Study on information entropy of combinatorics coding

Ideas of combinatorics coding and decoding are expatiated in this paper. It predicates that if only repetition exists,redundancy is inevitable. The relation between ordinal space and sequence space is studied. It opens out the characteristic that ordinal space is smaller than sequence space. By making use of this characteristic, combinatorics coding can be used to compress data. Combinatorics coding belongs to universal coding and is probability-independent. To compare with information entropy, average code length of combinatorics coding can be computed approximatively. It is proved that average code length of combinatorics coding can come to information entropy when the length of the section is long enough.