributions made by industries, academics, and regulatory authorities. It also highlights specific advances such as the integration of AI in traffic flow algorithms and the development of collision avoidance systems, addressing current technological and regulatory hurdles. Furthermore, this paper provides an in-depth review of the studies on the topic of urban air transportation network design, urban air traffic flow management, and urban flight safety management. The diverse scenarios and methodologies employed across those studies are discussed and compared, identifying gaps in current research and potential for future exploration. Lastly, this paper highlights opportunities and prospective directions for future research in relation to traffic management strategies, thus propelling UAM toward greater efficiency. rban Air Mobility (UAM) is an emerging transport mode, offering on-demand and automated air transit of passengers or cargo within urban ecosystems. Unrestricted by terrestrial road networks, UAM provides a timely and efficient way of transportation. Due to its profound significance, industries, regulatory authorities, and academic researchers have made extensive efforts toward its development and improvement. This paper contributes to these ongoing endeavors by providing a comprehensive and systematic review of existing literature on UAM from the perspective of airspace and traffic management. Specifically, this paper chronologically encapsulates the development history of UAM, aiming to foster a holistic understanding of contributions made by industries, academics, and regulatory authorities. It also highlights specific advances such as the integration of AI in traffic flow algorithms and the development of collision avoidance systems, addressing current technological and regulatory hurdles. Furthermore, this paper provides an in-depth review of the studies on the topic of urban air transportation network design, urban air traffic flow management, and urban flight safety management. The diverse scenarios and methodologies employed across those studies are discussed and compared, identifying gaps in current research and potential for future exploration. Lastly, this paper highlights opportunities and prospective directions for future research in relation to traffic management strategies, thus propelling UAM toward greater efficiency.

The escalating sophistication of cyber threats necessitates innovative detection mechanisms to safeguard digital infrastructures. Dynamic Enclave Partitioning (DEP) emerges as a novel approach, offering real-time isolation and monitoring of suspicious code activities within virtualized environments. DEP's adaptive segmentation enables continuous adjustment of enclave boundaries based on observed behavioral patterns, effectively distinguishing between benign and malicious processes. This methodology enhances detection accuracy by minimizing false positives and provides a robust defense against evolving ransomware tactics. The modular design of DEP facilitates seamless integration with existing security frameworks, extending its applicability beyond ransomware to a broader spectrum of malware threats. Comprehensive evaluation demonstrates DEP's efficacy in identifying and isolating novel ransomware behaviors, showing its potential as a significant component in contemporary cybersecurity strategies.

Previous research on jailbreak attacks has mainly focused on optimizing the adversarial snippet content injected into input prompts to expose LLM security vulnerabilities. A significant portion of this research focuses on developing more complex, less readable adversarial snippets that can achieve higher attack success rates. In contrast to this trend, our research investigates the impact of the adversarial snippet's position on the effectiveness of jailbreak attacks. We find that placing a simple and readable adversarial snippet at the beginning of the output effectively exposes LLM safety vulnerabilities, leading to much higher attack success rates than the input suffix attack or promptbased output jailbreaks. Precisely speaking, we discover that directly enforcing the user's target embedded output prefix is an effective method to expose LLMs' safety vulnerabilities.

The escalating sophistication and frequency of ransomware attacks have highlighted the inadequacies of traditional detection methods, which often fail to adapt to evolving attack patterns and advanced evasion techniques. Addressing the pressing need for more resilient detection mechanisms, the proposed approach introduces a novel framework, RansomNet, that leverages deep behavioral signature extraction to identify ransomware through unique operational patterns. Unlike conventional approaches that rely on static signatures or limited heuristic analysis, RansomNet dynamically analyzes ransomware behavior, capturing complex, adaptive signatures that accurately distinguish between benign and malicious activities. Through employing advanced deep learning models, RansomNet achieves high detection accuracy across diverse ransomware families and maintains exceptionally low false positive rates, making it suitable for deployment in various operational environments. Extensive empirical evaluations validate the framework's adaptability and precision, demonstrating its capacity to maintain effectiveness against both known and emerging ransomware variants. Ransom-Net's innovative architecture not only advances the reliability of ransomware detection but also enhances cybersecurity resilience through its real-time, adaptive analysis of evolving ransomware behaviors.

The cylindrical screen has bee proposed and utilized in studying the double slit/grating experiments. In this article we derived the math formulas for describing the patterns on the cylindrical screen.

This paper comprehensively presents 105 GHz multipath characterizations for indoor short-range communication environments and proposes a stochastic channel generator compatible with the third-generation project partnership (3GPP) standard. Using extensive wideband propagation measurements, we holistically derive the statistical distributions of both largescale parameters (LSPs) and small-scale parameters for various indoor short-range communication environments, such as desktops in conference rooms, corridors, and office rooms. These distributions not only capture the holistic propagation characteristics of this underexplored frequency band in the aforementioned environments but also serve as a complete stochastic model sufficient for developing a multipath channel generator to perform physical layer link-level simulations. The derived parameters are compared with those specified in the incumbent 3GPP stochastic channel model for an indoor hotspot office scenario, highlighting the fact that the cross-correlation between the azimuth angle spread of arrival and the K-factor demonstrates a major difference, requiring model amendments for short-range use cases in this band. Based on these results, we propose a 3GPP-compatible channel generation algorithm tailored for all three indoor short-range communication scenarios at 105 GHz, incorporating the derived statistical distributions. The extensive simulations of channel generation demonstrated consistency with our propagation measurements in terms of intra-cluster subpath characteristics and LSPs, demonstrating the validity of the proposed channel generation algorithm. Our results offer a foundation for accurate link-level simulations in various 105 GHz short-range communication use cases, which is crucial for advancing next-generation wireless communication systems. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.