Intelligent Wireless Networking
IDLab-iWiNe has about 25 years of research experience on wireless networking with a prime focus on end-to-end solutions covering all layers in the end-to-end protocol stack. The team currently focuses on following research lines:
- Deterministic Connected Systems: An emerging requirement is guaranteed and predictable wireless network performance. For this, we design connectivity services with deterministic end-to-end guarantees using programmable wireless & wireline hardware. This encompasses both digital hardware, software, protocol and algorithmic designs to monitor, control, manage and customize end-to-end programmable connected systems. Key technologies being considered are next-generation Wi-Fi, beyond 5G/6G, converged Wi-Fi/5G and time-sensitive networking (TSN).
For Wi-Fi, we have openwifi, our own full-stack standard-compliant Wi-Fi design that has been implemented on various software defined radio platforms and that has been extended with time-sensitive features compatible with Ethernet TSN, as well as various wireless TSN lab setups for experimentation. For 5G, we have at our disposal multiple O-RAN compliant units, both commercial-of-the-shelf and software-defined radio, that enable testing, experimentation and research.
- IoT and sensor networks: To enable reliable IoT applications running on constrained devices, we perform IoT system design and prototyping. We assess, extend and improve wireless IoT communication technologies, both short-range (BLE, UWB, 802.15.4) and long range (LoRaWAN, NB-IoT, 802.11ah), and embedded protocol stacks to increase resilience and performance, and/or to reduce the energy consumption. The whole device lifecycle, spanning commissioning, operation and management, is considered as well as the end-to-end system design, spanning data capturing, transmission and processing.
- AI/ML for Wireless: We focus on investigating AI/ML approaches in three main domains: 5G/6G/satellite networks, indoor localization, and connected healthcare. This includes several challenging research topics such as wireless technology recognition, radio resource management, traffic steering, networks of networks, dynamic spectrum sharing, UWB error correction, activity recognition, ECG classification, and more. Additionally, inspired by the progress of large language models, we are exploring foundation models for wireless communications and networks, with the aim of developing a general model that can be applied to multiple tasks without the need to train separate AI/ML models for each individual task.
- Radar & localization: We specialize in using advanced wireless technologies, including WiFi, UWB, 6G, and Bluetooth, for a range of applications such as location estimation and radar-based wearable-free person monitoring. Our work in person spans vital sign monitoring, activity recognition, and device-free localization. We develop cutting-edge software algorithms and AI models, grounded in a robust theoretical foundation and validated through comprehensive real-world experimentation. Our focus is on achieving multi-metric localization & radar improvements, such as enhanced accuracy, scalability, generalizability, and energy efficiency, under realistic conditions.
