Physical layer secret key can be extracted from wireless fading channels based on channel quantization with guardband (CQG). In this paper, we analyze the key generation efficiency of the CQG protocol. Analytical closed-form expressions of bit error rate (BER) and key generation efficiency are derived for the CQG protocol. In order to further improve the key generation efficiency, we concatenate the CQG protocol with reconciliation. And then, we present a upper bound of key generation efficiency, which can be maximized by careful selecting the guardband region and LDPC codes. The key generation efficiency is also analyzed by considering the attacks in the case of an unauthenticated wireless channel.
The adaptive fusion estimation problem was studied for the multi-sensor nonlinear under-observed systems with multiplicative noise. A one-step predictor with state update equations was designed for the virtual state with virtual noise first of all. An extended incremental Kalman filter (EIKF) was then proposed for the nonlinear under-observed systems. Furthermore, an adaptive filtering method was given for optimization. The fusion adaptive incremental Kalman filter weighted by scalar was finally proposed. The comparison analysis was made to verify the optimization of the state estimation using adaptive filtering method in the filtering process.