2
$\begingroup$

If we have a unitary map from Hilbert space $H$ to $H$, we get a unitary map from $e^{H}$ to $e^{H}$, where $e^{H}$ is the symmetric Fock space of $H$. But if we replace the unitary with partial isometry, will we get partial isometry at the Fock space level?

My main aim is to study $E_0$ semigroup (on type I factor) coming from semigroup of partial isometries on some Hilbert space.

Improve this question
$\endgroup$
1
  • 2
    $\begingroup$ If $U$ is your partial isometry on $H$, I don't see what prevents you from defining a partial isometry $S(U)$ on $e^H$ by $S(U)(\oplus_{n=0}^\infty(\xi_1\otimes...\otimes\xi_n)=\oplus_{n=0}^\infty(U\xi_1\otimes...\otimes U\xi_n)$. $\endgroup$
    – Alain Valette
    Commented Jan 23, 2013 at 17:14

1 Answer 1

Reset to default
4
$\begingroup$

For any contraction $X$ on $H$, the operator $\Gamma(X)$ (in Alain's notation S(X)) is a contraction. It is essentially algebraic to check that a pair of contractions $X,Y$ on $H$ will satisfy $\Gamma(X)\Gamma(Y)=\Gamma(XY)$ and $\Gamma(X)^\*=\Gamma(X^\*)$. So yes, you do get a partial isometry.

See Parthasarathy's book An Introduction to Quantum Stochastic Calculus (especially Exercise 20.22) for a very readable account of these facts.

(In fact, for any closed densely defined operator $X$ you can define a second quantisation $\Gamma(X)$ which is again closed and densely defined. It is bounded precisely when $X$ is a contraction.)

Improve this answer
$\endgroup$
2
  • 1
    $\begingroup$ thanks! So do you think CCR flow makes sense for semigroup of partial isometries? $\endgroup$
    – Sayan
    Commented Jan 23, 2013 at 18:35
  • 1
    $\begingroup$ If you're asking whether the map $W(f)\mapsto W(U_tf)$ defines an E$_0$-semigroup then the basic condition it has to satisfy is $Im<U_tf,U_tg>=Im<f,g>$ so it must be a symplectic map. If it's complex linear then it must be isometric. If you're asking whether the map $X\mapsto \Gamma(U_t)X\Gamma(U_t)^*$, (and other variants on this theme) defines an E$_0$-semigroup then note that the semigroup property implies $U_t$ should at least be an isometry. Interestingly this leaves space for E$_0$-semigroups from semigroups of symplectic maps $\endgroup$
    – Ollie
    Commented Jan 24, 2013 at 10:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .