Skip to main content
European Commission logo
English English
CORDIS - EU research results
CORDIS
CORDIS Web 30th anniversary CORDIS Web 30th anniversary

MassiVe MIMO Radio Channel- and Learning-based Indoor Passive Localization and Posture Identification for Multiple Humans

Project description

Advanced antenna technology for human posture recognition

Older people have the highest risk of death or serious injury from a fall. Funded by the Marie Skłodowska-Curie Actions programme, the V.I.P. project plans to develop radio technology for identifying the posture of multiple humans in indoor environments. Researchers will leverage massive multiple-input multiple-output antenna technology and neural network algorithms that should help recognise the interactions between the human form (body type/shape, posture, motion) and the indoor radio channel. The proposed research is multidisciplinary, combining networked systems, electromagnetism, smart building, human factors and data science.

Objective

The aim of the proposed V.I.P. project is to achieve passive localization and posture identification of multiple humans in indoor environment for assisted-living functionalities. We propose to use the temporal-spatial-delay domain radio channel properties with dedicated design of massive Multiple-Input Multiple-Output (MIMO) topology and configuration for specific physical environment, and use the neural-networks-based learning algorithm recognizing the interactions between the human form-factors (body type/shape, posture, motion) and the aforementioned indoor radio channel. Both simulation and measurement platforms will be built to estimate and measure the radio propagation resulting from the reconfigurable physical environment, multiple-antenna systems, and human form-factors.

The proposed research is multidisciplinary, connecting networked systems, electromagnetism, smart building, human factors, and data science. It proactively combines MIMO radio networked system with smart building, targeting at identifying MULTIPLE humans with the help of neural networks. It is proposed by a female researcher with rich international experience, and is supervised by a female professor with top track record. The proposed research includes gender-dependent modelling and parameterization on the propagation mechanism of human-radio interaction, potentially using the double-layer reflection and the uniform theory of diffraction. The body postures are simplified into geometries with homogeneous dielectric, e.g. a standing/laying human as ellipsoid, sitting on the ground with legs crossed as triangle.

Upon the completion of the research, training and networking of V.I.P. the researcher will profit from the first-hand experience on designing and developing the networked system for specific use case, developing neural-network-based pattern recognition methodology, as well as growing towards professional maturity from proposing, coordinating, finance managing, educating to impacting.

Coordinator

KATHOLIEKE UNIVERSITEIT LEUVEN
Net EU contribution
€ 166 320,00
Address
OUDE MARKT 13
3000 Leuven
Belgium

See on map

Region
Vlaams Gewest Prov. Vlaams-Brabant Arr. Leuven
Activity type
Higher or Secondary Education Establishments
Links
Total cost
€ 166 320,00