Date of Award

Summer 2022

Document Type


Degree Name

Master of Science (MS)


Electrical and Computer Engineering

First Advisor

Ababei, Cristinel

Second Advisor

Povinelli, Richard

Third Advisor

Gao, Jie


This thesis investigates the efficacy of side channel analysis based attacks to recover the encryption key of unmanned aerial vehicles (UAV) which have become increasingly important internet of things devices in both the public and military sectors. Additionally, this thesis proposes a method based on noise injection to safeguard against side channel analysis based attacks. The proposed method is unique because it is ultra-low power and uses as a noise source the brushless motors of the unmanned aerial vehicle. Verification through experimentation on an in-house build UAV system is presented. Firstly, side channel analysis techniques are overviewed and the correlation power analysis (CPA) algorithm along with the key components of the advanced encryption standard (AES) algorithm are detailed. An unprotected UAV system is attacked with CPA to demonstrate the ability to easily recover the AES key from the UAV. Once the vulnerability is proven, noise injection methods using the noise generated by any of the brushless motors on the UAV is detailed. To ensure the proposed noise injection method is effective, a mathematical study for randomness is completed on the noise generated by the motors to confirm it as a good source of entropy. Then, the noise injection hardware is implemented on the experimental UAV. Extensive experiments are conducted on the stability of the flight control algorithm and power consumption to ensure flight dynamics nor flight time are negatively affected. Finally, the implemented noise injection method is tested through the same experiments that were conducted earlier during the initial key theft procedure to demonstrate the efficacy of the proposed method in creating a UAV system that is resilient to side channel based attacks such as correlation power analysis.

Included in

Engineering Commons