Secure Coordinate Rotation in Embedded Systems Using Homomorphic Encryption: Implementation and Evaluation on the AI-Saqr Platform

This paper explores the viability of a method for performing coordinate rotation on encrypted data within a drone's mission computer. Utilizing the Joye-Libert additively Homomorphic Encryption scheme, we implement the CORDIC (COordinate-Rotation DIgital Computer) algorithm, optimized for addition operations, on the AI-Saqr system-on-chip platform, a RISC-V open-source SoC. This approach ensures data confiden-tiality and computational efficiency in real-time embedded applications. Our experimental results demonstrate the feasibility and performance trade-offs of integrating homomorphic encryption with the CORDIC algorithm, offering significant advancements in secure computations for unmanned aerial systems. To the best of our knowledge, this work, is the first of it's kind to present a vector rotation based computation scheme for encrypted vectors.