On the Feasibility of Attribute-Based Encryption on Internet of Things Devices

Abstract

Attribute-Based Encryption (ABE) could be an effective cryptographic tool for the secure management of Internet-of-Things (IoT) devices, but its feasibility in the IoT has been under-investigated thus far. This article explores such feasibility for well-known IoT platforms, namely, Intel Galileo Gen 2, Intel Edison, Raspberry Pi 1 Model B, and Raspberry Pi Zero, and concludes that adopting ABE in the IoT is indeed feasible.Accepted for publication --IEEE Micro Special Issue on Internet of Things (2016) Preprint versionAttribute-Based Encryption and IoT. In recent years, several security protocols adopted Attribute-Based Encryption (ABE) as a building block in different distributed environments [3], such as IoT [4], cloud services [5], and medical systems [6]. ABE is a public key scheme where both encryption and decryption are based on high-level data access policies. Considering the aforementioned requirements in distributed and heterogeneous IoT scenarios, ABE provides more efficient access control mechanism compared to conventional cryptographic algorithms [3], [6], [7]: (i) allows fine-grained access control based on recipients' attributes; (ii) scales independent from the number of authorized users; (iii) is resilient against collusion attacks; (iv) does not require key sharing or key management algorithms between the participating parties (data owner does not need to identify the destination client). Besides its noteworthy advantages, a proper key revocation algorithm is still a challenging issue in ABE (beyond the scope of this paper), and an ongoing research effort [3]. More relevant to our work, ABE suffers from high computational overhead [6], [8]. However, the literature is still missing a proper assessment of ABE efficiency on resource-constrained devices, widely used in the IoT domain.In order to shine a light on the feasibility of ABE in IoT, we perform a comprehensive analysis of the cost of ABE operations on resource-constrained devices. In particular, along the same line of our previous work [7], which investigated the feasibility of ABE on smartphone devices, in this paper we implement the original Key-Policy Attribute-Based Encryption (KP-ABE) [9] and Ciphertext-Policy Attribute-Based Encryption (CP-ABE) [10] on widely used IoT-enabling devices. Our work focuses on the evaluation of encryption and decryption (hereinafter called cryptographic operations ) on four boards: Intel Galileo Gen 2, Intel Edison, Raspberry Pi 1 Model B, and Raspberry Pi Zero. Due to space limitation, we only report the results for CP-ABE. However, we noticed that the KP-ABE experiments have a very similar quantitative behavior to CP-ABE results. Supported by our observations from thorough experimental results, we provide evidence of the feasibility of adopting ABE on resource-constrained devices. Moreover, we present a smart healthcare use case application to evaluate feasibility of using ABE in real world IoT scenarios.

PRTA: A Proxy Re-encryption based Trusted Authorization scheme for nodes on CloudIoT

“…Recent advances in sensing technologies and smart chips have promoted the progress of IoT. Based on various sensors and devices, IoT could collect the information of different things communicating 5 with Internet. The communications by Internet are changing from computers to computers to Man-to-Machine or Machine-to-Machine(M2M).…”

“…The communications by Internet are changing from computers to computers to Man-to-Machine or Machine-to-Machine(M2M). In a word, IoT integrates various sensors, objects and smart nodes that are capable of communicating with each other without human intervention [5]. The development of IoT has been blurring the boundaries among the physical, social, and cy- 10 ber worlds and fueling the astonishing number of Internet-connected devices, which has been increasing from 15 billion in 2014 to 17.6 billion in 2016 and will be 30 billion by 2020 [9] [1].…”

“…ReKeyGen(rkpara, f lag) → rk, this algorithm generates the re-encryption 450 key based on rkpara, and outputs the re-encryption key rk; if f lag = "download",then rk = (pk i , pk r i , H 1 (pk i ∥ α) · H 1 (pk IoT ) sk IoT , g −r ). if f lag = "upload",then rk = (pk IoT , pk r IoT , H 1 (pk IoT ) · H 1 (pk i ) ski , g −r ) (6)ReEnc 1 (C IoT , rk) → C IoT →i ,this algorithm outputs a ciphertext C IoT →i based on re-encryption key, where 5 ), otherwise, outputs the error information. The ciphertext C IoT →i can be decrypted with sk i .…”

Information collection system of duck products based on IoT

“…At present, the structure of the Internet of Things is mainly divided into the following categories [17,18]: The USN-based five-tier structure model includes sensor networks, access networks, basic network facilities, middleware, and application platforms. The physicalnet's four-layer structure includes service provider network, gateway layer, coordination layer, and application layer.…”

On the feasibility of attribute-based encryption for WLAN access control

“…A user is able to decrypt a ciphertext, if a subset of his attributes satisfies the access policy specified on the ciphertext. Due to the characteristics of ABE, several researchers have concentrated on ABE and proposed either new variations of ABE (to mention a few [9], [14], [15]), or new applications and performance evaluation of ABE (to mention a few [16]- [18]). …”