2020/02/22 by Vikram Krishnamurthy, Krishnamurthy, Vikram
Computer Science · Physics and Astronomy · #Distributed Sensor Networks and Detection Algorithms #FOS: Electrical engineering #Signal Processing (eess.SP) #Statistical Mechanics and Entropy #Target Tracking and Data Fusion in Sensor Networks #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2002.10910
openalex publication_date 2020/02/22 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
Cognitive sensing refers to a reconfigurable sensor that dynamically adapts\nits sensing mechanism by using stochastic control to optimize its sensing\nresources. For example, cognitive radars are sophisticated dynamical systems;\nthey use stochastic control to sense the environment, learn from it relevant\ninformation about the target and background, then adapt the radar sensor to\nsatisfy the needs of their mission. The last two decades have witnessed intense\nresearch in cognitive/adaptive radars.This paper discusses addresses the next\nlogical step, namely inverse cognitive sensing. By observing the emissions of a\nsensor (e.g. radar or in general a controlled stochastic dynamical system) in\nreal time, how can we detect if the sensor is cognitive (rational utility\nmaximizer) and how can we predict its future actions? The scientific challenges\ninvolve extending Bayesian filtering, inverse reinforcement learning and\nstochastic optimization of dynamical systems to a data-driven adversarial\nsetting. Our methodology transcends classical statistical signal processing\n(sensing and estimation/detection theory) to address the deeper issue of how to\ninfer strategy from sensing. The generative models, adversarial inference\nalgorithms and associated mathematical analysis will lead to advances in\nunderstanding how sophisticated adaptive sensors such as cognitive radars\noperate.\n