An Adaptive Information Reconciliation Protocol for Physical-Layer Based Secret Key Generation

Physical-layer based secret key generation (SKG) becomes a research focus as it solves key distribution problem which is difficult in traditional cryptographic mechanism. To remove the disagreements caused by imperfect reciprocity of communicating parties, information reconciliation is a critical process in SKG to obtain symmetric keys. Various reconciliation schemes are proposed, e.g. BBBSS, Cascade, BCH code and Turbo code. Reconciliation efficiency is a common evaluation index which takes reconciliation success rate and information leakage rate into account. However, time delay caused by information interaction and computation overhead may affect reconciliation performance under specific scenarios. Therefore, a comprehensive evaluation metric is required to compare existing schemes. Besides, channel condition changes all the time in real mobile communication systems, and most existing reconciliation schemes only work well in certain channel conditions. Hence, a reconciliation scheme which can adapt to time-varying channel condition is required. In this paper, we introduce a novel comprehensive reconciliation efficiency index (CREI) to evaluate existing reconciliation schemes and propose an adaptive information reconciliation scheme selection (AIRSS) protocol to maximize CREI. The simulation results show the superiority of AIRSS and present recommendations of reconciliation scheme selection in different scenarios.