On the counting function of the lattice profile of periodic sequences (Q2465280)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 5224936
Language Label Description Also known as
default for all languages
No label defined
    English
    On the counting function of the lattice profile of periodic sequences
    scientific article; zbMATH DE number 5224936

      Statements

      On the counting function of the lattice profile of periodic sequences (English)
      0 references
      9 January 2008
      0 references
      A sequence \(\mathbf{S} = s_0,s_1,\ldots\) over a finite field \(\mathbb F_q\) is said to pass the \(t\)-dimensional \(n\)-lattice test if the vectors \(\underline{\mathbf{s}}_i - \underline{\mathbf{s}}_0\), \(i = 1,\ldots,n-t\), span \(\mathbb F_q^t\), where \[ \underline{\mathbf{s}}_i = (s_i,s_{i+1},\ldots,s_{i+t-1}),\quad 0 \leq i \leq n-t. \] The lattice profile \(T(\mathbf{S},n)\) of \(\mathbf{S}\) at \(n\) is then defined to be the greatest integer \(t\) such that \(\mathbf{S}\) passes the \(t\)-dimensional \(n\)-lattice test, and the lattice profile of \(\mathbf{S}\) is defined as \[ T(\mathbf{S}) = \sup_{n \geq 2}T(\mathbf{S},n), \] which is finite if \(\mathbf{S}\) is a periodic sequence. The authors use relations between linear complexity and the lattice profile for periodic sequences presented in \textit{H. Niederreiter} and \textit{A. Winterhof} [AAECC 13, 319--326 (2002; Zbl 1033.11038)], and \textit{G. Dorfer} and \textit{A. Winterhof} [Proceedings of MCQMC 2002, 199--211 (2004; Zbl 1043.65007)] to obtain explicit formulas for the expected value and the variance for the lattice profile of a random \(N\)-periodic sequence. Furthermore for certain period length closed formulas for the number of sequences with given lattice profile are provided.
      0 references
      linear complexity
      0 references
      discrete Fourier transform
      0 references
      expected value
      0 references
      variance
      0 references
      0 references
      0 references

      Identifiers