Secure Estimation Using Partially Homomorphic Encryption for Unmanned Aerial Systems in the Presence of Eavesdroppers

Unmanned aerial systems (UASs) are attracting increasing attention thanks to the great mobility and flexibility of unmanned aerial vehicles (UAVs). This paper considers a typical UAS, which consists of a UAV, a sensing device that provides some sensed data to the UAV, and an end-user that operates the UAV. However, the information exchanged between these parties is vulnerable to eavesdropping attacks, emphasizing the need to develop privacy-preserving approaches. The cryptographic methods are undoubtedly effective, but their high computational overhead may adversely impact the normal operations of UASs. Additionally, the dynamic of a UAV has a high dimension, which is disadvantageous for both estimation and encryption. Therefore, this paper proposes a secure distributed estimation protocol with partially homomorphic encryption by encrypting the transmitted measurements and estimates. Attribute to distributed structure and partial homomorphism, the computation amount for secure estimation is greatly reduced. At the same time, the raw data that needs to be encrypted is transferred into the space of plaintexts by a uniform quantizer and a mapping strategy. Finally, the effectiveness of the proposed method is verified by computer simulation.