Wavelets and Besov spaces on Mauldin-Williams fractals (Q2371929)

The author extends the earlier work of Jonsson, who constructed Haar type wavelets of higher order on self-similar fractal sets and then used them to characterise Besov space on a class of totally disconnected self-similar sets, to a class of strong connected Mauldin-Williams (MW) fractals. To overcome the technical difficulties in using Jonsson's method, namely the difficulty in calculating the inner product of \(L^2(\mu)\) where \(\mu\) is a measure supported on a fractal set, the author extends the earlier work of \textit{Weiqiang Jin} [An explicit construction of wavelets on self-similar fractals, Research Reports No. 5, Dept. of Mathematics, Umeå University (1998)]. Jin showed that the Gram-Schmidt process used in Jonsson's work can be reduced to a calculation of moments which themselves can be calculated recursively. The author extends this to their class of strongly connected MW-fractals. That paper begins with a review of MW-fractals and the author establishes necessary and sufficient for a strongly connected MW-fractal of dimension \(d\) to satisfy the open set condition. The author then shows that a certain subclass of these fractals also preserve Markov's inequality. The paper continues with a review of the aforementioned method of calculating a wavelet basis for \(L^2(\mu)\) and includes the authors extension to the subclass of strongly connected MW-fractals discussed earlier. This section includes a number of illustrative examples used to describe the construction and the moment calculation. The paper concludes with the extension of Jonsson's characterisation of Besov space to the MW-fractals. This is accomplished by showing that the norm of a function in Besov space on this class can be calculated using the magnitudes of the coefficients of the wavelet expansion discussed in the previous section.