Let be an analytic periodic forcing on a complex strip around a torus, and let be a Diophantine frequency vector. The cohomological equation on a Diophantine torus
has an analytic periodic solution on every smaller strip exactly when has zero average. Integrating over the torus proves necessity, since each derivative has zero integral.
In the period- convention, comparison of Fourier coefficients gives
The Diophantine lower bound prevents zero denominators and limits their inverse growth by a power of . Exponential Fourier decay of dominates this loss on a smaller strip, as quantified by the analytic estimate for the torus cohomological equation. The zero coefficient of is arbitrary, so imposing zero mean makes the solution unique. Without analytic regularity of the forcing, zero average alone does not establish an analytic solution.
For a sufficiently small analytic periodic vector field , a near-identity angular diffeomorphism and a constant vector solve . They satisfy , ; if , then . For the defect , set , choose , and solve with zero mean. Updating by leaves a quadratic defect . The torus small-divisor estimate and geometrically decreasing strip losses yield a convergent analytic iteration.
Suppose is bounded and analytic on the complex torus strip , has zero mean, and
For , the normalized solution of the cohomological equation on a Diophantine torus obeys
Contour shifting bounds . On the smaller strip the Fourier solution is consequently bounded by
There are integer vectors with . Comparing the resulting sum with its exponential integral bounds it by . This gives both convergence and the stated loss-of-strip estimate. It is sufficient for analytic conjugacy iterations even when sharper small-divisor estimates are available.

Articles by others on the same topic (0)

There are currently no matching articles.