Correlation based passive imaging with a white noise source

Initial value problems for second-order hyperbolic equations (35L15) Inverse problems for PDEs (35R30) PDEs with randomness, stochastic partial differential equations (35R60) Geodesics in global differential geometry (53C22) Integral geometry (53C65)

Abstract: Passive imaging refers to problems where waves generated by unknown sources are recorded and used to image the medium through which they travel. The sources are typically modelled as a random variable and it is assumed that some statistical information is available. In this paper we study the stochastic wave equation

p a r t i a l_{t}^{2} u - D e l t a_{g} u = c h i W

, where

W

is a random variable with the white noise statistics on

{m a t h b b R}^{1 + n}

,

n g e 3

,

c h i

is a smooth function vanishing for negative times and outside a compact set in space, and

D e l t a_{g}

is the Laplace-Beltrami operator associated to a smooth non-trapping Riemannian metric tensor

g

on

{m a t h b b R}^{n}

. The metric tensor

g

models the medium to be imaged, and we assume that it coincides with the Euclidean metric outside a compact set. We consider the empirical correlations on an open set

m a t h c a l X s u b s e t {m a t h b b R}^{n}

,

C_T(t_1, x_1, t_2, x_2) = frac 1 T int_0^T u(t_1+s,x_1) u(t_2+s,x_2) ds, quad t_1,t_2>0, x_1,x_2in mathcal X,

for

T > 0

. Supposing that

c h i

is non-zero on

m a t h c a l X

and constant in time after

t > 1

, we show that in the limit

T o i n f t y

, the data

C_{T}

becomes statistically stable, that is, independent of the realization of

W

. Our main result is that, with probability one, this limit determines the Riemannian manifold

({m a t h b b R}^{n}, g)

up to an isometry. To our knowledge, this is the first result showing that a medium can be determined in a passive imaging setting, without assuming a separation of scales.

Review text

The authors consider the wave equation \(\partial _{t}^{2}u(t,x)-\Delta _{g}u(t,x)=\chi (t,x)W(t,x)\) posed in \((0,\infty )\times \mathbb{R}^{n}\), where \(\Delta _{g}\) is the Laplace-Beltrami operator associated to a matrix \( g\), \(\chi \) is a smooth function \(\chi (t,x)=\chi _{0}(t)\kappa (x)\), the function \(\chi _{0}\) being equal to 1 if \(t\leq 0\) (resp. 0 if \( t\geq 1\)), \(\kappa \in C_{0}^{\infty }(\mathbb{R}^{n})\) and \(W\) is the realization of a Gaussian white noise on \(\mathbb{R}^{1+n}\). The authors introduce the correlations \(C_{T}(t_{1},x_{1},t_{2},x_{2})=\frac{1}{T} \int_{0}^{T}u(t_{1}+s,x_{1})u(t_{2}+s,x_{2})ds\) for \(x_{1},x_{2}\) belonging to an open and non-empty subset \(\mathcal{X}\subset \mathbb{R}^{n}\) and prove that the limit \(\lim_{T\rightarrow \infty }\left\langle C_{T},f\otimes h\right\rangle _{\mathcal{D}^{\prime }\times C_{0}^{\infty }(\mathbb{R} ^{2+2n})}\) is deterministic for every \(f,h\in C_{0}^{\infty }(\mathbb{R} ^{1+n})\), assuming that \(g\) is non-trapping and coincides with the Euclidean metric outside a compact set. The main result of the paper proves that, with probability one, the empirical correlations defined in the sense of generalized random variables in \(\mathcal{D}^{\prime }((\mathbb{R}\times \mathcal{X})^{2})\) for \(T>0\), determine the Riemannian manifold \((\mathbb{R} ^{n},g)\) up to an isometry. The authors first prove that the correlations are well-defined. They then analyze the time reversed wave equation \( \partial _{t}^{2}v^{f}-\Delta _{g}v^{f}=f\) in \((-\infty ,S)\times \mathbb{R} ^{n}\) with the initial conditions \(v^{f}\mid _{t=S}=0=\partial _{t}v^{f}\mid _{t=S}\) and they establish a link between this solution and the limit \( \lim_{T\rightarrow \infty }\left\langle C_{T},f\otimes h\right\rangle _{ \mathcal{D}^{\prime }\times C_{0}^{\infty }(\mathbb{R}^{2+2n})}\). They quote from the book by \textit{B. R. Vainberg} [Asymptotic methods in equations of mathematical physics. Transl. from the Russian by E. Primrose. New York etc.: Gordon and Breach Science Publishers (1989; Zbl 0743.35001)] results concerning the local energy decay for the solution of the wave equation \(\partial _{t}^{2}u-\Delta _{g}u=0\) in \((0,\infty )\times \mathbb{R}^{n}\) with compactly supported initial data for \(u\) and \(\partial _{t}u\).

Recommendations

Cites work

Cited in

(35)

This page was built for publication: Correlation based passive imaging with a white noise source

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1653063)