Collecting and analyzing massive data generated from smart devices have become increasingly pervasive in crowdsensing, which are the building blocks for data-driven decision-making. However, extensive statistics and analysis of such data will seriously threaten the privacy of participating users. Local differential privacy (LDP) was proposed as an excellent and prevalent privacy model with distributed architecture, which can provide strong privacy guarantees for each user while collecting and analyzing data. LDP ensures that each user’s data is locally perturbed first in the client-side and then sent to the server-side, thereby protecting data from privacy leaks on both the client-side and server-side. This survey presents a comprehensive and systematic overview of LDP with respect to privacy models, research tasks, enabling mechanisms, and various applications. Specifically, we first provide a theoretical summarization of LDP, including the LDP model, the variants of LDP, and the basic framework of LDP algorithms. Then, we investigate and compare the diverse LDP mechanisms for various data statistics and analysis tasks from the perspectives of frequency estimation, mean estimation, and machine learning. Furthermore, we also summarize practical LDP-based application scenarios. Finally, we outline several future research directions under LDP.

Electronic voting plays a key role in the functioning of modern democracies. However, existing electronic voting protocols have some disadvantages, such as inability to trace the identity of malicious voters, insecurity of the voting process, and lack of a score voting paradigm. To address these issues, we propose a novel blockchain-based publicly traceable self-tallying voting protocol that implements scored voting. The protocol enables voters to score each candidate within a specified range. At the end of voting, the candidate with the highest total score is declared the winner of the election. Meanwhile, we take advantage of traceable ring signatures and homomorphic time-lock puzzles to balance the anonymity and accountability, and guarantee time-bounded privacy of the ballots. We prove that the proposed protocol satisfies the properties of full-anonymity, linkability, public traceability and time-bounded privacy. In addition, performance analysis shows that our protocol guarantees various properties of score voting with reasonable overhead.

Images are now often used as data carriers on the Internet, which has sped up the development of image encryption based on chaotic systems. The security of this image encryption algorithm is highly dependent on the stability of chaotic systems. This work suggests a two-dimensional chaotic mapping based on the improvement of classical chaotic mapping to address current chaotic mapping issues such small parameter range and poor correlation performance of generated sequences. The Exponent function is added as a disturbance and the Logistic and Cosine one-dimensional chaotic maps are coupled to create a two-dimensional chaotic map. The enhanced map significantly widens the chaotic map's parameter window and raises its level of complexity. The test results demonstrate that the proposed system has 5 parameters, a large key space, a Lyapunov exponent of 7, a slightly modified initial value, and a changeable system state after 2 to 4 iterations. These findings demonstrate that the two-dimensional chaotic map proposed in this paper has better chaotic performance and is therefore better suited for image encryption.

Aiming at the problems of image encryption, such as poor encryption effect and poor security, a compound chaotic map ILSC is designed with Logistic map and Sine map as seed maps. Based on this, an image encryption algorithm based on compound chaotic system and dynamic DNA coding is proposed. First, according to the DNA random coding rules, the plaintext image is converted into a DNA matrix, and the scrambling operation is performed on it, and then the rows and columns of the DNA matrix are XORed to obtain the ciphertext image. Theoretical analysis and simulation results show that the proposed algorithm has a larger key space, the ciphertext image has higher information entropy, and can effectively resist statistical attacks, brute force attacks and other attack methods, and has better performance.

To enhance the security and irreversibility performance of the fingerprint template, a fingerprint-template-generating method is proposed based on the 3D mapping of local minutiae. First, we extract fingerprint minutiae features after preprocessing the fingerprint image and select the minutiae using parameter adaptive circular areas. Second, we project the minutiae into a line. Subsequently, quantization, mapping, and modulo operation are performed on the projected vectors to generate a fixed-length binary bit string. Finally, a fingerprint template is generated by combining the users PIN code with a binary bit string. The experiments performed on FVC2002-DB1 and DB2 show that this template has advantages over the traditional ones in terms of recognition, revocation, non-invertibility, and performance.