Contents

1 Ideas

• Used then the distribution of probability of the symbols follows a geometric distribution.
• It is a particular case of the Huffman code where the number of bits of each code-word (minus one) is equal to the index of the symbol in the source alphabet. Example:

• The unary coding is only optimal when (see Equation Eq:symbol_information)

  $$p\left(s\right)=2^{-\left(s+1\right)}$$

  (Eq:Unary)

  where $s=0,1,\cdots$.