مقاله انگلیسی رایگان در مورد پروتکل رمزنگاری موثر و امن برای اینترنت اشیای ابری – الزویر 2018

In an Internet of Things (IoT) architecture, there are many realworld objects (also referred to as devices or things) connected to the Internet. These interconnected objects (e.g. sensors, mobile devices such as Android and iOS devices, wearable devices, and drones or unmanned aerial vehicles) are responsible for sensing, collecting, disseminating and exchanging data in a broad range of context, such as public/homeland security (e.g. smart cities), utility (e.g. smart grids), logistics (e.g. smart supply chains), and intelligent building (e.g. smart homes). The trend of IoT in our modern society is explained in a recent report from Gartner, which estimated that 63 million IoT devices will be attempting to connect to the network each second by 2020 [33]. Cloud computing can also play a supporting role in IoT architecture, as explained by Roopaei, Rad and Choo [28]. The authors used the Cloud of Things-based automated irrigation as a use case to illustrate how integrating cloud and IoT can make energy use more efficient and less costly. This is not surprising, considering that cloud computing can provide affordable and real-time computing and storage capabilities [30]. A typical cloud-based IoT system is shown in Fig. 1. In a CoT environment, data collected by various smart devices is uploaded to the cloud server. Then, the user obtain information of interest from the cloud server via a client device. Since it is not realistic to expect that a cloud server is completely trustworthy (e.g. the server may be compromised and there may exists a corrupted or malicious insider) [14,18], data outsourced to or stored in the cloud should be protected (e.g. using a secure encryption scheme). However, searching on encrypted data is challenging given today’s technology. Thus, Searchable Encryption (SE) has emerged as a salient research inquiry. A SE protocol is designed to allow one to search on encrypted data containing specified keywords and to obtain the response from the server based on the keyword trapdoor without the need to decrypt the data. The server is also prevented from learning the content of the user’s query. A number of SE protocols/schemes have been proposed in the literature since the work of Song et al. [29], and these SE protocols can be broadly categorized into searchable symmetric encryption (SSE) protocols (see [6,9,15,19,23,29,31]) and public-key encryption with keyword search (PEKS) protocols (see [1–3,5,12,16,17,24,34–36]). It is known that the SSE protocols generally are more efficient, but suffer from complex and expensive key management distribution [10] limitations (due to the fact that data owner needs to share a key with each user by the secure channel [4]). We refer the interested reader to a recent review of SSE schemes by Poh et al. [26]. PEKS protocols are known for their stronger security and flexibility, but a key limitation with such protocols is the inability to resist inside keyword guessing attacks [7] (e.g. from a malicious server or cloud employee). Specifically, the cloud server can use the user’s public key to encrypt some keywords and use the keyword’s ciphertext to test the content of a trapdoor. Recently in 2016, Chen et al. [13] proposed a PEKS protocol to resist such an attack, but their protocol is insecure against an external adversary.