On the classification of perfect codes: extended side class structures (Q1045141)

Two binary codes \(C_1,C_2 \subseteq \mathbb{F}_2^n\) are said to be equivalent if there is a word \(w \in \mathbb{F}_2^n\) and a permutation \(\pi \in S_n\), such that addition by \(w\) followed by the action of \(\pi\) maps the codewords of \(C_1\) onto the codewords of \(C_2\). The action of \(\pi\) can be obtained by multiplying the codewords by a permutation matrix. If one modifies the definition above to allow multiplication by an arbitrary nonsingular matrix (instead of a permutation matrix), then one gets the definition of extended equivalent codes. After introducing the concept of extended equivalent codes, the authors study binary perfect codes with respect to this concept, prove various related results, and classify such codes in some particular cases.