Tran et al. (2005) proposed a power estimation model for digital

CMOS circuits. The circuit was divided into five sections and the power

dissipation of each part has been estimated individually. Further the

implemented gates were also counted, the proposed power investigation

model suggest early guidelines for design of circuits. The leakage power

estimation is most important factor the study of design feasibility.

Derakhshandeh et al. (2005) identified the relationship between the leakage

power and threshold voltage, where initially the number of input gates were

identified, followed by identifying the number of inputs to the gates and their

corresponding states. The predicted results on benchmark ICs reported

improved accuracy than the conventional methods of estimation.

Ligang et al. (2006) introduced neural network models for power

estimation for VLSI circuits. In the proposed study the authors used ISCAS89

benchmark circuits and the corresponding experimental results were noted.

The neural network based power estimation techniques produced better

results as compared to the conventional methods such as Monte-Carlo and

other statistical techniques. A linear programming based leakage power

estimation technique has been proposed by Chen et al. (2006), where Genetic

Algorithm was implemented for Minimum Leakage Vector (MLV) searching,

the leakage power is estimated on gate level by linear programming method.

The study assists in reducing the leakage power of VLSI circuits with easy

implementations.

Chaudhry et al. (2006) presented accurate power estimation

strategy by extracting the switching and clock activity of macro power

models. Do et al. (2007) discussed a high power estimation models for

proposed 2-kB 6T - SRAM array, in the proposed study the authors computed

the threshold leakage by combining the probing methodology. The memory