This paper introduces the JAM Stream cipher, an improved version of the classical Vigenère cipher, designed to address its known vulnerabilities. The proposed cipher technique incorporates a multi-key, multi-round, and chaining scheme to significantly enhance security. By leveraging these techniques, the JAM Stream cipher creates a strong Avalanche effect, ensuring substantial changes in ciphertext for minimal alterations in plaintext or keys. Experimental results confirm the cipher's resilience against differential attacks and brute-force methods, demonstrating a marked improvement over traditional algorithms, namely the Vigenère cipher and DES. However, the JAM Stream cipher faces computational complexity challenges for real-time and high-performance applications. To facilitate further research and practical implementation, the source code is available at https://github.com/alberttuwan1/JAM_Stream_Cipher. Future research will explore optimization strategies to balance security and efficiency, paving the way for broader adoption in secure communication systems.

Modern modular service ecosystems rely on microservices for scalability and flexibility, yet performance constraints such as bottlenecks and resource contention remain critical challenges. This research presents a novel insight-driven framework for detecting and mitigating performance anomalies in microservices-based applications. By leveraging graph-based learning models and multi-source telemetry analysis, our approach effectively identifies independent, dependent, and cascading bottlenecks, facilitating timely intervention. We introduce and evaluate GAMMA, an advanced anomaly detection and localization model that integrates attention-based graph convolution networks with a multi-expert learning mechanism. Empirical evaluations on large-scale datasets demonstrate that GAMMA significantly outperforms existing methods in detecting and diagnosing multiple bottlenecks, achieving an F1 score of 0.92 for anomaly detection and 0.89 for bottleneck localization. The proposed framework enhances performance management in modular service architectures, ensuring resilient and adaptive cloud-native applications.

Prompt engineering has arisen as a pivotal discipline in optimizing the performance of Large Language Models (LLMs) by structuring inputs to enhance coherence, accuracy, and task alignment. This paper comprehensively surveys various prompting techniques, systematically categorizing them according to their application domains and methodological foundations. Fundamental approaches like zero-shot and few-shot prompting are examined along with advanced strategies, including chain-of-thought reasoning, retrieval-augmented generation, and self-consistency mechanisms. A rigorous qualitative analysis is conducted to evaluate each technique's strengths, limitations, and optimal use cases, offering a structured framework for selecting the most effective prompting strategies. Theoretical insights and empirical findings are consolidated to provide researchers and practitioners with advanced methodologies for refining prompt design and enhancing LLM capabilities in complex reasoning, decision-making, and knowledge synthesis while improving reliability and factual accuracy in generated outputs.

Deepseek and Qwen large language models (LLM) became popular at the beginning of 2025 with their low-cost and open-access LLM solutions. A company based in Hangzhou, Zhejiang, China, announced its new LLM, DeepSeek v3, in December 2024. Then, Alibaba released its AI model, Qwen 2.5 Max, on January 29, 2025. These tools, which are free, open-source, and have no specified or known query limit, have made a significant impact on the world. Deepseek and Qwen also have the potential to be used by many researchers and individuals around the world in academic writing and content creation. Therefore, it is important to determine the capacity of these new LLMs to generate high-quality academic content. This study aims to evaluate the academic writing performance of both Qwen 2.5 Max and DeepSeek v3 by comparing these models with popular systems such as ChatGPT, Gemini, Llama, Mistral, and Gemma. In this research, 40 articles on the topics of Digital Twin and Healthcare were used. The method of this study involves using generative AI tools to generate texts based on posed questions and paraphrased abstracts of these 40 articles. Then, the generated texts were evaluated through the plagiarism tool, AI detection tools, word count comparisons, semantic similarity tools and readability assessments. It was observed that plagiarism test result rates were generally higher for the paraphrased abstract texts and lower for the answers generated to the questions, but both were above acceptable levels. In the evaluations made with the AI detection tool, it was determined with high accuracy that all the generated texts were detected as AI-generated. In terms of the generated word count comparison, it was evaluated that all chatbots generated satisfactory amount of content. Semantic similarity tests show that the generated texts have high semantic overlap with the original texts. The results of the readability tests of the generated texts showed that the texts were insufficient in terms of readability.  

The volume of healthcare-related data is increasing, and accurate, efficient information retrieval is crucial. Traditional lexical retrieval methods have limitations, often leading to suboptimal results due to their inability to capture semantic meaning. This exploratory study investigates the combination of Retrieval-Augmented Search (RAS) with Facebook AI Similarity Search (FAISS), with the aim of improving medical information retrieval, specifically in diabetes-related content. We compared the performance of 'bge-large-en-v1.5' BGE-large-en-v1.5 model (Big Graph Embeddings-Large English Version 1.5) and 'all-mpnet-base-v2', Allmpnet-base-v2 model (a small variant of MPNet) noting that 'all-mpnet-base-v2' achieved a superior similarity score of 0.5508, outperforming earlier models. Additionally, scaling the corpus by a factor of ten demonstrated that while embedding time. Accurate and efficient retrieval of information is essential. This study highlighted the scalability of FAISS for large-scale medical text retrieval.Based on these results, it appears that the hybrid retrieval approach, combining RAS and FAISS, is effective.The study demonstrates that FAISS, when combined with RAS, effectively enhances retrieval accuracy and scalability for large-scale healthcare applications.

This work presents a practical implementation that integrates three emerging wireless technologies: mmWave communications to access a broadband underutilized spectrum, non-orthogonal multiple access (NOMA) as the multiple access strategy, and intelligent reflective surfaces (IRS) as the relaying strategy. The goal is to provide multiuser mmWave indoor coverage in a specific indoor setting. Coverage in the mmWave band, often affected by blind spots, is extended by the ad-hoc design and deployment of a fully passive IRS that requires no power supply. Contrary to the use of a Reconfigurable Intelligent Surface (RIS), this solution lacks the dynamic reconfigurability of the relay but offers a low-cost, easy-to-deploy alternative with zero power consumption to provide coverage tailored to the architecture of the scenario. The integration with power-domain NOMA provides increased spectral efficiency by multiplexing two users in the same frequency, time and code slot. To the best of our knowledge, this is the first mmWave NOMA implementation and the first prototype that includes both NOMA and IRS. The radio frequency (RF) design ensures a bandwidth of 400 MHz centered at 24.45 GHz, and the communication link is evaluated through over-the-air (OTA) measurements of channel power, error vector magnitude (EVM) and symbol error rate (SER), considering different user positions and superposition coding strategies.

This paper introduces a multi-stage hybrid framework for fault detection, isolation, and identification in large-scale vehicle systems characterized by cascaded estimation structures. The hybrid framework combines model-based residual quantification with data-driven techniques, using derived mathematical equations to model both the temporal and spatial dependencies of residuals. This enhances fault isolation by accurately capturing fault propagation in complex systems. By integrating model-based likelihood quantification with deep learning models, including a recurrent neural network (RNN) layer, the approach improves recall while addressing the limitations of model inaccuracies, unmodeled dynamics, and unknown disturbances. High accuracy in fault detection and isolation is achieved by leveraging the strengths of both methods. Additionally, the multi-stage approach enhances the performance by separating the fault detection, isolation, and identification stages. The hybrid approach demonstrates improved fault isolation results, even with limited faulty data, by utilizing model-based priors that guide the deep learning algorithms. Precise identification of bias fault magnitudes is realized through model-based equations. Extensive numerical evaluations, , featuring a high-fidelity vehicle model from CarSim, validate the framework's effectiveness, demonstrating its capability to accurately detect, isolate, and identify faults in a complex cascaded vehicle estimation.

This work investigates the reliability of complementary field-effect transistors (CFETs) by addressing both design-time variability arising from process variations and run-time variability due to temperature and aging effects. A rigorously calibrated TCAD model, validated against experimental CFET data, is employed to quantify the impact of metal gate granularity (MGG)-induced work function variation (WFV) and random dopant fluctuation (RDF) on key device parameters, including the threshold voltage (V TH), on-state current (I ON), and off-state current (I OFF). Temperature-dependent variability is systematically analyzed to further elucidate the behavior of these advanced devices. To capture the dynamic effects of aging, the reaction-diffusion (RD) framework-which accounts for defect generation due to Negative Bias Temperature Instability (NBTI)-is implemented in TCAD, enabling detailed modeling of trap generation and the corresponding V TH shifts in p-type transistors under varying gate stress biases (V GSTR) and operating temperatures. At the circuit level, a full array of 6T-SRAM cells with the requisite peripheral circuits is simulated using SPICE after careful calibration of the the industry-standard compact model of gate-all-around (BSIM-CMG) against the TCAD data. Variability analysis reveals that the access disturb margin achieves a cell sigma (µ/σ) of 17.4 at nominal supply voltage, significantly exceeding the 6σ robustness criterion for read disturbances. Moreover, as the operating temperature increases from 300 K to 398 K, the read static noise margin (RSNM) and hold static noise margin (HSNM) degrade by 13.7% and 6.37%, respectively, while the write static noise margin (WSNM) improves by 18.3 %. These findings provide critical insights into the design trade-offs and reliability challenges of CFET-based SRAMs.

The Location Service ETSI standard has evolved to closing a critical gap between the standards for positioning mechanism of 5G mobile networks and the Multi-access Edge Computing (MEC) standard framework. Location Service (LS) generally allows, among other functions, to optimize services according to users' geographical location. Although LS already offers features typical of the next generation of mobile networks, such as 3D positioning of user equipment, the LS standardization framework is still evolving to meet the demands in support of next-generation mobile networks, most notably the sixth generation paradigm. Some of the features in the current standards, for instance, the concepts of zones, carry the promise of possibly transformative LS applications in future mobile networks. For instance, LS can help the building of a "6G network of (sub-)networks", with various technology solutions, towards the fully decentralized network operation. These and the related developments are driving further evolution of MEC and LS standards to support such new features. This article addresses the evolution of current Location API standards towards 6G mobile networks. We propose extensions and improvements to the current Location API data model to include data fields to characterize novel features and the envisioned novel use cases in the next-generation mobile networks. We show that the proposed modifications allow us to abstract the user's location into a tuple of physical and network locations, allowing MEC applications to perform complex data flow optimization.

A sparse-array metasurface (MTS) antenna corresponds to a continuously modulated sheet impedance on top of a grounded slab, periodically repeated over many unit cells (with a few wavelengths cell size). The structure is fed with an array of surface-wave (SW) feeds placed at the center of each cell. The modulated MTS is designed to generate a sharp embedded element radiation pattern (EEP) able to filter out possible grating lobes that may appear due to the large size of the unit cell (as compared to the classically adopted half-wavelength size in traditional phased arrays). This paper proposes a Method of Moments (MoM) tool developed to efficiently analyze modulated MTSs with periodic boundary conditions (PBC). Such structures may be used to provide continuous beam scanning with a sparse feeding network. In this letter, the unknown current distribution as well as the sheet impedance are expanded into a set of regularly spaced Gaussian basis functions. The adopted Gaussian functions along with the mesh regularity are employed to dramatically accelerate the computation of the MoM matrix with FFT-based accelerations. The numerical results are validated with that from the commercial software HFSS.

Lattice-based Post-quantum cryptography and Homomorphic Encryption schemes have become the key methodologies for today's and the future's secure world. This comes at the cost of a vastly increased computational load due to the multiplication of wide-integer coefficient polynomials. NIST recommends Number Theoretic Transform (NTT) as an efficient remedy. Nevertheless, NTT strongly requires acceleration for large numbers of coefficients. This paper explores the use of systolic arrays as NTT accelerators and finds an optimal hardware architecture configuration across problem sizes. Design-space exploration is performed, resulting in a new design configuration for an efficient two-dimensional NTT accelerator without losing the ability to execute other workloads. Our finding indicates that for 22 nm technology, an optimal systolic array accelerator requires an area of 53.04 mm 2. The accelerator can efficiently execute and apply NTT on a polynomial with 4096 32-bit integer coefficients requiring 3296 cycles, and 1794.92 nJ.

Ultra-dense 3D ICs, with ultra-dense 3D connections (pitch < 100 nm), are projected to achieve large energy and throughput benefits compared to today's ICs. To enable many high-power compute engines on 3D tiers, 3D thermal and power delivery challenges must be overcome. We use a recent idea called 3D thermal scaffolding to overcome both challenges simultaneously. 3D thermal scaffolding cools ultra-dense 3D ICs (e.g., monolithic 3D ICs) using a combination of (1) a new thermally conductive dielectric (the 'thermal dielectric'), (2) scaffolding vias for heat conduction paths to the heat sink, and (3) efficient (and experimentally demonstrated) heatsinks. In this paper, we present new algorithms which place scaffolding vias and the thermal dielectric with minimal footprint impact while satisfying peak temperature and worst-case IR drop constraints. These algorithms are implemented during physical design and demonstrated using a 12-tier open-source 7nm AI accelerator design. Compared to approaches that do not consider power delivery and thermal constraints simultaneously, our approach reduces the footprint penalty due to 3D thermal scaffolding from 10% to 5.5%-an 80% improvement-and simultaneously meets worst-case IR drop constraint of <20 mV at 0.7V supply and peak temperature constraint of <125℃.

Large urban areas generate vast amounts of data that pose challenges inaccessibility and understanding. To address these challenges, our study introduces an innovative framework for Interactive Data Exploration within Smart City Analytics, centering around user needs and experiences. By leveraging intuitive visualization techniques, we allow users to dynamically engage with urban data, facilitating better comprehension and interaction. The framework incorporates user feedback to refine the exploratory experience, thereby enhancing accessibility and usability. We conducted user studies to evaluate its effectiveness, which demonstrated marked improvements in user satisfaction and understanding when compared to conventional static visualizations. The approach not only enriches users' grasp of smart city dynamics but also fosters community engagement in urban planning initiatives. Our results underscore the crucial role of user-centered design in transforming complex data into accessible insights for stakeholders involved in urban development.

In intersection scenarios, a single intelligent connected vehicle (ICV) is limited in making rational decision in a system due to the lack of cooperation with other vehicles and infrastructures, preventing it from acquiring comprehensive information from complex traffic environments and achieving global optimization. Under this circumstance, irrational decisions may lead to unnecessary energy consumption. To overcome this challenge, this paper proposes a novel swarm path decisionmaking method that can enhance the efficiency of ICV swarms' decision-making at consecutive signalized intersections, ensure travel efficiency and safety, and reduce the energy consumption. Firstly, consecutive intersections are modeled to obtain the information of topological nodes and lane-level waypoint. Then, a multi-objective function is formulated by integrating safety distance constraints, kinematic constraints, and risk constraints. After that, a novel adaptive co-evolutionary algorithm (ACA) is developed to solve this multi-objective function by incorporating a diversity-based search allocation mechanism (DSAM) and traffic volume-guided mutations (TVGM) to dynamically balance global exploration and local optimization, while adapting to varying traffic conditions. Finally, a hardware-in-the-loop test platform is built based on Prescan, MATLAB/Simulink, a driving simulator, and a V2X channel emulator to validate the effectiveness of the proposed method. The experimental results demonstrate that the proposed method achieves a 10.42% improvement in traffic efficiency and 42.95% reduction in energy consumption at consecutive signalized intersections while maintaining swarm safety under varying traffic volumes.

The research looks at user impressions of large language model (LLM) DeepSeek through a dual analytical framework encompassing sentiment and thematic analysis. The study aims to find user responses' predominant moods and repeated themes. The bifocal approach enhances perceptions of user satisfaction, concerns, and expectations by including sentiments and themes. A mixed-methods approach combining rule-based VADER sentiment analysis and thematic evaluation was used to analyze user feedback to elucidate platform-specific advantages and systemic challenges. Results indicate DeepSeek was perceived as highly positive overall by users (+0.80 aggregate sentiment) as well as high scores for Accessibility (+0.93) and Intelligence & Reasoning (+0.88). Nevertheless, emerging criticism revolved around its Censorship and Content policies (-0.20). The users applauded analytical accuracy through free access but criticized the interface navigation issues. Thematic analysis pinpoints DeepSeek as affluent with analytically driven users while emphasizing its highly tenacious capacity towards domain-specific problems. However, Cross-cutting challenges include latency, stability, and the inherent tension between content moderation and user autonomy. The AI evaluation frameworks are methodologically advanced by adding computational sentiment tools and qualitative perceptions. The intuitions from these findings allow for a better understanding of the user experience and have profound implications for the current development and practical usability of DeepSeek. The relationship between functional specialty and user expectations equates the competitiveness of AI platforms with DeepSeek, seen as a robust, free analytical tool while signaling possible areas for strategic enhancements.

This paper analyzes the outage probability and ergodic capacity of a RIS-assisted fluid antenna system (FAS)enabled non-orthogonal multiple access (NOMA)-based satellite communication system, considering practical impairments such as imperfect successive interference cancellation (SIC) and channel state information (CSI). We model the satellite links with shadowed-rician distribution and the terrestrial links with Nakagami-m distribution while employing maximum ratio transmission (MRT) at the LEO satellite. By leveraging the central limit theorem (CLT) and the moment method, the cascaded RIS-assisted channel is approximated as a Gamma distribution. Additionally, the clayton copula effectively models the correlation between FAS ports, enabling the derivation of approximate outage probability and ergodic capacity expressions for arbitrary fading distributions. The Gauss-Laguerre quadrature approximation is employed to efficiently solve complex integrals in the ergodic capacity analysis. Furthermore, asymptotic outage probability expressions in the high SNR regime are derived, revealing the diversity order and the impact of various system parameters on outage probability. Numerical results highlight the detrimental effects of these imperfections, underscoring their significance in system design. Performance improvements are demonstrated with an increasing number of RIS elements, transmit antennas, and FAS ports in RIS-assisted NOMA satellite communications. Furthermore, the impact of power allocation on outage probability is analyzed, and a trade-off between the number of FAS ports and the width of the fluid channel is established to achieve optimal performance.

Process models constitute essential tools in business process management and software engineering for representing and managing real-world business processes. Hence, the proper comprehension of these models is crucial for enabling an effective and efficient communication among stakeholders. While several studies have examined factors affecting process model comprehension, such as the used modeling notation or process complexity, there is a lack of in-depth research on the cognitive processes important to comprehend process models deeper. This systematic literature review explores the cognitive mechanisms underlying process model comprehension by integrating insights from relevant disciplines such as cognitive neuroscience and psychology. Key areas of cognition include perception, attention, memory, language, problem solving, emotion, and metacognition. The review was conducted following the well-established Kitchenham methodology and included an extensive search in the following digital libraries (Web of Science, IEEE Xplore, ScienceDirect, ACM Digital Library, PubMed, and Springer-Link). By analyzing 47 studies, the literature review identifies gaps in current research, emphasizing the need for further investigation into these cognitive processes to improve model comprehensibility.

This paper presents a study on the infield self-calibration of two rigidly connected IMUs' intrinsic parameters, without the aid of any external sensors, equipment, or specialized procedures. Specifically, we consider the calibration of gyroscope biases, gyroscope scale factors, and accelerometer biases, using only IMU data and known extrinsics between the two IMUs. We focus on the observability analysis of this system, and show that all gyroscope intrinsic parameters and a portion of accelerometer biases are observable, with information from both IMUs and sufficient motion. Moreover, we identify the additional unobservable directions in the intrinsic parameters that arise from various degenerate motions. Finally, we validate our observability findings through numerical simulations, and assess our system's calibration accuracy using real-world data.