On positivity, criticality, and the spectral radius of the shuttle operator for elliptic operators (Q2564684): Difference between revisions

Consider an open subset \(D\) of a Euclidean space and \(P\) a second order elliptic operator on it with smooth coefficients. The question addressed in this paper is the following: do there exist positive solutions of the equations \(Pu=0\) in \(D\)? No conditions on the behaviour of \(u\) at infinity are imposed. One instance when such solutions exist is when \(P\) admits a positive minimal Green's function. In such an instance \(P\) is called subcritical. Even if such positive Green's functions cannot be found then positive solutions still may exist in which case \(P\) is said to be supercritical. In the worst case when no positive solutions exist \(P\) is said to be critical. The (sub, super) criticality of \(P\) also depends on the domain \(D\). Currently, there are no intrinsic characterizations of (sub, super) criticality. The main result of this paper provides such a characterization under additional assumptions on \(P\) and \(D\). The criticality is decided by the norm of the shuttle operator which is an exchange operator between an interior and an exterior Dirichlet problem. The operator \(P\) is (sub, super) critical if and only if the norm of the shuttle operator is \((\leq, \geq) =1\).