Comparative evaluation of pairing-free and pairing-based CP-ABE schemes for resource constrained environments
摘要
Pairing-free ciphertext-policy attribute-based encryption (PF-CP-ABE) schemes rely on either elliptic curve cryptography (ECC)-based public key cryptosystems (PKCs) or conventional PKC for their underlying cryptographic construction. On the other hand, pairing-based ciphertext-policy attribute-based encryption (PB-CP-ABE) schemes utilize bilinear pairings. The selection of a suitable CP-ABE scheme, whether PF-CP-ABE or PB-CP-ABE, significantly impacts a device’s lifespan and performance in resource constrained environments (RCEs), improves processing power, storage, battery life, and communication bandwidth. To enhance the lifespan and performance of the resource constrained devices, a lightweight and efficient CP-ABE scheme is essential. To the best of our knowledge, this article is the first to present a comparative evaluation of 23 selected CP-ABE schemes (12 PF-CP-ABE and 11 PB-CP-ABE) for RCEs. The selection of attributes in CP-ABE schemes affects the performance of communication parameters (i.e., size of ciphertext, public key, and secret key) and computational parameters (i.e., encryption and decryption costs). In this article, we discuss three scenarios related to attribute requirements of an application or environment, namely an attribute set of 0 to 20 attributes as scenario 1, an attribute set of 80 to 100 attributes as scenario 2, and a dynamic attribute set of any number of attributes from 0 to 100 as scenario 3. In addition, quantitative analysis and the comprehensive overview of all selected PB-CP-ABE and PF-CP-ABE schemes are provided for each parameter related to communication and computation. The study considers CP-ABE schemes to be most suitable if and only if the scheme performs well across all parameters. To the best of our knowledge, it appears that the DLL18, DN23, and YZDZ24 schemes are suitable for scenario 1, while the ODG15 and OD16 schemes are suitable for both scenario 2 and scenario 3.