Operator integrals and phase space observables

From MaRDI portal





According to M. A. Naimark's dilation theorem any positive additive operator-valued measure may be dilated to an orthogonal spectral measure. The authors consider an example of a spectral measure generated on a product space \(H\times L_2(R_2)\) by the Lebesgue measure \(dx\wedge dy\) on the plane and the unitary transform \(U= e^{i\sqrt 2(yP- xQ)}\), with momentum \(P\) and coordinate generated by the pair of annihilation/creation operator in \(E\), \(dE= U^+ dx\wedge dy\cup\). The projection of it onto the sector in \(E\) with a certain number of particles \(S\); in \(H\), \(P_s U^+dx\wedge dy\cup P_s:= dA_s\) is a positive additive operator-valued measure in \(P_sH\times L_2\). The authors consider the operator integrals \[ \int(x+ iy)^k dA_s\equiv a^k,\quad \int(x- iy)^k dA_s\equiv (a^*)^k,\quad k> 0 \] and derive the estimation of the ``uncertainty product for the operators \[ {1\over 2} (a^k+ a^{*k}),\quad {1\over 2i} (a^k- a^{*k}). \] Remarks. Spelling Newmark instead of Naimark is misleading. The authors do not check the commutation relations for operators \(a\), \(a^*\) obtained as operator integrals and do not check the multiplicative property \[ \int(x\pm iy)^k dA_s= [\int(x\pm iy) dA_s]^k, \] which is nontrivial since the new measure is not orthogonal. Thus the connection between the initial creation-annihilation pair and the pair defined by the operator integrals is still in question.











This page was built for publication: Operator integrals and phase space observables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4701847)