Propagation of singularities theorem

From testwiki
Jump to navigation Jump to search

In microlocal analysis, the propagation of singularities theorem (also called the Duistermaat–Hörmander theorem) is theorem which characterizes the wavefront set of the distributional solution of the partial (pseudo) differential equation

Pu=f

for a pseudodifferential operator P on a smooth manifold. It says that the propagation of singularities follows the bicharacteristic flow of the principal symbol of P.

The theorem appeared 1972 in a work on Fourier integral operators by Johannes Jisse Duistermaat and Lars Hörmander and since then there have been many generalizations which are known under the name propagation of singularities.[1][2]

Propagation of singularities theorem

We use the following notation:

  • X is a C-differentiable manifold, and C0(X) is the space of smooth functions u with a compact set KX, such that uXK=0.
  • Lσ,δm(X) denotes the class of pseudodifferential operators of type (σ,δ) with symbol a(x,y,θ)Sσ,δm(X×X×n).
  • Sσ,δm is the Hörmander symbol class.
  • L1m(X):=L1,0m(X).
  • D(X)=(C0(X))* is the space of distributions, the Dual space of C0(X).
  • WF(u) is the wave front set of u
  • charpm is the characteristic set of the principal symbol pm

Statement

Let P be a properly supported pseudodifferential operator of class L1m(X) with a real principal symbol pm(x,ξ), which is homogeneous of degree m in ξ. Let uD(X) be a distribution that satisfies the equation Pu=f, then it follows that

WF(u)WF(f)charpm.

Furthermore, WF(u)WF(f) is invariant under the Hamiltonian flow induced by pm.[3]

Bibliography

References