Digital Signal Processing for Multimedia Systems

Flipping structure: an efficient VLSI architecture for lifting-based discrete wavelet transform

“…All noise signals can be assumed zero mean, uniform distributed, and independent. The variance is dependent on how many bits are quantized [21]. Nonetheless, it is not certain whether the original lifting-based architecture or the corresponding flipping structure has less roundoff noise.…”

“…Both the scaling effect and roundoff noise will be changed entirely. Although a systematic method using state variable description to solve these problems is presented in [21], only the calculation of roundoff noise can be adopted here because the method of scaling is only suitable for single input systems and possible to change the critical path.…”

“…8(a), and it is equivalent to performing quantization only at the outputs of the adders on the accumula- tive path. Thus, the roundoff noises of two output signals for each architecture can be calculated by the method with state variable description and matrix computation [21].…”

Low-Power Multipliers with Data Wordlength Reduction

“…Each bit in the data has equal probability of bit switching such as zero to one or one to zero, when new input data are given in previous data locations. The probability of the switching of each bit is 1 2 . The switching probability in X has binomial distribution:…”

“…Many methods have been developed to reduce power consumption. Lowering the supply voltage and minimizing the hardware are used for low-power hardware [1]. Changing the instruction order and reducing the number of operations are used for lowpower software [2].…”