Method and apparatus for performing a multiplication or division operation in an electronic circuit
View Patent ↗A multiplication or division operation X·K or X·1/K is performed in an electronic circuit. A software circuit area of the circuit calculates a digit shift sv such that p sv is an approximate value for K. In a hardware circuit area, the value X is shifted sv digits to the left in the case of multiplication or sv digits to the right in the case of division. The software circuit area calculates a suitable correction factor Kf. The value X is multiplied by the correction factor Kf.
1. A method for calculating a final value, which depends on the result of a multiplication operation X·K or a division operation X/K, wherein X and K are real numbers in a numerical representation to a base p, comprising:
calculating a digit shift sv in a software circuit area of an electronic circuit wherein p sv is an approximate value for shifting X to the left by the digit shift sv for multiplication or to the right by the digit shift sv for division within a hardware circuit area of the electronic circuit to obtain a shifted value of X;
processing the shifted value of X in a hardware block to obtain a value derived from the shifted value of X;
calculating a correction factor Kf in the software circuit area on the basis of K and p sv ; and
multiplying the value derived from the shifted value of X by the correction factor Kf to obtain the final value.
2. The method of claim 1 , wherein the digit shift sv equals round(log p K).
3. The method of claim 1 , wherein the correction factor Kf is about equal to (p sv )/K for division and K/p sv for multiplication.
4. The method of claim 1 , wherein X changes at a greater rate than K.
5. The method of claim 1 , wherein multiplying the value derived from the shifted value of X by the correction factor Kf is performed in the software circuit area.
6. The method of claim 1 , further comprising saturating the shifted value of X according to a defined threshold value to prevent overflows and reject unnecessary least significant bits.
7. The method of claim 1 , wherein shifting X is performed by a barrel shifter.
8. The method of claim 1 , wherein the final value is power of a signal in a mobile radio network.
9. The method of claim 1 , wherein p equals 2.
10. The method of claim 1 , wherein a non-linear operation is used for processing the shifted value of X in the hardware block to obtain the value derived from the shifted value of X.
11. An electronic circuit for calculating a final value, which depends on the result of a multiplication operation X·K or a division operation X/K, wherein X and K are real numbers in a numerical representation to a base p, comprising:
a software circuit area;
a hardware circuit area coupled to the software circuit area;
a first calculation mechanism located in the software circuit area for calculating a digit shift sv wherein p sv is an approximate value for a controllable digit shifter located within the hardware circuit area for shifting X to the left by the digit shift sv for multiplication or to the right by the digit shift sv for division to obtain a shifted value of X;
a hardware block located within the hardware circuit area for processing the shifted value of X and outputting a value derived from the shifted value of X;
a second calculation mechanism located within the software circuit area for calculating a correction factor Kf on the basis of K and p sv ; and
a multiplication mechanism for multiplying the value derived from the shifted value of X by the correction factor Kf to obtain the final value.
12. The electronic circuit of claim 11 , wherein the digit shift sv equals round(log p K).
13. The electronic circuit of claim 11 , wherein the correction factor Kf is about equal to (p sv )/K for division and K/p sv for multiplication.
14. The electronic circuit of claim 11 , wherein X changes at a greater rate than K.
15. The electronic circuit of claim 11 , wherein the multiplication mechanism is located in the software circuit area.
16. The electronic circuit of claim 11 , further comprising a saturation stage for saturating the shifted value of X according to a defined threshold value to prevent overflows and reject unnecessary least significant bits.
17. The electronic circuit of claim 11 , wherein the controllable digit shifter is a barrel shifter.
18. The electronic circuit of claim 11 , wherein the final value is power of a signal in a mobile radio network.
19. The electronic circuit of claim 11 , wherein the final value is a DPCH power calculation.
20. The electronic circuit of claim 11 , wherein the final value is a noise power calculation.
21. The electronic circuit of claim 11 , wherein the hardware block uses a non-linear operation for processing the shifted value of X to obtain the value derived from the shifted value of X.