Proposed by Peter Nightingale

Informally, the problem is to find a set (optionally of maximal size) of codewords, such that any pair of codewords are Hamming distance $d$ apart. Each codeword (which may be considered as a sequence) is made up of symbols from the alphabet $\{1, \ldots, q\}$, with each symbol occurring a fixed number $\lambda$ of times per codeword.

More precisely, the problem has parameters $v$, $q$, $\lambda$, $d$ and it is to find a set $E$ of size $v$, of sequences of length $q\lambda$, such that each sequence contains $\lambda$ of each symbol in the set $\{1, \ldots, q\}$. For each pair of sequences in $E$, the pair are Hamming distance $d$ apart (i.e. there are $d$ places where the sequences disagree).

For the parameters $v=5$, $q=3$, $\lambda =2$, $d=4$, the following table shows a set $E=\{c_1, c_2, c_3, c_4, c_5 \}$.

Symbol | Codeword |
---|---|

$c_1$ | 0 0 1 1 2 2 |

$c_2$ | 0 1 0 2 1 2 |

$c_3$ | 0 1 2 0 2 1 |

$c_4$ | 0 2 1 2 0 1 |

$c_5$ | 0 2 2 1 1 0 |