GRANIT, Green Radio and Adaptive Nodesfor IoT
In the outcoming sober low carbon society, designing low power sensors capable to be autonomous while ubiquitous with negligible energy and carbon footprint is a key challenge designers will have to face. GRANIT team purpose is to design algorithms and architectures able to adapt to environment parameters (wireless channel, traffic, possibilities of energy harvesting), while respecting various applications requirements (data rate, reliability, latency, life time). The GRANIT members have a strong experience on wireless sensor network (WSN) protocols (MAC and PHY layers) and hardware architectures, and developed several WSN platforms for physiological or environmental monitoring applications. In particular, GRANIT team has a strong expertise in these fields:
Energy Harvesting and power management : One of the objectives of the GRANIT team is then to design power management strategies, coupled to above-mentioned adaptive algorithms in order to reach real energy autonomy of the sensor nodes. Cooperation between nodes, (distributed computing or cooperative relaying schemes) represents also a key challenge for the design of energy-efficient wireless systems, where emergence of machine learning solutions represents appealing opportunities.
Wireless Sensor Networks: At the application level, GRANIT team has recently proposed innovative platforms and algorithms for indoor positioning or distributed audio processing. Ultra Wide Band technology achieves very accurate distance estimation and is embedded in last generation smartphones (e.g. Iphone11). Coupled with Inertial Motion Units, it allows to dramatically enhance the indoor positioning, offering new perspectives for logistics and smart factories. GRANIT is also strongly involved in sport monitoring and developed several platforms for cycling and tennis performance enhancement.
Hardware/Software methods : The aim of GRANIT team is also to efficiently implement these algorithms onto different targets, from low power microcontrollers and/or low power FPGAs for WSN solutions to powerful system-on-chip and multi-core systems for more computing-intensive applications. To answer the demand of agile devices, software defined radio (SDR) solutions are especially considered, from high data-rate mobile applications (e.g 5G), to low data-rate wireless sensor networks. Moreover, as SDR represents an efficient way to survey and process a large spectrum, new challenges in terms of security arises (signal interception and associated countermeasures). As most of considered targets are heterogeneous, GRANIT members pay a particular attention to HW/SW partitioning methodology.