Performance Evaluation of Secret Key Generation System for Static and Dynamic Condition

Received signal strength as one of the physical layer characteristics has been proven to be extracted as a secret key to secure communications carried out by two legitimate parties. Some of the previous secret key generation systems had low key generation speed and were indicated by the key generation value parameter and high bit difference indicated by the key disagreement value parameter. This paper proposes a system that can improve the similarity of RSS between legitimate parties by adding the Kalman Filter pre-process method. The increase in the similarity of received signal strength will be followed by a decrease in the difference in bits obtained so that it will reduce the number of bits wasted in the error-correcting component and will automatically increase the key generation value. The performance evaluation of the built system will be tested in an indoor environment with several scenarios and conditions. There are 2 scenarios used, namely the scenario in the classroom and laboratory and each of them has 2 conditions, namely static and dynamic. The test results show that the best performance is obtained when the Cl scenario and static condition with the key disagreement value of 0.41 and key generation value of 6.48 bps. The proposed system also shows better performance in terms of key disagreement value and key generation value compared to the previous system.