Watermarking-Aided dimensionality reduction encryption for distributed fusion estimation against eavesdropping

Abstract

The existing defense strategies offer robust protection against eavesdropping attacks. However, in practical engineering scenarios, the system must navigate numerous challenges, particularly when communication resources are limited. Under such constraints, the estimator performance can suffer significantly, especially when faced with both eavesdropping attacks and resource limitations. This decline is due to the lack of effective adaptation mechanisms. To remedy this issue, we propose an encryption method that incorporates watermarking signals and dimensionality reduction techniques. The core principle of this method aims to find a balanced compromise between the performance demands and the resource constraints. Specifically, dimensionality reduction strategy is applied to reduce the information density of the transmitted signal, which alleviates the burden on network bandwidth. Additionally, the incorporation of the watermarking signal not only enhances the complexity of the signal but also maximizes the estimation error for potential eavesdroppers, which ensures the secure transmission of information. A compensation mechanism, which incorporates techniques for watermarking removal and prediction, has been specifically developed for users at the fusion center. This mechanism undergoes rigorous feasibility and boundedness analyses to ensure the robustness of the encryption strategy and estimator. Extensive simulation experiments conduct to validate the effectiveness and superiority of the developed algorithm.