Title :
Low-power arithmetic for the processing of video signals
Author_Institution :
Liesegang Electron., Hannover, Germany
Abstract :
In this paper, it is shown that power reduction of complementary metal-oxide-semiconductor integrated circuits (CMOS IC´s) is possible by exploiting statistical properties of the input signals. The statistical properties of video signals are analyzed, and it is shown that a new low-power signed-digit representation and an offset representation can reduce the number of signal changes for the input data compared to binary representation. The increased effort for data processing has been taken into account and it is shown for which types of video signals the alternative representations are advantageous. Circuit simulations have been performed to verify the theoretical results.
Keywords :
CMOS logic circuits; digital arithmetic; statistical analysis; video signal processing; CMOS; low-power arithmetic; low-power signed-digit representation; power reduction; statistical properties; video signal processing; Arithmetic; Band pass filters; CMOS integrated circuits; Data processing; Gaussian distribution; Image analysis; Power dissipation; Signal analysis; Signal processing; Testing;
Journal_Title :
Very Large Scale Integration (VLSI) Systems, IEEE Transactions on