Zoom device for residue number system
A remainder system and scaling technology, which is applied in the field of signal processing, can solve the problems of simple scaling operation, etc., and achieve the effect of reducing the delay of the critical path and having a simple and clear structure
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Publication Date
- 2011-10-12
- Estimated Expiration
- Not applicable · inactive patent
Smart Images
Figure 1 Figure 2 Figure 3
Abstract
Description
technical field
[0001] The present invention relates to the field of signal processing, in particular to the implementation structure of a numerical scaling circuit based on the remainder system (RNS) used in communication and signal processing. Background technique
[0002] In the process of numerical calculation, a large number of multiplication and addition operations will cause the expansion of the dynamic range of data, resulting in the occurrence of overflow. Numerical scaling is one of the most common methods to solve arithmetic overflow. In the two's complement system, the truncation operation is usually used to perform 2 n Scaling, does not require any additional resources when implemented in VLSI. Similarly, in the application of the remainder system, for those multiplication-intensive digital signal processing systems, overflow will inevitably occur. Therefore, the scaling operation can also be used to suppress the overflow of the operation, so that the operati...
Examples
Embodiment Construction
[0019] figure 1 shows that the basis is {2 n ,2 n -1, 2 n-1 The remainder of -1} is 2 of the system n Scale the basic implementation structure. Number of RNS (x 1 , x 2 , x 3 ) if the corresponding decimal number is a negative number, then perform 2 according to formula (1) n Scale; otherwise, scale according to formula (2). From the point of view of mathematical expressions, the only difference between formula (1) and formula (2) is: when calculating y 1 , formula (1) has one more term than formula (2) (2 n-1 -1). 101 realizes the symbol detection of the RNS number, and 102 is used to calculate y 3 , 103 is used to calculate y 2 . the y 1 The calculation needs to use y 2 and y 3 information, and the results of symbol detection. Modules 104-108 are used to calculate y 1 .
[0020] y 1 = y ...