Pu (Perry) Wang is a Principle Research Scientist at the Mitsubishi Electric Research Laboratories (MERL), Cambridge, MA, USA. He received his Ph.D. degree in Electrical Engineering at the Stevens Institute of Technology, Hoboken, NJ, USA, in 2011. His research interests include signal processing, statistical learning, Bayesian inference, sparse signal recovery, and their applications to sensing, wireless communications, and networks. In this talk, he will present an overview of their recent works on 1) millimeter wave (mmWave) automotive radar and 2)terahertz (THz) sensing. For the mmWave FMCW-based automotive radar, one issue is inaccuracy in the range estimation due to source nonlinearity, especially when open-loop voltage-controlled oscillators(VCOs) are used due to the cost constraint. Traditional approaches require an elaborate step to measure nonlinearities in the VCO from a reference at a known range. They present a reference-free nonlinearity compensation approach to simultaneously estimate the range profile of multiple reflectors and the source nonlinearity function over a selected basis. For the THz sensing, he will introduce a THz-based encoder system consisting of a single THz transceiver and a scale encoded by pseudo-random code pattern (e.g., m-sequence) and present a variational Bayesian approach to estimate the pseudo-random code pattern from compressed THz measurements. Both works were recently presented at IWS 2018 and ICASSP 2018.