A federated learning evaluation across several contemporary IoT and IIoT intrusion detection datasets, benchmarking algorithms such as FedAvg, FedProx, and FedNova with LSTM and Transformer models in in-domain, cross-dataset, and multi-dataset federation scenarios.
IoT-DH is a real-world IoT DDoS honeypot dataset collected from a honeypot deployment and converted from PCAP to CSV with traffic features and labels for DDoS classification, identification, and detection tasks.
A pair of labeled MQTT and UDP DDoS datasets for Healthcare-IoT networks, generated with Cooja and ns-3 simulators to support evaluation of DDoS detection and mitigation techniques in H-IoT environments.
A real-world IoT dataset from a multi-purpose university building at University of Sharjah, capturing appliance-level energy consumption, temperature, humidity, and occupancy, along with 2D Markov Transition Field (MTF) image representations for deep learning, published in Data in Brief.
Massive dataset from solar panels, wind turbines, and smart grid infrastructure for energy forecasting, demand prediction, and efficiency optimization using big data analytics, machine learning, Hadoop, and Spark distributed processing frameworks.
Detailed energy monitoring dataset from smart home testbed with five common household appliances (refrigerator, washing machine, microwave, air conditioner, TV) each connected to individual smart meters for appliance-level consumption analysis and NILM research.
Multi-resolution smart building energy dataset for forecasting competition with three versions: 1-year at 5-min intervals (v1.0), 40-day at 5-min (v2.0), and 1-day hourly (v3.x), designed to benchmark state-of-the-art energy prediction techniques.
Standardized energy flexibility data model enabling uniform communication of energy flexibility potentials within industrial companies and in exchange with external energy systems (grids, aggregators), supporting demand response and sector coupling.
Multimodal dataset from 24 participants with synchronized wrist-worn PPG from Galaxy Watch 5 and Empatica E4, plus chest ECG from Polar H10, collected during diverse activities in semi-naturalistic settings for evaluating consumer-grade wearable performance.
Large public dataset from Microsoft Research Aurora Project with ECG and PPG signals from wrist-worn wearables, balanced for gender, age, and hypertension status, enabling causal inference research for non-invasive blood pressure prediction with 205 extracted features.
Deep learning-ready dataset combining real vehicle CAN bus traffic and simulated attack scenarios (DoS, fuzzing, spoofing) for training intrusion detection systems to protect autonomous and connected vehicles from cyber-attacks.
Real operational electricity system records from ISO New England (ISONE) covering urban smart city energy activity including demand, generation, pricing, and grid operations across multiple metropolitan areas for energy analytics and forecasting.