Figure 1

Passive NLoS localization process using uncontrolled noise sources (numerical example). (a) The simulated scene (top view): a target is hidden behind an occluder. A 16-detectors array records the continuous broadband noise emitted by a nearby uncontrolled source, which reverberates in the scene. The recorded noise contains directly arriving signals, single reflections (in green, i), diffracted reflections (cyan, ii), and multiple reflections (magenta, iii) allowing NLoS localization. (b) Noise fields \(v_1(t), v_2(t)\) recorded by detectors 1,2, respectively. (c) Cross-correlation of the recorded fields \(C_{12}(t)\), reveals pulse-echo-like ToF information containing: (i) direct reflections from the wall; (ii) fields that originate from diffraction by the occluder edge to the target; (iii) fields that reflect by the wall to the target and back. These are used for direct localization of the target mirror image. (d) Delay-and-sum beam-forming reconstruction from 16 \(\times\) 16 cross-correlations (as in c) for all detector pairs. The positions of the wall (green arrow), the target mirror image (magenta arrow), as well as the edge diffraction artifact (cyan arrow) are visible. (e,f) Same as (c,d), for a scene without the hidden target. (g) Difference between the cross-correlations of (c) and (e). (h) Difference between (d) and (f) shows only the hidden target contributions. The figure was created using MATLAB R2022a and INKSCAPE 1.2.