Realization of reduced computation and power efficient DCT architecture for image compression
Design and characterization of null convention logic based floating point multiplier