Partial identification of the potential from phase shifts (Q536276): Difference between revisions
From MaRDI portal
Added link to MaRDI item. |
Set profile property. |
||
Property / MaRDI profile type | |||
Property / MaRDI profile type: MaRDI publication profile / rank | |||
Normal rank |
Revision as of 01:34, 5 March 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Partial identification of the potential from phase shifts |
scientific article |
Statements
Partial identification of the potential from phase shifts (English)
0 references
16 May 2011
0 references
Consider the inverse scattering problem for the operator \[ \phi ^{^{\prime \prime }}(r,\lambda)-\frac{\lambda ^{2}-\frac{1 }{4}}{r^{2}}\phi (r,\lambda)+(1-q(r))\phi (r,\lambda)=0,\quad r\geq 0, \] \[ \phi (r,\lambda)=\gamma (\lambda) r^{\lambda + \frac{1}{2}}(1+o(1))\text{ as }r\rightarrow 0+, \] \[ \phi(r,\lambda)=\sin \left( r-\tfrac{\pi }{2}\left( \lambda - \tfrac{1}{2}\right) +\delta (\lambda)\right) +o(1) \text{ as }r\rightarrow +\infty \] with fixed energy \(k^{2}=1\). The quantities \(\delta \left( n+\frac{1}{2} \right) =\delta _{n}\) are called phase shifts and \(\gamma \left( n+\frac{1}{2 }\right) =\gamma _{n}\) for \(n\geq 0.\) The inverse scattering problem here consists of the recovery of the potential \(q\)\ from the phase shifts \(\delta _{n}.\) The scattering amplitude \(F(t)\) can be expressed by the phase shifts: \[ F(t)=\sum_{n=0}^{\infty }(2n+1) F_{n}P_{n}(t),\quad F_{n}=e^{i\delta _{n}}\sin \delta _{n}, \] where the functions \(P_{n}(t)\) are the Legendre polynomials. This paper shows that the tail of the potential can be determined by the asymptotical behavior of the scattering amplitude and of the phase shifts. The following results are obtained: Assume that the potential \(q\) satisfies \(rq(r),rq^{\ast }(r)\in L_{1}(0,\infty)\) and let \(0<a<\infty \). (a) If \(q=q^{\ast }\) a.e. on \((a,\infty)\) then \(F(t) -F^{\ast }(t)\) is an entire function and satisfies \[ |F(t)-F^{\ast }(t)| \leq c(1+|t|)\exp (a\sqrt{2|t|}). \] Conversely, if \(q\) and \(q^{\ast }\) have compact support, \(F(t) -F^{\ast }(t)\) is an entire function and \[ F(t) -F^{\ast }(t) =O(\exp (a_{1}\sqrt{ 2|t|})) \] for all \(a_{1}>a\), then \(q=q^{\ast }\) a.e. on \((a,\infty)\). (b) If \(q=q^{\ast }\) a.e. on \((a,\infty),\) then, for all sufficiently large \(n\), it holds \[ |\delta _{n}-\delta _{n}^{\ast }| \leq \frac{c}{n^{2}} \left( \frac{ae}{2n}\right) ^{2n}. \] Conversely, if \(q\) and \(q^{\ast }\) have compact support and \[ \delta_{n}-\delta _{n}^{\ast }=O\left( \left( \frac{a_{1}e}{2n}\right) ^{2n}\right)\text{ as } n\rightarrow \infty \] holds for all \(a_{1}>a\), then \(q=q^{\ast }\) a.e. on \((a,\infty)\).
0 references
inverse scattering
0 references
phase shifts
0 references
scattering amplitude
0 references