NS41B-1942
Mono-static GPR without transmitting anything for pavement damage inspection: interferometry by auto-correlation applied to mobile phone signals

Thursday, 17 December 2015
Poster Hall (Moscone South)
Ralph Feld, Delft University of Technology, Delft, Netherlands and Evert C Slob, Delft University of Technology, Delft, 5612, Netherlands
Abstract:
Creating virtual sources at locations where physical receivers have measured a response is known as seismic interferometry. A much appreciated benefit of interferometry is its independence of the actual source locations. The use of ambient noise as actual source is therefore not uncommon in this field. Ambient noise can be commercial noise, like for example mobile phone signals. For GPR this can be useful in cases where it is not possible to place a source, for instance when it is prohibited by laws and regulations. A mono-static GPR antenna can measure ambient noise. Interferometry by auto-correlation (AC) places a virtual source on this antenna’s position, without actually transmitting anything. This can be used for pavement damage inspection.

Earlier work showed very promising results with 2D numerical models of damaged pavement. 1D and 2D heterogeneities were compared, both modelled in a 2D pavement world. In a 1D heterogeneous model energy leaks away to the sides, whereas in a 2D heterogeneous model rays can reflect and therefore still add to the signal reconstruction (see illustration). In the first case the amount of stationary points is strictly limited, while in the other case the amount of stationary points is very large. We extend these models to a 3D world and optimise an experimental configuration.

The illustration originates from the journal article under submission ‘Non-destructive pavement damage inspection by mono-static GPR without transmitting anything’ by R. Feld, E.C. Slob, and J.W. Thorbecke. (a) 2D heterogeneous pavement model with three irregular-shaped misalignments between the base and subbase layer (marked by arrows). Mono-antenna B-scan positions are shown schematically. (b) Ideal output: a real source at the receiver’s position. The difference w.r.t. the trace found in the middle is shown. (c) AC output: a virtual source at the receiver’s position. There is a clear overlap with the ideal output.