Optimizing the performance of large language models (LLMs) on large-scale AI training and inference systems requires a scalable and expressive mechanism to model distributed workload execution. Such modeling is essential for pre-deployment system-level optimizations (e.g., parallelization strategies) and hardware design-space explorations. While recent efforts have proposed collecting execution traces from real systems, access to large-scale infrastructure remains limited to major cloud providers. Moreover, traces capturing execution on a specific platform cannot be easily adapted to study alternate software and/or hardware configurations, especially at scale. We introduce STAGE1, a framework that synthesizes high-fidelity execution graphs to accurately model distributed AI workloads (including LLMs and MoEs). STAGE supports a comprehensive set of parallelization strategies, allowing users to systematically explore a wide spectrum of model architectures and system configurations. STAGE demonstrates its scalability by synthesizing high-fidelity LLM traces spanning over 128K GPUs, while preserving tensor-level accuracy in compute, memory, and communication.
This paper introduces a novel l-shaped slots self-nonuplexing antenna in a nested substrate integrated waveguide cavity for LTE 5537.5/WiFi/WiMAX (5.21-5.50 GHz), WLAN/ISM (5.71-5.89 GHz), satellite communication (6-6.27 GHz), super ext. C-band (6.5-6.74 GHz), space communication (6.91-7.23 GHz), LTE/LTE advanced (7.45-7.69 GHz), meteorological satellite (7.90-8.12 GHz), ITU (8.21-8.79 GHz), and weather RADAR (8.75-9.33 GHz) applications. The antenna comprises eight different l-shaped slots in four subcavities and one square ring slot in the nested cavity. The inset feed of the 50 line has been used to excite the l-shaped slots, and the ring slot is excited by a coaxial feed line. Each slot radiates at nine different frequencies, such as 5.35, 5.8, 6.15, 6.6, 7.1, 7.65, 8.05, 8.65, and 9.2 GHz, respectively. The corresponding gains are 5.56, 5.84, 5.82, 6.32, 6.61, 7.50, 6.67, 6.77, and 7.08 dBi at the respective resonant frequencies. This antenna offers better than 20 dB measured isolation between any two ports and independent tenability features across any band. This antenna is appropriate for high-integration portable devices because of its planar design and its ability to eliminate intermodal electromagnetic interference. This antenna ends the requirement of RF switches in multiband transceivers and Internet of Things-based applications.
Smart electronic devices are emerging as a foundational component of next-generation technological systems, enabling intelligent sensing, communication, computation, automation, and real-time decision-making across diverse application environments. The rapid convergence of advanced semiconductor technologies, embedded systems, artificial intelligence, Internet of Things (IoT), flexible electronics, wireless communication, and energy-efficient circuit architectures has transformed conventional electronic devices into highly connected and adaptive platforms. This research paper examines the evolving landscape of smart electronic devices and their potential to support future applications in healthcare, smart homes, intelligent transportation, industrial automation, agriculture, environmental monitoring, wearable technology, and infrastructure management. Particular attention is given to devices equipped with integrated sensors, low-power processors, wireless connectivity, and intelligent data-processing capabilities that allow them to respond dynamically to changing operating conditions. The study explores how miniaturization, improved materials, edge computing, energy harvesting, flexible substrates, and advanced fabrication techniques contribute to enhanced device functionality, portability, reliability, and sustainability. Smart electronic devices can collect physical and environmental information, process data locally or through connected platforms, and generate actionable responses with reduced dependence on conventional computing infrastructure. Their integration with artificial intelligence and machine learning further enables predictive analysis, personalized services, anomaly detection, and autonomous operation. At the same time, the widespread deployment of such devices introduces challenges associated with power consumption, data security, privacy, interoperability, hardware reliability, scalability, electronic waste, and ethical use of intelligent technologies. The paper therefore considers both the technological opportunities and practical limitations associated with developing and deploying next-generation smart electronic systems. Emphasis is placed on the need for multidisciplinary design approaches that combine electronics engineering, computing, communication technologies, materials science, and application-specific requirements. The research highlights that future smart electronic devices will increasingly depend on seamless integration between sensing, processing, communication, and intelligent control while maintaining compact form factors and efficient resource utilization. By examining these developments collectively, the study provides a broad perspective on how smart electronics can contribute to more responsive, connected, sustainable, and human-centered technological ecosystems and identifies important directions for continued research and innovation in next-generation electronic applications.
cancer remains one of the leading causes of cancer-related mortality worldwide, and early diagnosis is essential for improving treatment outcomes and patient survival. Computed tomography (CT) plays a vital role in the detection and characterization of pulmonary abnormalities; however, the increasing volume and complexity of imaging data can create substantial challenges for accurate and timely interpretation. Artificial intelligence (AI), particularly deep learning, has emerged as a promising approach for automated medical image analysis and precision diagnosis. This study presents an AI-enabled precision CT imaging framework for the early detection and classification of lung cancer. The proposed framework incorporates image pre-processing, lung-region segmentation, region-of-interest extraction, deep feature learning, and automated classification to identify suspicious pulmonary abnormalities. CT images are initially enhanced through noise reduction, contrast improvement, and intensity normalization to improve image quality and consistency. Lung regions are subsequently isolated using segmentation and morphological operations, followed by convolutional neural network-based feature extraction and classification. The proposed model achieved an accuracy of 95.2%, sensitivity of 94.1%, specificity of 96.0%, precision of 94.8%, F1-score of 94.4%, and ROC-AUC of 0.97. The findings demonstrate the potential of AI-enabled CT imaging to support accurate and efficient identification of lung cancer patterns. By integrating automated image analysis with clinical expertise, the proposed framework may contribute to earlier diagnosis, improved diagnostic consistency, and precision-oriented lung cancer management. Further validation using large and diverse clinical datasets is required before routine clinical implementation.
Dynamic pricing is commonly used to regulate congestion in shared service systems. This paper is motivated by the fact that when heterogeneaous user groups (in terms of price responsiveness) are present, conventional monotonic pricing can lead to unfair outcomes by disproportionately excluding price-elastic users, particularly under high or uncertain demand. The paper's contributions are twofold. First, we show that when fairness is imposed as a hard state constraint, the optimal (revenue maximizing) pricing policy is generally non-monotonic in demand. This structural result departs fundamentally from standard surge pricing rules and reveals that price reduction under heavy load may be necessary to maintain equitable access. Second, we address the problem that price elasticity among heterogeneous users is unobservable. To solve it, we develop a robust dynamic pricing and admission control framework that enforces resource capacity and fairness constraints for all user type distributions consistent with aggregate measurements. By integrating integral High Order Control Barrier Functions (iHOCBFs) with a worst case robust optimization framework, we obtain a controller that guarantees forward invariance of safety and fairness constraints while optimizing revenue. Numerical experiments demonstrate improved fairness and revenue performance relative to monotonic surge pricing policies.
