Publication Type : Journal Article
Publisher : Indian Journal of Science and Technology
Source : Indian Journal of Science and Technology, Volume 9, Issue 1 (2016)
Url : http://www.indjst.org/index.php/indjst/article/viewFile/107899/77630
Campus : Coimbatore
School : School of Engineering
Center : Amrita Innovation & Research
Department : Electronics and Communication
Verified : Yes
Year : 2016
Abstract : Present day electronic chips are outsourced by the manufacturer and done by third-parties. Such practices are widelyprevalent as it works out to be more profitable in terms of cost and effort to complete these processes offshore inVLSI foundries. This comes with a potential risk of compromised security due to hardware Trojans. There are existingmethodologies which aim at tacking the issue of Hardware Trojans by detecting it. Hardware Trojan detection bydecomposing the design into blocks to detect combinational Trojan and Sequential Trojan is proposed in this paper. Bycomparing the leakage and total power values consumed by both the design i.e. with and without the presence of thehardware Trojans, we can identify the presence of the hardware Trojans. Moreover if the measurement noise gets high itmasks or hides the effect of the variation in power profile, which leads to a wrong decision. So by taking the measurementnoise under consideration the classification is effectively done for magnifying the differences by using Principle ComponentAnalysis approach and tested with 8, 16 and 32 bit ALU designs.
Cite this Research Publication : N Mohankumar and NirmalaDevi, M., “Improving the Classification Accuracy in Detecting Hardware Trojan in ALU Using PCA”, Indian Journal of Science and Technology, vol. 9, no. 1, 2016.