The noisy shuffling channel models the conditions encountered in DNA storage systems, where transmitted data segments experience random permutation and substitution errors. Reliable communication over this channel requires effective indexing and channel coding strategies for segment order restoration and error correction. This paper introduces a concatenated coding approach for communication over the noisy shuffling channel, using Reed-Solomon (RS) codes as outer codes and polar codes as inner codes. A coset-based indexing method, derived from polar codes, is proposed. A joint decoder is designed to detect the permutation pattern and perform polar decoding simultaneously. An approximate analysis of the frame error rate (FER) using random coding is conducted. Additionally, a mapping between the cosets of the polar code and subsets of its frozen bits is established to design cosets achieving lower FERs compared to a commonly used explicit indexing method. Furthermore, a list decoding approach is devised, providing a trade-off between the computational complexity of the joint decoder and its performance.

This paper introduces a dataset characterizing the parameters of a stepped piezoelectric cantilever beam using various geometric specifications. The dataset also shows the electric potential generated at each layer when exposed to a specific excitation frequency. The primary application of the cantilever is in ambient vibration energy harvesting, where the piezoelectric cantilever beam plays a crucial role in converting mechanical vibrations into electrical energy. By shedding light on the complex interaction between geometric features and electric potential distributions, this dataset is valuable for advancing knowledge in energy harvesting technologies. The availability of this dataset facilitates further analysis, model validation, and optimization efforts, contributing to the ongoing enhancement of ambient energy harvesting systems by optimizing geometries.

Assessing the severity of gait impairment in Parkinson’s disease (PD) using the Movement Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) is typically performed by clinical experts, but this process is time-consuming, subjective, and costly. To address these challenges, we propose a Guided Diffusion Model with an encoder-only transformer that automatically predicts gait severity by learning the underlying distribution of PD gait and leveraging domain knowledge critical for clinical evaluations. Our diffusion model enables us to generate synthetic PD gait video frames conditioned on clinical features determined by experts to assess disease severity. These synthetic samples contain novel movement patterns not present in the observed data; systems trained on this information have better prediction performance. In addition, we propose a novel classification algorithm that can learn a predictive model, from both observed training data and synthetic samples, to accurately assess PD severity. We evaluate the effectiveness of the proposed method using two human motion datasets across two tasks: PD severity prediction and action classification. Our approach in predicting PD, and our action classification is sufficiently accurate that it can be applied to general applications with healthy subjects performing similar tasks. The full codebase is available on GitHub: https://github.com/arshakRz/DiffuseGaitNet.

Solid-state batteries (SSBs) represent a transformative leap in energy storage, offering significant advantages in safety, energy density, and environmental impact compared to conventional lithium-ion batteries (LIBs). However, the high costs and complex manufacturing processes of SSBs hinder their large-scale adoption. This research evaluates cost-reduction strategies and efficiency enhancements in the production of SSBs, focusing on roll-to-roll (R2R) processing. Key innovations include the integration of infrared (IR) drying systems and automated start/stop control mechanisms, combined with iterative machine calibration for quality control. The results demonstrate energy savings of ~50%, processing speed improvements of 20-50%, and manufacturing cost reductions of ~20%. These findings contribute to the broader adoption of SSB technology, paving the way for a sustainable energy future.

A document by Yuchen Yang. Click on the document to view its contents.

In the rapidly evolving field of multimedia services, video streaming has become increasingly prevalent, demanding innovative solutions to enhance user experience and system efficiency. This paper introduces a novel approach that integrates user digital twins-a dynamic digital representation of a user's preferences and behaviors-with traditional video streaming systems. We explore the potential of this integration to dynamically adjust video preferences and optimize transcoding processes according to real-time data. The methodology leverages advanced machine learning algorithms to continuously update the user's digital twin, which in turn informs the transcoding service to adapt video parameters for optimal quality and minimal buffering. Experimental results show that our approach not only improves the personalization of content delivery but also significantly enhances the overall efficiency of video streaming services by reducing bandwidth usage and improving video playback quality. The implications of such advancements suggest a shift towards more adaptive, user-centric multimedia services, potentially transforming how video content is consumed and delivered.

Scalability, heterogeneity, energy efficiency, costeffectiveness, robustness, interoperability, and low latency data transfer are some of the critical challenges posed by the Internet of Things in the modern era of the Internet. Content Centric Networks (CCN) and Named Data Networks (NDN) are some proposed solutions that can meet the abovementioned challenges. In-network caching, multicasting, content security, and decoupling of data from location are the significant advantages offered by the CCN. Despite the diverse benefits of CCN, it has to face some significant problems, such as how the key management and distribution are performed among the consumers and key managers efficiently, how the nodeto-node communication occurs, and how to provide services to the neighboring nodes while saving resources, etc. This paper proposes an Elliptic Curve Cryptography-based secure communication protocol supporting broadcast/multicast IoT data. Push-based data generation and transfer, typical requirements of IoT devices, have been integrated with the proposed protocol. The proposed communication protocol has been compared with existing protocols regarding computation and communication overheads as well as functional and security features. The proposed protocol has been evaluated in AVISPA and NS3 simulators for security and efficiency, and it is found that the proposed protocol outperforms the existing protocols in all aspects.

This research explores the application of possibility theory in the stock market, aiming to improve trading outcomes by integrating various technical indicators. The study focuses on the Relative Strength Index (RSI), Moving Average Convergence/Divergence (MACD), and Moving Average Crossover (MAC) to generate actionable buy, sell, or hold signals. By using possibility theory to combine these indicators, the research seeks to provide a more accurate assessment than individual indicators alone. Results indicate that the possibility theory-based approach significantly enhances returns while mitigating risk compared to traditional methods.

In order to improve the efficiency of automated warehouse robots in picking and packaging operations and enhance the control precision of their motion execution, this study incorporated machine learning algorithms into the design of the control system of automated warehouse robots. Mechanical learning and inductive learning were used to design machine learning methods applicable to these robots. Meanwhile, the learning process was optimized with the help of PID feedback adjustment, thus constructing an efficient execution system for the picking and packaging operations of automated warehouse robots. Taking the quantity of goods picked and packaged within the same period of time as the task indicator for the operations, simulation was carried out on the operation situations of robots adopting different machine learning methods. The results show that by using the combination of mechanical and inductive learning methods and the PID feedback adjustment approach, automated warehouse robots can handle more goods within the same period of time, significantly improving their motion execution efficiency in picking and packaging operations and providing strong support for the efficient operation of automated warehouses.

This paper presents a broadband low-loss fanin wafer level ball grid array (WLB) vertical through-moldvia (TMV) interconnect that enables highly integrated systemin-package (SiP) applications beyond 220 GHz. The dedicate advantage of the proposed approach is to further minimize the separation between package components (e.g., antenna in package (AiP)) and on-chip receiver or transmitter chain. To demonstrate the robustness of the TMV interconnect design, the necessity for co-simulation of the integrated circuit (IC) package and the IC itself during the package-design is shown. Therefore, an in-depth analysis of the suppression of parasitic mode oscillation inside silicon is carried out. Furthermore, the parasitic radiation loss at the TMV interconnect discontinuity is given and a verification of the proposed TMV due to high agreement of simulation and measurement results is demonstrated. The interconnect has been verified with a measured 10 dB return loss bandwidth of 38 GHz and a de-embedded insertion loss of less than 2.8 dB at 275 GHz, which enables a compact low-loss broadband signal transition from active IC components in the backend-of-line (BEOL) to passive components in package. Compared to fan-out WLB, the developed fanin WLB solution enables higher integration density of active and passive components with the shortest interconnects for minimal insertion loss. Thus, the proposed fan-in TMV packaging provides a very compact, easy-to-assemble and cost-efficient alternative for SiP designs for future broadband communication and sensing solutions.

The paper presents relatively simple formulations of the problem of acoustic scattering by elastic shells immersed in fluids, which can serve as a basis for efficient numerical models. The full rigorous formulation of the problem, which involves the Helmholtz equations for acoustic pressures in the fluids and the Navier equation for 3D displacements in the elastic material, is reduced to a boundary value problem only for the Helmholtz equations with effective boundary conditions relating the boundary pressures and normal displacements on both sides of the shell. To that end, the thin elastic shell is regarded as a neighborhood of its midsurface and the boundary values of the elastic quantities (displacements and stresses) are expressed via their values at the midsurface. The shell thickness is considered a small parameter. Special consideration is given to the cases of flooded and hollow shells. Under certain conditions, the problem is reduced to a system of hypersingular integral equations, appropriate for solving by the Boundary Element Method (BEM). The resulting numerical models are validated by the comparison with the exact solutions for the case of spherical elastic shells. In particular, the BEM numerical solutions reproduce the resonant peaks and dips of the exact solutions.

Sequential Model-Based Optimization (SMBO) is a highly effective strategy for hyperparameter search in machine learning. It utilizes a surrogate model that fits previous trials and approximates hyperparameters' response surface (performance). This surrogate model primarily guides the decision-making process for selecting the next set of hyperparameters. Existing classic surrogates, such as Gaussian processes and random forests, focus solely on the current task of interest and lack the capability to incorporate trials from historical tasks. This limitation hinders their efficacy in various applications. Inspired by the state-of-the-art convolutional neural process, this paper proposes a novel meta-learning-based surrogate model for efficient and effective hyperparameter optimization. Our surrogate is trained on the meta-knowledge from a range of historical tasks, enabling it to accurately predict the hyperparameter response surface even with a limited number of trials on a new task. We tested our approach on the hyperparameter selection problem for the well-known support vector machine (SVM) and residual neural network (ResNet) across hundreds of real-world classification datasets. The empirical results demonstrate its superiority over existing surrogate models, highlighting the effectiveness of metalearning in hyperparameter optimization.

Ship seaworthiness is a critical factor in maritime engineering, ensuring vessels can navigate safely and efficiently across diverse environmental conditions. This project employs advanced Computational Fluid Dynamics (CFD) calculations, coupled with GPU acceleration, to comprehensively assess ship seaworthiness. Through meticulous CFD simulations, the complex fluid-structure interactions between ships and water bodies are modeled, enabling precise predictions of hydrodynamic forces, fluid flow patterns, and stability parameters crucial for ship performance evaluation. The primary objective is to deepen understanding of ship behavior under various environmental scenarios, including different sea states, vessel speeds, and hull designs. By conducting detailed CFD simulations accelerated by GPUs, potential areas for enhancing ship designs, propulsion systems, and operational strategies can be identified, significantly improving seaworthiness and operational efficiency. Additionally, this approach aims to establish a robust framework for continuous improvement in maritime engineering practices, contributing to safer and more efficient maritime operations globally. The integration of advanced turbulence models and machine learning techniques further enhances the predictive capabilities of our simulations, offering new insights into optimal ship design and performance under challenging conditions.

Data centers increasingly depend on Operational Data Analytics (ODA) for real-time insights from vast streams of telemetry data. They typically utilize NoSQL databases for scalability and data diversity, which leads to unstructured data representation and presents significant challenges for the data interoperability. Indeed, the lack of standardization, combined with schema flexibility and complex data structures, makes it difficult for system administrators to write and execute queries, ultimately complicating the automation of data retrieval tasks. Pre-trained Large Language Models (LLMs), with their latent knowledge, promise a ready-to-use AI-driven data interoperability layer, enabling data retrieval through natural language input. However, they often generate inaccurate or hallucinated query code when handling heterogeneous data sources and complex data structures. This manuscript introduces EXASAGE, the first ODA co-pilot to leverage a Knowledge Graph (KG)-based approach, addressing these LLM limitations and simplifying data retrieval tasks in data center facilities through a prototype implementation of a conversational LLM agent. EXASAGE employs an LLM Agent as an interoperable layer to convert natural language into SPARQL queries (native to KGs), executed at a graph database endpoint. In evaluations on 1,000 prompts, EXASAGE achieved a 92.5% accuracy rate in generating correct SPARQL code, significantly outperforming the 25% accuracy of NoSQL/SQLite queries, which frequently exhibited hallucinations. Additionally, SPARQL queries are more concise, execute faster, and have shorter inference time compared to NoSQL/SQLite queries. The average time overhead for EXASAGE is about 20.36 seconds per prompt, covering the entire process, including the generation of the KG, LLM inference time, and query execution time. The maximum storage overhead for the generated KGs is just 52.62 MiB.

Higher-order boundary conditions are a useful tool for simplification of complex electromagnetic problems, such as spatially dispersive metasurfaces. A technique known as the Extended Generalized Sheet Transition Conditions (GSTCs) method has recently been proposed that allows for the systematic simulation of spatially dispersive metasurfaces using higherorder boundary conditions. A general metasurface, in general requires a boundary condition of infinite order spatial derivatives to accurately model the surface, making them unsuitable for conventional full-wave modeling techniques where higherorder derivatives are computed numerically. Here we propose a mixed spatial-spectral domain representation of the boundary conditions that avoids the explicit computation of higher-order derivatives while still being in a form suitable for integration into standard numerical methods. We integrate these boundary conditions into an integral equation solver which is then successfully validated through several numerical examples.

Aspect-based summarization has become increasingly important in scenarios where specific information needs to be extracted and condensed from vast amounts of text data, such as in e-commerce, healthcare, and finance. Introducing a novel approach that fine-tunes the Mistral Large model to target aspect-specific content, this research offers significant advancements in generating high-quality summaries that are both contextually accurate and domain-specific. The methodology involved a rigorous process of data collection, preprocessing, and aspect annotation, which collectively enhanced the model's ability to handle diverse linguistic structures and deliver precise summarization results. Extensive experiments conducted across multiple domains demonstrated the model's superior performance in key metrics such as precision, recall, and F1-score, highlighting its adaptability and effectiveness. The findings not only validate the robustness of the fine-tuned model but also reveal the potential for further optimization, particularly in complex domains. Through this research, a foundation has been laid for future innovations in generating aspect-based summaries, offering valuable insights for developing more targeted and sophisticated applications.

We propose a model for the atom which would explain the weak and strong nuclear forces as well as the electromagnetic force. We then demonstrate that General Relativity with some minor changes would constitute a 'theory of everything' with this new atom model.

Geometric problem-solving remains a challenging area for artificial intelligence due to the necessity for precise rule application and spatial reasoning. A novel approach is introduced in this research that incorporates rule-based alignment within the architecture of an open-source language model, Llama, to enhance its geometric reasoning capabilities. Through the embedding of explicit geometric rules into the model's neural network, the modified Llama demonstrates improved accuracy and efficiency in solving a wide range of geometric problems, from basic shape recognition to complex theorem application. The study employs a geometry-focused curriculum for training, which progressively increases in complexity, enabling the model to develop a robust understanding of geometric principles. Experimental results, compared with a baseline model, reveal significant improvements in problem-solving accuracy, consistency, and adherence to geometric rules, highlighting the efficacy of the rule-based alignment strategy. The findings suggest that integrating structured knowledge into language models can lead to substantial advancements in their ability to perform specialized mathematical tasks, thereby broadening the scope of applications for artificial intelligence in scientific and technical domains.