Localization-of-things in 5G and beyond ecosystems - Develop machine learning approaches, called soft information (SI), with unprecedented performance in complex wireless environments. Design of SI-based localization algorithms that learn the environment and fuse data from heterogeneous sensors. Perform experimentations with ultra-wideband (UWB) radios. Quantifying localization performance in full conformity with the 3GPP specifications by ETSI at both sub-7 GHz and mmWaves.

Device-free multi-target tracking, identification, and activity recognition – Establish foundations for device-free localization of multiple targets in cluttered environments. Develop algorithms for integrated tracking and identification of multiple targets and for activity recognition based on reflected signals. Perform experimentation with mmWaves MIMO radar.

Adaptive diversity communication systems – Design and analysis of adaptive diversity communication systems that provide reliable and efficient operation in wireless environments. Derive a new class of upper and lower bounds on multichannel communication performance with non-ideal channel estimation.

Wireless resource optimization – Determine network performance metrics as functions of the wireless resources and nodes deployment for the design and operation of communication and location-aware networks. Develop strategies for deploying assisting (i.e. cooperative) nodes to improve network performance in complex wireless environments. In the framework of 5G and beyond localization, develop network operation strategies for node prioritization, node selection, and node deployment in complex wireless environments.

Multidimensional stochastic sampling – Design and analysis of wireless sensor networks with application to multidimensional signal reconstruction from spatiotemporal stochastic samples. Derive of the optimal interpolator that minimizes the reconstruction error as a function of multidimensional signal characteristics (signal spectrum and spatial correlation) and sampling properties (sensor spatial distribution, sample availability, and sensor position knowledge).

Distributed inference – Determine the accuracy of decentralized inference in wireless networks and established both necessary conditions and sufficient conditions for boundedness of the inference error in terms of the nodal sensing and communication capabilities. Develop real-time encoding strategies for generating the information-carrying messages exchanged among different nodes for decentralized inference.