Math 529: Functional Analysis
Introduction to Harmonic Analysis
David L Ragozin
Spring 2000, MWF 12:30-1:20
Prerequisite: Functional Analysis(Math 527/8) although its prerequisites
-- Real Analysis (Math 524/5/6) or Linear Analysis (Math 554/5/6) -- will
suffice for much of what we will cover.
Text: H. Dym and H. P. McKean,
Fourier Series and integrals.
Content: An introduction to both classical and modern harmonic
analysis in concrete contexts with applications. These chapters in
Dym and McKean will be covered:
-
Fourier series (development and convergence of expansions of periodic functions
or functions on the circle -- R/Z or higher-dimensional tori
Rn/
Zn
-- in terms of exponentials
exp(2
ik
.t)
),
-
Fourier integrals (the non-periodic expansion and convergence questions
on R or Rn),
-
Fourier integrals and complex function theory (what happens when a function
on the circle or the real line extends to a holomorphic function on some
subset of the complex plane?),
-
Fourier series and integrals on groups (topics like the preceding for functions
on the rotation group O(3) of 3 x
3 orthogonal matrices and the group SL(2,R) of 2 x
2 matrices of determinant 1).
These contain many applications, such as the isoperimetric problem, Jacobi's
identity for the theta function, heat flow, wave motion, random walks,
the central limit theorem, Heisenberg's inequality, Minkowski's theorem
in the geometry of numbers, polynomial approximation, the distribution
of primes, Gauss' law of quadratic reciprocity, representations of the
rotation group SO(3).
In parallel to the Dym and McKean presentation, the course will develop
the perspective of Fourier series and integrals relative to operators on
some class of functions on a geometric space X on which a group
G
acts when those operators commute with the action of the group.
Examples will include X = Sn
(the n-sphere) with the action of the orthogonal matrices and any
rotation invariant kernel operator or differential operator, and X =
the upper half-plane with the action of SL(2,R) via linear fractional
transforms and various kernel and differential operators.