In the existing theory of self-affine tiles, one knows that the Lebesgue measure of any integral self-affine tile corresponding to a standard digit set must be a positive integer and every integral self-affine tile admits some lattice Γ ⊆ Z n as a translation tiling set of R n . In this paper, we give algorithms to evaluate the Lebesgue measure of any such integral self-affine tile K and to determine all of the lattice tilings of R n by K. Moreover, we also propose and determine algorithmically another type of translation tiling of R n by K, which we call natural tiling. We also provide an algorithm to decide whether or not Lebesgue measure of the set K ∩ (K + j), j ∈ Z n , is strictly positive. Proposition 1.1 [10, 17, 19]. The Lebesgue measure of any integral self-affine tile K(B, D) is a positive integer.However, an efficient algorithm to evaluate this measure is not yet available, to our knowledge. We say that a measurable set E ⊂ R n tiles R n using the discrete translation tiling set T if t∈T χ E+t = 1 almost everywhere on R n , where χ A denotes the characteristic function of A. In this case, we also say that E + T is a tiling of R n . One also knows that certain sets of integral translates of K tile R n (see [8,10,16]). In particular, Lagarias and Wang [18] as well as Conze, Hervé and Raugi [4] independently proved that every self-affine tile admits a lattice tiling.Proposition 1.2 [4, 18]. Every integral self-affine tile K(B, D) admits some lattice Γ ⊆ Z n as a tiling set for R n .
Natural tiling and Lebesgue measure of K(B, D)Our first goal is to give a characterization for the integral points in K, which will be needed later on. Before doing so, we need to introduce some terminology.Definition 2.1. Given an integral expansive matrix B and associated standard digit set D, with 0 ∈ D, we have that, for any k ∈ Z n , and every m 11) where d 0 , . . . , d m−1 ∈ D, x m ∈ Z n . So for each k ∈ Z n , we can associate a unique sequence (d 0 , d 1 , d 2 , . . . ), and formally write k ∼ (d 0 , d 1 , d 2 , . . . ).It is easy to prove that, for each given k ∈ Z n , the sequence {x m : m 1} ⊂ Z n in (2.1) is bounded and can, therefore, take only finite distinct values.