Sustainable Engineering and Innovation (SEI - E-Journal)
Not a member yet
142 research outputs found
Sort by
A scalable and explainable framework for detecting Ponzi schemes in Ethereum smart contracts using a stacking model
Full text linkBlockchain technology has reshaped digital finance, enabling decentralized applications (DApps) on platforms like Ethereum. However, these innovations have also facilitated fraudulent schemes such as Ponzi schemes, which deceive users with false promises of high returns. These schemes cause financial losses and weaken trust in blockchain systems. Existing detection methods face key challenges, including limited labeled data, over-reliance on transaction history, and failure to identify scams early. To address these issues, we propose a framework that combines static and dynamic features of smart contracts for early Ponzi detection. Our feature set includes opcode patterns, developer behavior, temporal trends, and metadata, crafted to work independently of transaction data. We enhance feature representation using TF-IDF, CountVectorizer, and Word2Vec for deeper semantic understanding. These features are used to train multiple machine learning and deep learning models such as Random Forest, XGBoost, CNNs, and BiGRUs. A stacking ensemble with a neural meta-learner integrates predictions for improved performance. The model achieves 99% accuracy and an AUC of 0.9522 on a curated Ethereum dataset, handling class imbalance through oversampling and synthetic data generation. We also employ SHAP for model explainability, offering insights into feature importance and promoting transparency. Our framework is scalable and supports real-time monitoring of contracts, helping prevent financial damage by detecting fraud at deployment. This solution enhances the security and reliability of decentralized finance platforms
Blockchain technologies and their application in security software development
Full text linkCurrent factors like the rising frequency of cyber threats and vulnerabilities on centralized platforms indicate the inefficiency of conventional network security frameworks, leading to new solutions such as the blockchain. This review systematically reviews developments of blockchain technologies in the context of security software (2021-2023) to evaluate its efficacy and challenges and explore the future potential. 77 peer-reviewed papers from ScienceDirect, IEEE Xplore and Scopus; adopting the PRISMA guideline, records were screened down from 1,532 to 77. Empirical evaluations (35%), case studies (28%), and theoretical frameworks (37%) using Joanna Briggs Institute tools and the Newcastle-Ottawa Scale were used in mitigating bias. Our results show that blockchain has strengths that add to data integrity (89% of studies) and security of the Internet of Things (IoT) ecosystem (28 studies) and supply chains (15 studies). Nevertheless, blockchain-based authentication has reduced latency by 284% (342 ± 112 ms) compared to a traditional system and has tradeoffs with scalability and performance. Research is skewed towards finance (47%), missing healthcare (9%), and critical infrastructure (6%). It does not include sufficient interoperability standards, post-quantum cryptographic validation, etc. The adaptive regulations are urged for policy implications for editable blockchains and hybrid Artificial Intelligence (AI) blockchain architectures. Interoperability should be taken care of by cross-chain protocols, scalability trilemmas and real-world adversarial testing must be addressed by the researchers and practitioners must put priority on scalability. This review, in its totality, brings out the singular role of blockchain in complementing the existing security solutions instead of replacing them. It calls for cross-disciplinary involvement and partnership in harnessing technical innovation in a regulatory framework to tackle cybersecurity threats through outsider and insider security approaches
Interactive installations and innovative design solutions using artificial intelligence
Full text linkArtificial intelligence (AI) technologies have increasingly penetrated the field of modern design, particularly in the creation of interactive installations. This study aimed to analyze the impact of AI-generated interactive installations on user experience and aesthetic perception. The research employed a comparative approach, content analysis, and case study methodology. A total of 50 scholarly sources were collected and analyzed using the PRISMA protocol to ensure systematic selection and review. The findings demonstrated that key AI tools used in interactive installations include generative adversarial networks (GANs), variational autoencoders (VAEs), computer vision systems, natural language processing (NLP), behavioral analytics, and adaptive machine learning algorithms. These tools, while powerful, require high levels of digital competence, precise configuration, and substantial financial investment. Case analyses of installations such as Living Light, The AI Van Gogh Museum, AI-Driven Storefront, and AI Classcape revealed the following benefits: enhanced interactivity, user personalization, innovative use of AI capabilities in aesthetic experiences, and the emergence of AI as a co-author in artistic creation. AI-driven interactive installations offer significant potential in design and digital art. However, their effectiveness is currently limited by the lack of intuitive human-like creativity, reliance on pre-programmed datasets, and the cost of implementation. The results highlight both the transformative potential and the current limitations of AI as a creative agent in modern design environments
Design of lightweight neural networks for resource-constrained devices
Full text linkPractical realization of modern artificial intelligence systems, especially deep neural networks, on edge platforms presents a daunting challenge. The root cause lies in the critical gap between the computational requirements of these models and the drastically limited capabilities of edge platforms in terms of processing power, memory, storage, and energy consumption. This constraint often requires applications to rely on cloud processing, which presents essential problems in terms of added latency, privacy, and persistent internet connectivity. To overcome this problem, this study presents an architecture for designing and deploying compact neural networks. The methodology begins with the choice of the MobileNet architecture as an initial reference, then adopts advanced model compression schemes, i.e., pruning to eliminate redundant neural connections, and quantization to reduce the numerical precision of weights, substantially contributing to model size reduction and computational requirements. The optimized models were then implemented and evaluated on Raspberry Pi and Arduino Nano boards to test their usability in practical situations. Experimental results clearly demonstrate that optimized models realized an energy consumption reduction of 40% and a latency reduction of 43.75%, while retaining an impressive level of performance in terms of an accuracy loss of less than 3%. This research provides evidence in support of bridging the gap between complex AI and resource-limited hardware, thereby enabling the realization of real-time, compact, and secure on-device intelligent applications
Modeling and simulation of electromagnetic interference from 2G–6G mobile phones on implanted cardiac pacemakers
Full text linkUsing cell phones in proximity to implanted sensitive medical electronic devices has raised concerns about the potential negative impacts of electromagnetic interference (EMI) on pacemaker functioning. This study presents a comprehensive simulation-based analysis of induced fields and voltages in pacemaker leads resulting from electromagnetic emissions of 2G through 6G mobile phones. The EMI model incorporates parameters including distance, orientation, frequency, antenna gain, and burst factor. Simulation results show that 2G and 3G phones, particularly within 10-15 cm at 0° alignment, can induce electric fields exceeding the pacemaker immunity threshold of 3 V/m. Conversely, 5G and 6G technologies, due to higher frequencies and directional emissions, exhibit minimal EMI risks. These findings support updated safety guidelines and call for revised EMI testing protocols considering emerging wireless standards
Using fruit fly and dragonfly optimization algorithms to estimate the Fama-MacBeth model
Full text linkThis research proposes the application of the dragonfly and fruit fly algorithms to enhance estimates generated by the Fama-MacBeth model and compares their performance in this context for the first time. To specifically improve the dragonfly algorithm's effectiveness, three parameter tuning approaches are investigated: manual parameter tuning (MPT), adaptive tuning by methodology (ATY), and a novel technique called adaptive tuning by performance (APT). Additionally, the study evaluates the estimation performance using kernel weighted regression (KWR) and explores how the dragonfly and fruit fly algorithms can be employed to enhance KWR. All methods are tested using data from the Iraq Stock Exchange, based on the Fama-French three-factor model. The results show that the dragonfly algorithm, particularly when using MPT and APT, demonstrates superior performance in improving the accuracy of Fama-MacBeth estimates and enhancing the effectiveness of the KWR approach
Detecting gradual trends: Integrating EWMA control charts with artificial intelligence algorithms (LSTM)
Full text linkControl charts are widely used in statistical process control (SPC) to detect small, gradual shifts in process behavior, although effective at mitigating noise, such as the exponentially weighted moving average (EWMA). Traditional EWMAs, however, face significant challenges and limited adaptability in complex and dynamic environments. In this paper, we propose an improved hybrid approach that integrates EWMAs with artificial intelligence algorithms, such as anomaly detection models, deep learning networks, and unsupervised learning, to enhance the early detection of non-random variations and subtle process trends. Simulations and real-world datasets were used to validate the effectiveness of the integrated model in identifying slow-developing faults
Improving IoT support in Smart Cities through LoRa technology upgrading
Full text linkThe Internet of Things (IoT) has advanced Smart City services through extensive device connectivity. LoRa, a leading LPWAN technology, provides long-range communication with low power consumption but suffers from scalability, latency, and energy-efficiency challenges in dense urban settings. To address these issues, this study introduces an integrated optimization framework that combines adaptive data rate (ADR) control, multi-channel communication, and dynamic resource allocation. The framework aims to reduce transmission delays, minimize packet collisions, and improve overall energy performance. It leverages multi-channel communication to distribute traffic, resource scheduling to prioritize critical data, and ADR to adjust transmission power and data rate based on real-time network conditions. Large-scale simulations conducted in OMNET++ demonstrate significant improvements over standard LoRa configurations, including baseline, ADR-only, and multi-channel setups. In a representative urban environment, the proposed framework achieved packet delivery rates of approximately 92.3% at 300 nodes and 85.7% at 900 nodes, while maintaining low latency and energy consumption. Overall, the integrated approach delivers robust performance across varying node densities, making it a strong candidate for future large-scale IoT deployments in Smart City architectures
A method of representing design solutions in complex systems through model-parametric spaces
Full text linkOn the side of highly complicated systems, it is necessary to have powerful frameworks that can present solution design integration and visualization in the best possible manner, dealing with clarity, scalability, and adaptability. This work aims to formulate an innovative approach for modeling design solutions using model-parametric spaces to create a systematically structured yet convenient method for dealing with multidimensional design complexities. This investigation is conducted within a mixed-method research design combining qualitative assessments of system architecture with quantitative modeling techniques to formulate parametric spaces where design variables and their interrelations are parameterized systematically. The validation of that methodology was done in a way that involves simulated operational scenarios and expert-driven evaluation, which shows the robustness and versatility of the understanding achieved using the approach in different fields of study. Results reveal the utility of model-parametric spaces in vastly increasing the interpretability, modularity, and optimization capability of complex design processes. Therefore, the study argues that this methodology framework is a solid and reasoned basis for decisions in the systems engineering domain and positively explains both research and industrial applications. These future research trajectories determined by this study include further extensive validation within actual project settings to make the developed software more applicable and impactful in the physical world
Improved QPSK modem communication performance for acoustic signals using adaptive digital equalizers
Full text linkTransmission reliability in voice communications is an important challenge due to various effects such as multipath propagation, rapid channel changes, and Doppler shift. In this study, QPSK modulation is combined with the transmitter channel equalizer in the receiver part as well as in the transmitter, where a unit for suppressing audio interference and distortions is developed to reduce the effects of the transmitter channel on the waves transmitted through it. An in-depth analysis of the improved digital equalizer device is discussed by comparing different results obtained by varying the design parameters. An optimized digital channel equalization unit is selected from the simulation results, which is used to create a multimedia wireless audio communication system. Better BER performance is expected at the minimum signal-to-noise ratio (SNR) required at the detector input