Least-square estimation of average power in digital CMOS circuits

摘要

The estimation of average-power dissipation of a circuit throughnexhaustive simulation is impractical due to the large number of primaryninputs and their combinations. In this work, two algorithms based onnleast square estimation are proposed for determining the average powerndissipation in complementary metal-oxide-semiconductor (CMOS) circuits.nLeast square estimation converges faster by attempting to minimize thenmean square error value during each iteration. Two statisticalnapproaches namely, the sequential least square (SLS) estimation and thenrecursive least square estimation are investigated. The proposed methodsnare distribution independent in terms of the input samples, unbiased andnpoint estimation based. Experimental results presented for the MCNC'91nand the ISCAS'89 benchmark circuits show that the least squarenestimation algorithms converge faster than other statistical techniquesnsuch as the Monte Carlo method and the DIPE