![](https://csdnimg.cn/release/download_crawler_static/15512626/bg3.jpg)
290 R. Zhou et al.
– We analyze the performance evaluation of our scheme. The evaluation result
shows that the scheme is practical for IoT applications.
1.1 Related Work
Since cloud-assisted technology is provided to improve IoT, the researchers are
focus on data management stored in cloud server server. Moreover, data privacy
becomes the concern in IoT data storage system. To ensure data security, the
data owner needs to encrypt the data before sending it to the cloud server. In
order to crease the quality of service, data manager needs to share IoT data
with authorized users for data analysis. Keyword searchable encryption, which
meets the requirement in data management, is proposed for secure data sharing.
Therefore, the design of lightweight keyword searchable encryption construction
is a challenge in IoT data sharing application.
To reduce the cost of massive key management in searchable symmetric
encryption construction [1], the concept of public key encryption with keyword
search (PEKS) is introduced by [2]. Moreover, multi-key searchable encryption
and forward secrecy searchable encryption are presented to complete the func-
tions of fine-grained document sharing and user right revocation in practical
system.
Multi-key Searchable Encryption. The multi-key searchable encryption
framework is designed by Popa et al. [5], and the first web application are built
on Mylar in [6]. Only one trapdoor is provided to server, and keyword search
match is completed by in different documents, which is encrypted by different
keys. Moreover, an new security model for multi-owner searchable encryption
areproposedbyTang[7] in this framework. Liu et al. [8] design a scheme in
data sharing. Unfortunately, low performance problem is inevitable, because the
trapdoor size is linear with the number of documents, in their scheme. To com-
plete provable security, Rompay et al. [9] constructed a new scheme on proxy
method. However, heavy overhead does not be ignored on proxy.
Combining with the study of [10], Cui et al. [11] proposed a new key-aggregate
searchable encryption scheme, where an aggregation method on file keys is intro-
duced to compute authorization key generation. One user generates a trapdoor to
complete keyword search in all authorized file set to this user. However, Kiayias
et al. [12] design two key guessing attacks to the study of [11]. Li et al. [14]and
Liu et al. [15] proposed their two improved schemes to maintain data verifica-
tion and multi-owner functions. However, in [14] and [15], the similar drawbacks
as in [11] can be found. Kiayias et al. [12] proposed their improved scheme for
the study of [11], and more communication and computation are inevitable. To
reduce the computation and communication cost, [13] designs a new scheme.
Forward Secrecy Searchable Encryption. In recent studies [16–18], to
achieve the function of fine-grained access control in searchable encryption, many
schemes, in different scenario, are proposed based on attribute-based encryp-
tion. The study [17] implements user revocation in a practical multi-user and
multi-owner scenario. In [18], a key-policy and a ciphertext-policy key searchable