As data storage and processing increasingly shift to the cloud, the risk of data breaches also rises. One way to address this is using Homomorphic Encryption (HE), which allows for data processing while the data remains encrypted, unlike traditional methods. However, current HE libraries support only addition and multiplication, requiring users to implement other mathematical functions themselves. To this end, we developed and analyzed basic mathematical functions in a previous work. Since polynomial approximations are more common in HE, this paper expands on that by examining and comparing polynomial approximations of these functions with the previously implemented methods. Our findings indicate that while polynomial approximations offer the benefit of low multiplication depth, the previously implemented methods generally outperform them in most scenarios despite their higher computational cost.

In recent years, digital services have experienced significant growth, exemplified by platforms like Netflix achieving unprecedented revenue levels. Some of these services employ subscription models, with certain content requiring additional payments or offering third-party products. To ensure the widespread availability of diverse digital services anytime and anywhere, providers must have control over content accessibility. To address the multifaceted challenges in this domain, one promising solution is the adoption of attribute-based encryption (ABE). Over the years, various approaches have been proposed in the literature, offering a wide range of features. In a prior study [18], we assessed the security of one of these proposed approaches and identified one that did not meet its promised security standards. In this research we focuses on conducting a security analysis for another ABE scheme to pinpoint its shortcomings and emphasize the critical importance of evaluating the safety and effectiveness of newly proposed schemes. Specifically, we uncover an attack vector within this ABE scheme, which enables malicious users to decrypt content without the required permissions or attributes. Furthermore, we propose a solution to rectify this identified vulnerability.