Linear-to-log converter for power estimation in a wireless data network receiver
View Patent ↗A converter to convert an N-bit input in a linear scale to an M-bit output in a logarithmic scale. The converter includes a set of K subrange converters each coupled to a respective number of bits of the N-bit input that represents a subrange of the N-bit input, each subrange converter to convert the subrange into a respective output that would be the M-bit output if the most significant set bit in the N-bit input was in the subrange. The converter further includes a range selector having the N-bit input as an input, to indicate which of the subranges to select for an output, and a selector having as inputs the outputs of the subrange converters and coupled to the range selector to select the output of one of the subrange converters as the M-bit output. In one embodiment, the converters operate in parallel.
1. A method of converting an N-bit input in a linear scale to an M-bit output in a logarithmic scale comprising:
dividing the input range into a set of K subranges each defined by a number of bits of the N-bit input;
converting each of the subranges into a respective converted output such that, for each subrange, the subrange's converted output is the M-bit output of the method for the case that the most significant bit in the N-bit input that is set is in the subrange;
determining from the N-bit input an indication of which of the subranges to select for an output; and
selecting the output of the converting step for the selected output as the M-bit output.
2. A method of converting as recited in claim 1 , wherein the subranges are overlapping.
3. A method of converting as recited in claim 2 , wherein the converting steps of the subranges occur in parallel.
4. A method of converting as recited in claim 3 , wherein the determining step occurs in parallel with the converting steps.
5. A method of converting as recited in claim 2 , wherein N is 16 and wherein there are 5-subranges, the four lowest order subranges comprising 5-bits of the input, and the highest subrange including 4 bits of the input.
6. A method of converting as recited in claim 2 , wherein the N-bit input is an average power measurement determined from a set of samples received in a radio receiver.
7. A method of converting as recited in claim 6 , wherein the M-bit output is for indexing a lookup table to set the gains of the radio receiver.
8. A converter to convert an N-bit input in a linear scale to an M-bit output in a logarithmic scale comprising:
a set of K converting means each for converting one of K sets of bits of the N-bit input, each set of bits defining a subrange of the N-bit input, each converting means for converting its respective subrange into a respective converted output such that, for each subrange, the subrange converting means' converted output is the M-bit output for the case that the most significant bit in the N-bit input that is set is in the subrange;
determining means having the N-bit input as an input, the determining means for indicating which of the subranges to select for an output; and
selecting means connected to the set of determining means, the selecting means for selecting the output of the converting means for the selected output as the M-bit output.
9. A converter as recited in claim 8 , wherein the subranges are overlapping.
10. A converter as recited in claim 9 , wherein the converting means carry out the converting in parallel.
11. A converter as recited in claim 10 , wherein the indicating means indicates in parallel with the converting of the converting means.
12. A converter as recited in claim 9 , wherein N is 16 and wherein there are 5-subranges, the four lowest order subranges comprising 5-bits of the input, and the highest subrange including 4 bits of the input.
13. A converter as recited in claim 9 , wherein the N-bit input is an average power measurement determined from a set of samples received in a radio receiver.
14. A converter as recited in claim 13 , wherein the M-bit output is for indexing a lookup table to set the gains of the radio receiver.
15. A converter to convert an N-bit input in a linear scale to an M-bit output in a logarithmic scale comprising:
a set of K subrange converters each coupled to a respective number of bits of the N-bit input that represents a subrange of the N-bit input, each subrange converter to convert the subrange into a respective converted output such that, for each subrange, the subrange converters converted output is the M-bit output for the case that the most significant bit in the N-bit input that is set is in the subrange;
a range selector having the N-bit input as an input, to indicate which of the subranges to select for an output; and
a selector having as inputs the outputs of the subrange converters and coupled to the range selector to select the output of one of the subrange converters as the M-bit output.
16. A converter as recited in claim 15 , wherein the subranges are overlapping.
17. A converter as recited in claim 16 , wherein the subrange converters carry out the converting in parallel.
18. A converter as recited in claim 17 , wherein the range selector indicates in parallel with the converting of the subrange converters.
19. A converter as recited in claim 16 , wherein N is 16 and wherein there are 5-subranges, the four lowest order subranges comprising 5-bits of the input, and the highest subrange includng 4 bits of the input.
20. A converter as recited in claim 16 , wherein the N-bit input is an average power measurement determined from a set of samples received in a radio receiver.
21. A converter as recited in claim 20 , wherein the M-bit output is for indexing a lookup table to set the gains of the radio receiver.