Method for common source and channel coding
View Patent ↗In a method, and a signal processing arrangement, for coding information, the information symbol sequences are first sorted in accordance with the probability of occurrence of the symbols in the sequences, and are then mapped to the natural binary code. This allows redundant information contained in the symbol sequences to be used for decoding the bit positions.
1. A method for coding information consisting of symbol sequences containing symbols which occur with different probabilities, comprising the steps of:
mapping said symbols to binary code words, each having a plurality of bit positions; and
in said mapping, sorting said symbols dependent on their respective probability of occurrence, and allocating a natural code words to said symbols to obtain sorted symbols, and allocating a natural binary code to said sorted symbols.
2. A method as claimed in claim 1 wherein the step of sorting said symbols comprises sorting a substantial proportion of said symbols, thereby obtaining a substantial proportion of sorted symbols, and comprising allocating said natural binary code to said substantial proportion of sorted symbols.
3. A method as claimed in claim 1 wherein the step of sorting said symbols comprises sorting all of said symbols, and allocating said natural binary code to all of said sorted symbols.
4. A method as claimed in claim 1 wherein the step of allocating said natural binary code comprises:
allocating a code word which exhibits a first binary value at all bit positions to a symbol which occurs most frequently; and
allocating a code word which exhibits a second binary value at all positions to a symbol occurring most infrequently.
5. A method as claimed in claim 1 comprising producing said symbol sequences from a source encoding.
6. A method as claimed in claim 1 comprising interchanging bit positions of code words obtained from said mapping.
7. A method as claimed in claim 1 wherein said symbol sequences contain redundant information, and comprising decoding said natural binary code using said redundant information as a priori information for determining respective values of said bit positions.
8. A method as claimed in claim 1 wherein said symbol sequences contain redundant information, and comprising decoding said natural binary code using said redundant information as a posteriori information for determining respective values of said bit positions.
9. A method as claimed in claim 1 wherein said bit positions of said code words contain redundant information, and comprising decoding said natural binary code using said redundant information as a priori information for determining respective values of said bit positions.
10. A method as claimed in claim 1 wherein said bit positions of said code words contain redundant information, and comprising decoding said natural binary code using said redundant information as a posteriori information for determining respective values of said bit positions.
11. A signal processing arrangement for coding information consisting of symbol sequences containing symbols which occur with different probabilities, comprising the steps of:
mapping said symbols to binary code words, each having a plurality of bit positions; and
in said mapping, sorting said symbols dependent on their respective probability of occurrence, and allocating a natural code words to said symbols to obtain sorted symbols, and allocating a natural binary code to said sorted symbols.
12. A signal processing arrangement as claimed in claim 11 wherein said symbol sequences contain redundant information, and comprising decoding said natural binary code using said redundant information as a priori information for determining respective values of said bit positions.