Efficient and secure secret sharing-based data outsourcing suitable for Internet of Things environments

Advancements in the Internet of Things (IoT) environment and cloud computing have created new business opportunities, such as cloud-based IoT data outsourcing, in which data that are collected by IoT devices are outsourced to cloud servers for further processing. However, concerns exist regarding the privacy and security of the collected data. Moreover, IoT devices often operate at low power and possess restricted memory, processing capabilities, and storage capacities. Consequently, a secure data-outsourcing method with minimal computational requirements is essential. Secret sharing schemes are recognized for providing robust security while requiring minimal computational resources. Hence, secret sharing schemes have recently been utilized as alternative methods to address privacy protection concerns. In a perfect $(k,n)$ threshold secret sharing scheme, a secret is converted into $n$ different shares, thereby requiring a total share size of $n$ times the secret. A $(k,L,n)$ ramp secret sharing scheme provides better storage efficiency, albeit at the cost of realizing only weak security. In this study, we propose a new protocol for a secure and storage-efficient $(k,L,n)$ ramp secret sharing scheme that incorporates the idea of encrypting each secret with a random number. By introducing new randomization steps based on the one-time pad encryption approach to randomize the inputs, we prove that even if part of the information can be leaked because of the nature of $(k,L,n)$ ramp secret sharing, the actual secret remains secure, provided that the product of the random numbers is not leaked from the unauthorized shares. Furthermore, we performed an experimental evaluation of the distribution phase of the proposed method with C++ using a Raspberry Pi 4 Model B as an IoT device and showed that the proposed method can be executed in significantly less time than most conventional secret haring schemes, particularly when the parameter $k$ increases.