Analysis and Design of Low Area and Highly Energy Efficient Hybrid Adder for Signal Processing Applications

Mammogram imaging provides very useful support for the radiologist in detecting and treating the breast cancer. All the detection methods need pre-processing support to make the image clear and free from any unwanted information. Filters with high accuracy are the major requirement for all pre-processing methods. Adders are the main building blocks used in the filter design. A new Quality Confirmed Approach (QCA) adder has been proposed by combining the existing Brent Kung, Sklansky and Kogge Stone adder logic by using Tree Grafting Technique (TGT) for improvement in speed, reduction in complexity and power consumption. The proposed new adder performs well in the Modified Low Range Modification (MLRM) filter, which is used for the effective pre-processing of mammogram image towards the detection of breast cancer. The existing and proposed adder based MLRM method has been tested for Power reduction, Power Delay Product (PDP) and accuracy. The proposed QCA adder based MLRM performed well and have consumed 891.842 µW power with 0.21 % of power saving over Brent Kung adder based approach, achieved the PDP value of 16.613 pJ, which is 0.81 % less than that of the Han Carlson Adder based approach. The existing and proposed MLRM methods have been tested for contrast improvement, mean square error (MSE) reduction and peak signal to noise ratio (PSNR) improvement. For the test image mdb072, 7.4 % improvement achieved in contrast percentage than the next best BKA based approach.