Electronic medical records (EMRs) are extremely important for patients’ treatment, doctors’ diagnoses, and medical technology development. In recent years, the distributed healthcare blockchain system has been researched for solving the information isolated island problem in centralized healthcare service systems. However, there still exists a series of important problems such as the patients’ sensitive information security, cross-institutional data sharing, medical quality, and efficiency. In this paper, we establish a lightweight privacy-preserving mechanism for a healthcare blockchain system. First, we apply an interleaving encoder to encrypt the original EMRs. This can hide the sensitive information of EMRs to protect the patient’s privacy security. Second, a ( $t,n$ )-threshold lightweight message sharing scheme is presented. The EMRs are mapped to n different short shares, and it can be reconstructed by at least t shares. The EMR shares rather than the original EMRs are stored in the blockchain nodes. This can guarantee high security for EMR sharing and improve the data reconstruction efficiency. Third, the indexes of the stored EMR shares are employed to generate blocks that are chained together and finally form a blockchain. The authorized data users or institutions can recover an EMR by requesting at least t shares of the EMR from the blockchain nodes. In this way, the healthcare blockchain system can not only facilitate the cross-institution sharing process, but also provide proper protections for the EMRs. The security proof and analysis indicate that the proposed scheme can protect the privacy and security of patients’ medical information. The simulation results show that our proposed scheme is more efficient than similar literature in terms of energy consumption and storage space, and the healthcare blockchain system is more stable with the proposed message sharing scheme.

An adversary can deploy parasitic sensor nodes into wireless sensor networks to collect radio traffic distribu-tions and trace back messages to their source nodes. Then, he can locate the monitored targets around the source nodes with a high probability. In this paper, a Source-location pri-vacy Protection scheme based on Anonymity Cloud (SPAC) is proposed. We first design a light-weight (t,n) -threshold message sharing scheme and map the original message to a set of message shares which are shorter in length and can be processed and delivered with minimal energy consumption. Based on the shares, the source node constructs an anonym-ity cloud with an irregular shape around itself to protect its location privacy. Specifically, an anonymity cloud is a set of active nodes with similar radio actions and they are statisti-cally indistinguishable from each other. The size of the cloud is controlled by the preset number of hops that the shares can walk in the cloud. At the border of the cloud, the fake source nodes independently send the shares to the sink node through proper routing algorithms. At last, the original message can be recovered by the sink node once at least t shares are re-ceived. Simulation results demonstrate that SPAC can strongly protect the source-location privacy with an efficient manner. Moreover, the message sharing mechanism of SPAC increases confidentiality of network data and it also brings high tolerance for the failures of sensor nodes to the data transmission process.

JavaScript has been widely used on the Internet because of its powerful features, and almost all the websites use it to provide dynamic functions. However, these dynamic natures also carry potential risks. The authors of the malicious scripts started using JavaScript to launch various attacks, such as Cross-Site Scripting (XSS), Cross-site Request Forgery (CSRF), and drive-by download attack. Traditional malicious script detection relies on expert knowledge, but even for experts, this is an error-prone task. To solve this problem, many learning-based methods for malicious JavaScript detection are being explored. In this paper, we propose a novel deep learning-based method for malicious JavaScript detection. In order to extract semantic information from JavaScript programs, we construct the Program Dependency Graph (PDG) and generate semantic slices, which preserve rich semantic information and are easy to transform into vectors. Then, a malicious JavaScript detection model based on the Bidirectional Long Short-Term Memory (BLSTM) neural network is proposed. Experimental results show that, in comparison with the other five methods, our model achieved the best performance, with an accuracy of 97.71% and an F1-score of 98.29%.

With the rapid development of cloud computing, massive amounts of data are uploaded to cloud servers for storage. For privacy protection, sensitive data should be encrypted before outsourcing, and ciphertext retrieval technologies based on similarity come into being. In cloud computing with massive data, the efficiency and accuracy of retrieval are crucial. However, most of the current similarity retrieval schemes do not perform well in these two aspects. Therefore, we propose SESR scheme, a secure and efficient similarity retrieval scheme based on homomorphic encryption. Firstly, we use Hamming distance to calculate the similarity between the feature vector of the data and query vector from the data user. Secondly, the homomorphic encryption algorithm is used to encrypt data to protect data privacy. Furthermore, we creatively design a BK-KD tree structure that hierarchically implements similarity search and fine-grained access control, thereby speeding up the retrieval efficiency. In addition, we design a two-cloud-server cooperative retrieval model and a message authentication scheme, which ensure access pattern privacy security and the integrity of the transmitted data simultaneously. We also propose an improved SESR scheme. In this scheme, we use Simhash algorithm to generate feature vectors and query vectors, which reduces storage overhead. Finally, the security of SESR is formally proved and the simulation results show the efficiency and accuracy of the retrieval scheme.

With the rapid development of 5G/6G, most Internet of Things (IoT) devices will embrace wireless connection in the near future. A public concern is how to securely organize, store and retrieve data generated from IoT devices. Many cloud-based IoT data storage schemes have been proposed recently. However, for an untrusted or vulnerable cloud server, the stored IoT data can be easily accessed, modified and even destroyed given that the IoT data are stored in total centralization. Moreover, the servers in a cloud are generally homogeneous and thus vulnerable to attacks. For improvements, we design a novel framework for secure and efficient IoT data storage based on secret sharing and a collaborative blockchain. First, an ultra-lightweight secret sharing algorithm is designed to map original messages generated by IoT devices to a set of shorter message shares. Second, all the shares of IoT messages are separately delivered to different clouds for storage. To guarantee the security of shares, the delivery is notarized on a proposed blockchain. Specifically, both hash values of the shares and their information of location are embedded in blocks which are then chained to form a blockchain. Third, we create a balanced index structure about the shares for each cloud storage node based on the information in the blockchain, and we also propose a depth-first data search algorithm to improve IoT data retrieval efficiency. Theoretical analysis and simulation results illustrate that our scheme can store and retrieve the IoT data securely and efficiently.

Internet of Vehicles (IoV) enables vehicles and other smart terminals to share data in order to improve traffic safety and efficiency. With the increase in data volume and data interaction frequency, the burden of the central server increases dramatically, and it is difficult for resource-limited vehicles to store and verify all messages in the untrusted environment as well. Therefore, a secure and distributed IoV data-sharing scheme needs to be constructed. This paper proposes a hybrid proofof-stake (PoS) blockchain consensus and constructs a secure and efficient distributed IoV data-sharing scheme based on it. In this scheme, participants can store and share data securely in a distributed way through blockchain. The hybrid PoS blockchain consensus, where block producers are chosen based on their own stakes, and multiple trusted managers supervise each other updating head blocks periodically, maintains a more secure blockchain as a consistent and tamper-proof record for data sharing, than some researches using existing consensus, as shown in the security analysis. Every entity can participate in the blockchain consensus equally, without consuming high computation power, which improves the distribution degree and achieves co-governance of the IoV system. The performance evaluation shows that referring to the Gini coefficient, our scheme achieves a higher degree of decentralization than other schemes with only RSUs participants. In this way, vehicles collaborating with roadside units, realize distributed storage and secure data sharing to whoever with their authorization.