This work introduces an alternative radio-frequency (RF) sensing paradigm termed absence-driven RF sensing. Unlike conventional radar and RF sensing systems that rely on reflected or scattered energy, the proposed framework infers object presence from structured packet loss patterns within a synchronized RF network. Spatially distributed RF nodes exchange known packetized transmissions and monitor reception outcomes. Physical obstructions introduce coherent and spatially correlated packet loss across intersecting propagation paths. By aggregating these loss patterns at the network level, object presence and approximate location can be inferred without reliance on waveform-level measurements, signal amplitude, or radar cross-section. A theoretical detectability bound is derived, demonstrating that occlusion-induced loss can be distinguished from stochastic channel effects using packet-level statistics alone. Numerical simulations validate the framework and illustrate coarse localization via loss aggregation.
Suman Pramanik (Fri,) studied this question.