| 摘要 |
PURPOSE:To shorten the time required for the scaling by storing the scale value in the memory. CONSTITUTION:The scaling memory MSR has a plurality of memories arranged in a matrix shape. The X axis corresponds to the respective increment value SIGMADELTAYi obtained every division quadrature, the Y axis corresponds to the respective value of the scale value and the product of the values of the X axis and the Y axis is stored on an intersecting portion thereof. Accordingly, when the increment SIGMADELTAYi and the scale value S are given, the product SSIGMAYi is immediately obtained and the value is outputted to the second adder AD2. According to this construction, the operation becomes unnecessary in the scaling and the extremely high speed scaling can be done. |
