Tight wavelet frames on the space of \(M\)-positive vectors (Q6580207)

The paper involves the study of wavelets on the sets of \(M\)-positive vectors in the Euclidean space. An \(M\)-positive vector arises from a dilation matrix, i.e., a square integer matrix whose eigenvalues are all greater than \(1\) in absolute value. \(M\)-positive vectors are analogous to the set of non-negative numbers. The space of \(M\)-positive vectors is denoted by \(X.\) The nice thing about the space \(X\) is the existence of a class of compactly supported functions, called test functions, whose Fourier transform is also compactly supported. A function \(\phi \in L^2(X)\) is called refinable if its Fourier transform satisfies a certain refinement equation. The mask of a refinable function \(\phi\) is a Walsh polynomial of a certain order. In the paper, wavelet frames have been constructed in the traditional way by a multiresolution analysis generated by a refinable function. The tight wavelet frames that are constructed consist of test functions. A complete description of the masks generating such frames is given, and an algorithmic method for constructing them is developed. It is expected that these frames will be useful for applications to signal processing.