Search

The purpose of this exploration of blockchain in vehicle management based on Digital Twins in Vehicular Adhoc Networks (VANETs) is to further improve intelligent transportation in smart cities. In view of the complexity of pedestrians in the real road network, the Digital Twins (DTs) technology is used to map the traffic situation in the real road network to the virtual space. Furthermore, the concrete interaction of vehicle data information is stored and transmitted by using blockchain technology. Finally, the DTs model of vehicle-mounted Ad Hoc network based on blockchain is constructed, and its performance is analyzed by simulation. The results suggest that the model algorithm adopted in this work shows a lower average delay time, its data message delivery rate is basically stable at 80%, the data message leakage rate is basically stable at approximately 10%, and the communication overhead does not exceed 700 bytes. Therefore, the in-vehicle self-organizing network model constructed in this work shows high network security performance while ensuring low latency performance, enabling information to interact more efficiently. Therefore, it can provide an experimental basis for the intelligent development and safety performance improvement of the transportation field of smart cities. [ABSTRACT FROM AUTHOR]

The purpose is to solve the security problems of the Cooperative Intelligent Transportation System (CITS) Digital Twins (DTs) in the Deep Learning (DL) environment. The DL algorithm is improved; the Convolutional Neural Network (CNN) is combined with Support Vector Regression (SVR); the DTs technology is introduced. Eventually, a CITS DTs model is constructed based on CNN-SVR, whose security performance and effect are analyzed through simulation experiments. Compared with other algorithms, the security prediction accuracy of the proposed algorithm reaches 90.43%. Besides, the proposed algorithm outperforms other algorithms regarding Precision, Recall, and F1. The data transmission performances of the proposed algorithm and other algorithms are compared. The proposed algorithm can ensure that emergency messages can be responded to in time, with a delay of less than 1.8s. Meanwhile, it can better adapt to the road environment, maintain high data transmission speed, and provide reasonable path planning for vehicles so that vehicles can reach their destinations faster. The impacts of different factors on the transportation network are analyzed further. Results suggest that under path guidance, as the Market Penetration Rate (MPR), Following Rate (FR), and Congestion Level (CL) increase, the guidance strategy’s effects become more apparent. When MPR ranges between 40% ~ 80% and the congestion is level III, the ATT decreases the fastest, and the improvement effect of the guidance strategy is more apparent. The proposed DL algorithm model can lower the data transmission delay of the system, increase the prediction accuracy, and reasonably changes the paths to suppress the sprawl of traffic congestions, providing an experimental reference for developing and improving urban transportation. [ABSTRACT FROM AUTHOR]

Biometrics is one of the most effective solutions for personal authentication. With the wide use of biometrics technology, the scale of the biometrics database is increasing continuously. In this situation, fast matching and retrieval techniques become more and more necessary for large-scale biometrics recognition. Recently, palmprint recognition has become one of emerging biometrics technologies. Both low-resolution palmprint recognition and 3D palmprint recognition technologies were proposed and can be used for many civilian applications. However, the research on fast low-resolution palmprint and 3D palmprint retrieval is very preliminary. Hashing is one of the most popular techniques for image search due to its fast-computational speed and low storage cost. In this paper, we conduct a performance evaluation of hashing techniques for 2D and 3D palmprint retrieval and recognition. Several representative hashing methods are selected for performance evaluation including four unsupervised hashing methods, four supervised hashing methods, and four deep hashing methods. Experimental results demonstrate that some hashing methods such as column sampling based discrete supervised hashing (COSDISH) can obtain promising results. [ABSTRACT FROM AUTHOR]