International Electronics Symposium (IES)
Enhancing Key Generation Rate in LoRa-Based V2I Secret Key Generation Using Segment-Based Privacy Amplification
Muhammad Nizamudin Mahfud, Mike Yuliana, Weny Mistarika Rahmawati, Reni Soelistijorini, Nailul Muna
Abstract
Physical layer secret key generation (SKG) provides a promising solution for securing vehicle-toinfrastructure (V2I) communications by exploiting the inherent reciprocity of wireless channels. However, received signal strength indicator (RSSI) fluctuations, quantization mismatches, and constrained key generation rates present significant challenges in highly dynamic LoRa-based environments. This paper proposes an SKG framework that integrates Kalman filter, Guard Band Multibit Quantization (GBMQ), Cascade reconciliation, and SHA-1 based privacy amplification. Experimental evaluations were conducted across multiple urban V2I operational scenarios. The results demonstrate that the Kalman filter effectively enhances RSSI correlation, while the GBMQ scheme substantially minimizes the key disagreement rate (KDR) compared to the conventional Adaptive Multibit Quantization method. Furthermore, the remaining bit mismatches are completely eliminated via Cascade reconciliation, achieving a 0% KDR in all evaluated environments. Finally, the privacy amplification stage successfully generates secure cryptographic keys that exhibit robust randomness properties, confirming the efficacy of the proposed framework for resilient physical layer security in V2I networks.