A Lightweight and Anonymous Application-Aware Authentication and Key Agreement Protocol for the Internet of Drones

The drone technology has continuously been evolving since the beginning of the first decade of the 21st century with exceptional growth over the last several years. To pave the way for an interoperable aerial-ground communication platform, the Internet of Drones (IoD) framework has emerged to systematically organize a batch of drones to collect multiple application-specific data simultaneously and report them to a close ground station. As the collected data might contain sensitive information, people become more critically aware of data security and privacy issues associated with IoD applications. Authentication and key agreement protocols are able to protect IoD data from unauthorized access. However, the recent schemes fail to distinguish between types of data during the authentication and key establishment process, which leads to data leakage that sensitive data are being accessed by unauthorized entities. To address the data leakage issue and fill the research gap, this paper proposes a lightweight and anonymous application-aware authentication and key agreement protocol (also called liteA4) for IoD systems. The fundamental idea of liteA4 is that the ground station and the drone perform data type-aware mutual authentication and establish separate session keys for different types of data before the drone delivers the collected data to the ground station. The major techniques such as hash function, bitwise XOR, and physical unclonable function (PUF) are used to implement liteA4. We select the Automated Validation of Internet Security Protocols and Applications (AVISPA) tool to verify the security of liteA4 in the cyber-threat environment. We also set up a simulation framework and conduct comprehensive and comparative experiments to validate the performance of liteA4. Extensive experimental results demonstrate that liteA4 not only is a safe and reliable protocol in the adversarial setting, but also provides better results than its counterpart approaches in terms of communication overhead, computational time, storage cost, as well as energy consumption.