Compression using mu-law approximation
Techniques and mechanisms provide a technique for compression using an approximation of a mu-law algorithm.
1. A compression circuit to receive input data and generate compressed data based on the received input data, the compression circuit to generate the compressed data to have a first portion and a second portion, the first portion having a first value representing a value range of the received input data, the second portion having a second value associated with a bit shifting operation of the received input data, the bit shifting operation based on the first value of the first portion.
2. The compression circuit of claim 1 , wherein the first portion of the compressed data represents an exponent and the second portion of the compressed data represents a mantissa.
3. The compression circuit of claim 1 , wherein a number of bits shifted in the bit shifting operation is based on the first value of the first portion.
4. The compression circuit of claim 1 , wherein the compression circuit is further to add an offset to the bit shifted received input data.
5. The compression circuit of claim 4 , wherein the offset is based on the first value of the first portion.
6. The compression circuit of claim 1 , wherein the compression circuit is further to generate a positive representation and a negative representation of the compressed data.
7. The compression circuit of claim 6 , wherein the compression circuit is further to select between the positive representation and the negative representation of the compressed data based on the received input data.
8. A circuit receiving input data and providing compressed data, the circuit comprising:
a first logic unit to determine a first portion of the compressed data representing a value range of the input data; and
a second logic unit to determine a second portion of the compressed data based on a value of the first portion.
9. The circuit of claim 8 , wherein the first portion represents an exponent of the compressed data and the second portion represents a mantissa of the compressed data.
10. The circuit of claim 8 , wherein the second portion of the compressed data is based on the input data being shifted by a shift register by a number of bits based on the value range of the input data.
11. The circuit of claim 10 , wherein the second portion of the compressed data includes an offset based on the value range of the input data.
12. The circuit of claim 11 , wherein the shifted input data is summed with the offset to provide the second portion.
13. The circuit of claim 10 , wherein different value ranges of the input data indicate a different number of bits of the input data to be shifted.
14. The circuit of claim 13 , wherein different value ranges of the input data indicate different offsets to be summed with the shifted input data.
15. A method for receiving input data and approximating a compression algorithm to provide compressed data, the method comprising:
generating exponent data as a first portion of the compressed data representing a value range of the input data;
generating mantissa data as a second portion of the compressed data based on a value of the first portion; and
providing the first portion and the second portion as the compressed data.
16. The method of claim 15 , wherein the second portion of the compressed data is based on the input data being shifted by a shift register by a number of bits based on the value range of the input data.
17. The method of claim 16 , wherein different value ranges of the input data indicate a different number of bits of the input data to be shifted.
18. The method of claim 17 , wherein different value ranges of the input data indicate different offsets to be summed with the shifted input data.
19. The method of claim 15 , wherein the second portion of the compressed data includes an offset based on the value range of the input data.
20. The method of claim 19 , wherein the shifted input data is summed with the offset to provide the second portion.