Cube-embedding based state encoding for low power design

We consider the problem of minimizing power consumption of a sequential circuit using low power state encoding. One of the previously published results is based on recursive matching. In general, a matched pair can be considered as a 1-cube being embedded in a hypercube. We generalize this idea of 1-cube embedding and propose a new encoding algorithm based on r-cube embedding. We then present an efficient 2-cube embedding based state encoding approach for low power design. It considers both Hamming distance and the complexity of the logic function (by estimation). Experimental results show that this approach is competitive to other existing techniques.