Enhancing Privacy and Security in IoT Environments through Secure Multiparty Computation

Abstract

With the increasing influence of IoT devices in our daily lives, secure data-sharing is becoming ever more important. Sensors and other devices are communicating vast amounts of possibly unencrypted data, which poses a significant privacy concern. To tackle this problem, this research implements two Partially Homomorphic Encryption (PHE) schemes, RSA and the Paillier cryptosystem, to perform Secure Multiparty Computation (SMPC) in the resource-constrained IoT environment. The environment consists of a laptop connected to an Arduino Uno through a serial connection. The RSA-based SMPC protocol has an average completion time of 2007ms. However, due to the inability to use padding, RSA lacks semantic security. Conversely, the Paillier-based protocol is semantically secure but cannot complete the encryption due to dynamic memory issues. Even if resolved, the estimated encryption time exceeds 103.3 minutes. Despite the potential of SMPC in IoT environments for secure data handling, the results from this research suggest that directly implementing PHE schemes on Arduino is not practical based on the observed limitations.