| 摘要 |
PURPOSE:To provide the method for enabling multiplication to a fixed point, for which the sum of a bit length at an integral part and a bit length at a decimal part is made much longer, in the case of providing the fixed point multiplying method for a computer by using integral multiplication. CONSTITUTION:In the case of multiplying a multiplier and a multiplicand expressed by integral part L bits and decimal part M bits, a value (11) shifting the multiplier to the least significant side for M bits and a value (12) ignoring the decimal point position of the multiplicand are respectively defined as integers and a first intermediate result (13) multiplying those integers is provided. A value (21) shifting the multiplicand to the least significant side for M bits and a value (22) ignoring the decimal point position at the decimal part of the multiplier are respectively defined as integers, and second intermediate result (23) multiplying those integers is provided. Further, a fourth intermediate result (40) is provided by shifting a third intermediate result (33), for which decimal point positions of a decimal part (31) of the multiplier and a decimal part (32) of the multiplicand are ignored and respectively defined as integers and those integers are multiplied, to the least significant side for M bits. Then, a final multiplied result (51) of the multiplier and the multiplicand is provided by integrally adding the first, second and fourth intermediate results (13, 23 and 40). |
