The growing importance of digital data security has led to increased exploration of advanced cryptographic methods, including innovative bio-inspired techniques such as DNA cryptography. In this paper, we present an enhanced encryption technique that combines a Feistel Network with DNA cryptography. This is achieved by using DNA encoding alongside an R-Box for mapping octal values to RNA codons, integrated with an XOR operation and random round keys. This hybrid approach achieves an impressive avalanche effect of 56 % with just four Feistel rounds, compared to the 53 % achieved by the DES algorithm in 16 rounds. The results highlight the efficiency and robustness of our method in protecting digital information while reducing computational overhead.
Pollution has a significant impact on public health, particularly for those living in densely populated urban areas affected by industrial and commercial activities. To address this challenge, this paper focuses on the design, development, and prototyping of an Intelligent Network of Anti-Pollution Stations, designed for real-time prediction and monitoring of atmospheric pollution. This network enables the detection of various types of pollution, tracking their geographic spread, predicting emergency situations, and reporting them through real-time monitoring of specific areas. Artificial intelligence plays a key role in enhancing predictive capabilities. Four machine learning algorithms were compared - XGBoost, LightGBM, LSTM, and Random Forest - for predicting the Air Quality Index (AQI), particularly for O3 and NO2.
System of Systems (SoS) represents a complex integration of independent and operational constituent systems that interconnect temporarily to achieve overarching objectives. These sub-systems retain a degree of autonomy and can exhibit significant diversity in terms of technology, context, function, and geography. While SoS can provide capabilities that surpass the sum of their constituents, they also introduce challenges related to unpredictability and complexity due to their dynamic and secure behavior. This paper addresses these challenges by proposing a multi-level framework that investigates how Model-Based System Engineering (MBSE) and formal approaches could be leveraged to model reliable SoS and identify high-impact solutions early in the system development life cycle. This framework, called FRAME-archSoS, comprises (1) an Architectural Description Language (ArchSoS) dedicated to define SoS architectures in a modular and reusable manner, (2) M-SoSRe, a meta model for SoS to identify and represent their reliable scenarios, this step is automated thanks to a generated graphical editor (GE4ArchSoS) to simplify accessibility of people with limited knowledge, (3) a transcription approach from ArchSoS specifications to Maude modules enabling simulation and reconfiguration of SoS behavior while supporting fault tolerance. We illustrate the validity of our proposed framework through an elderly care system (ECG-SoS) example, demonstrating effective complexity management and system integrity.
The development and availability of high-quality datasets are critical to the success of machine learning and artificial intelligence (AI) systems. Datasets provide the essential foundation for training, validating, and testing models. The need for domain-specific datasets is particularly important in closed-domain systems like Movie NoSQL Question Answering (QA). A well-curated dataset enables models to translate natural language questions into NoSQL queries effectively. This paper introduces an efficient method for developing a tailored dataset for MongoDB-based Movie QA systems. The resulting dataset, M2$Q$2+, facilitates the translation of natural language questions into MongoDB queries (MQL). The dataset development process includes creating templates, performing data augmentation with the pre-trained T5-base model, and conducting a data revision process to produce the final dataset. A combination of evaluation methods and metrics was applied to ensure that M2$Q$2+ is high quality.
The Agricultural Internet of Things (AIoT) network integrates agriculture with IoT to enhance productivity and improve quality. This infrastructure connects many devices, facilitating extensive information exchange within the network. However, this increased data exchange also heightens the network's susceptibility to attacks and vulnerabilities, posing risks and potentially causing damage to the network infrastructure. Intrusion Detection Systems (IDS) are effective solutions for detecting attacks and preserving productivity. Many existing approaches in the literature combine deep learning with IDS to create robust anomaly-based IDS. However, these solutions often encounter challenges related to data collection, which compromises user privacy by centralizing data from various entities into a single entity. Our paper proposes a privacy-preserving IDS based on federated learning techniques and Convolutional neural networks (CNN) to develop a robust IDS with significant performance. We evaluate our experiments using a new dataset, Edge-IIoT, which consists of real IoT traffic across binary, 6-class, and 15-class scenarios with 10 and 15 clients. The results demonstrate the performance of our model across all scenarios, achieving 100% and 99.99% accuracy in binary classification and both 6-class and 15-class scenarios, respectively.
