International Journal on Advanced Science, Engineering and Information Technology, Vol. 9 (2019) No. 1, pages: 100-108, DOI:10.18517/ijaseit.9.1.7583

Performance Analysis of Loss Multilevel Quantization on the Secret Key Generation Scheme in Indoor Wireless Environment

Mike Yuliana, Wirawan Wirawan, Suwadi Suwadi

Abstract

The necessity for secured communication devices that has limited computing power has encouraged the development of key generation scheme. The generation of a symmetric key scheme that utilizes randomness of wireless channels offers a most promising solution as a result of the easy distribution of secret key mechanisms. In the last few years, various schemes have been proposed, but there are trade-offs between the performance parameters used. The expected parameters are the low Key Disagreement Rate (KDR), the high Key Generation Rate (KGR), and the fulfillment of standard of randomness. In this paper, we propose the use of a combination of pre-processing methods with multilevel lossy quantization to overcome the trade-off of performance parameters of the Secret Key Generation (SKG) scheme. Pre-process method used to improve reciprocity so as to reduce KDR, whereas multilevel quantization is used to improve the KGR. We use Kalman as the pre-processing method and Adaptive Quantization, Modified Multi-Bit (MMB), and 2-ary Quantization as the multilevel lossy quantization. Testing is conducted by comparing the performance between direct quantization with the addition of the pre-processing method in various multilevel lossy quantization schemes. The test results show that the use of Kalman as pre-processing methods and multilevel lossy quantization can overcome the trade-off performance parameters by reducing KDR and increasing KGR, with the best performance, was obtained when we use adaptive quantization. The resulting secret key has also fulfilled 6 random tests with p values greater than 0.01.

Keywords:

KDR; KGR; secret key generation; Kalman; multilevel lossy quantization.

Viewed: 256 times (since Sept 4, 2017)

cite this paper     download