Next: Problem 4 - Operating
Up: Information Science I
Previous: Problem 2 - Digital
Consider an information source consisting of 5 characters , ,
, ,
with probability
,
,
,
,
respectively.
- 1.
- Find a Huffman code for this information source, and compute the
average code length, where
.
- 2.
- Compute the entropy of this information source
- 3.
- Describe relation between the entropy and the average code length.
- 1.
- The code words are:
The average code length is:
- 2.
- The entropy is:
- 3.
- The efficiency of the code is:
which means that only
of the bits are redundant.
We can also point out that:
Huffman code can deliver code word sequence that asymptotically approach
the entropy. Which means that for large source alphabet, the amount of
redundant bits is very small.
Reynald AFFELDT
2000-06-08