RSS Signal Processing for Secret Key Generation: A Comparison of Kalman Filtering and Quantization in Indoor Environments
Rahmadani Najwa Alfriza, Mike Yuliana, Nailul Muna · 2025
Received Signal Strength (RSS) is a parameter that is widely used in physical layer security systems because of its reciprocal and spatially decorrelated nature. However, RSS values fluctuate, so they require pre-processing before they can be used for reliable secret key generation. This study evaluates the effectiveness of combining Kalman filtering with two quantization methods, Jana multi-bit and multi-bit quantization, on the difference in signal strength (DOSS) data obtained from indoor environments. Testing is carried out by comparing the performance of the Key Generation Rate (KGR) and Key Disagreement Rate (KDR) as indicators of bitstream quality. The experimental results show that the combination of Kalman filtering and multi-bit quantization produces the best performance with a KDR as low as 1% and a KGR of 3.67 kbit/s. In contrast, the Jana multi-bit method without Kalman produces a higher KGR but with a much larger KDR. This finding confirms that quantization strategies-especially multi-bit ones-are critical in improving RSS-based bit quality even before the key generation stage is entirely performed.