Structure/Order of explaining distribution-theory

75 Views Asked by At

please tell me if this is too off-topic, then I will delete the post. In my master thesis I want to use (and explain) distributions in order to solve some PDE, but since covid still is present it is hard for me to get academic feedback, so I thought maybe you guys can have a quick look and tell me whether the following structure is fine for introducing distributions. My global structure will be something like this:

  • Introduction: PDEs are everywhere and solving them is hard
  • Main part
    • PDEs
    • Distributions
    • Fourier Transforms
    • The Laplace Equation
    • The Wave Equation
    • A hypoelliptic operator of fourth order
  • Conclusion: PDEs are still hard, but with distributions and fourier transforms specific problems can often be solved and surprising facts be derived

Now for Distributions I wanted to introduce it with "everything can be differentiated" and "Then for example the divergence theorem reduces to calculating the derivative of a characteristic function." The structure should be like:

  • Derivative of the Heaviside function
    • A first example.
    • There is no weak derivative for $\mathcal{H}$, but we want to have something like $\delta$.
  • Test functions
    • There are functions in $\mathcal{C}_c^\infty$, for example the standard mollifier.
    • Also $lim_{\epsilon \rightarrow 0} \eta_\epsilon = \delta$
    • $\phi \mapsto \phi^{(k)}(0)$ should be a distribution, so there needs to be a topology on $\mathcal{C}_c^\infty$ that respects the derivatives.
    • $\mathcal{D}$ has the topology induced by $\sup_{x \in K} sup_{\alpha \in \mathbb{N}^n} \partial^\alpha \phi$.
    • This is not metrizable.
  • Definition of Distributions
    • $\mathcal{D}'$ is the dual space of $\mathcal{D}$.
    • One can easier work with the usual definition of a linear map T with $|T(\phi) \leq C \sum_{|\alpha| \leq k} sup_{x \in K} |\partial^\alpha \phi(x)|$.
    • Operators can be defined by its behaviour on test function, for example: $<\partial^\alpha T ,\phi> = <T , (-1)^\alpha \partial^\alpha \phi >$
    • Every distribution is the limit of a series of Testfunctions
  • Support of a Distribution
    • One can define the support of a distribution
    • A distribution on $ X \subset \mathbb{R}^n$ can be extended to $\mathbb{R}^n$
    • We can restrict Distributions to a subset
    • Multiplication of Distributions
    • Definition of singsupp
  • Operators on Distributions
    • The Tensor product
    • Convolution with a function
    • Convolution in general
  • Working with Distributions
    • if $T' = 0$ then T is constant
    • calculating some derivatives
    • Complex Distributions
    • Calculating some limits
  • Fundamental solutions
    • Definition
    • Application of $P (E_P * f) = f$ and $ u = E_P * (P u)$

I'm aware that this is somewhat unstructered. But this just reflects that I can't differ between technicalities and important theorems, apart from that fundemantal solutions are important for solving the inhomogeneous equation.

Thanks already and as I said, if it is too off-topic I will delete the question.