Intelligent transport systems are the future in matters of safe roads and comfortable driving. Integration of vehicles into a unified intelligent network leads to all kinds of security issues and cyber threats common to conventional networks. Rapid development of mobile ad hoc networks and machine learning methods allows us to ensure security of intelligent transport systems. In this paper, we design an authentication scheme that can be used to ensure message integrity and preserve conditional privacy for the vehicle user. The proposed authentication scheme is designed with lightweight cryptography methods, so that it only brings little computational and communication overhead. We also conduct experiments on vehicular ad hoc network segment traffic generation in OMNeT++ tool and apply up-to-date machine learning methods to detect malicious behavior in a given simulated environment. The results of the study show high accuracy in distributed denial-of-service attack detection.
In this paper, we consider the problem of selecting the most efficient optimization algorithm for neural network approximation—solving optimal control problems with mixed constraints. The original optimal control problem is reduced to a finite-dimensional optimization problem by applying the necessary optimality conditions, the Lagrange multiplier method and the least squares method. Neural network approximation models are presented for the desired control functions, trajectory and conjugate factors. The selection of the optimal weight coefficients of the neural network approximation was carried out using the gravitational search algorithm and the basic particle swarm algorithm and the genetic algorithm. Computational experiments showed that evolutionary optimization algorithms required the smallest number of iterations for a given accuracy in comparison with the classical gradient optimization method; however, the multi-agent optimization methods were performed later for each operation. As a result, the genetic algorithm showed a faster convergence rate relative to the total execution time.
As part of this study, an approach is proposed that allows us to identify, on the basis of methods of intellectual data analysis of monitoring systems, the vulnerable elements of the network infrastructure of a cyber-physical system located on the basis of a cloud platform from the point of view of reliability. To conduct a consolidated assessment of the current state of network elements, the study developed a model for ensuring the reliability of the operation of network infrastructure objects, presented as a weighted multigraph, which forms a plan for collecting, analyzing and verifying the data, received from the monitoring system. At the same time, as the vertices of the graph, we chose the reliability parameters for the hotel components and nodes, both the infrastructure of the cyber-physical system and the cloud platform. As the arcs of the graph, the relationships between the established reliability criteria are presented, reflecting the relationship between the state and parameters of the operation of the associated cyber-physical system infrastructure nodes, taking into account the current data flows. This will allow identifying the system segments that require reconfiguration, which will reduce the overhead required to make changes. At the same time, the neural network approach is used to predict the uninterrupted infrastructure of the cyber-physical system. The proposed hybrid approach allowed predicting the behavior of the infrastructure over time and warning possible failures in the operation of individual components and critical nodes.
The rapid growth in the volume of medical data is pushing the development and implementation of artificial intelligence (AI) tools. One of the directions of the application of AI in the field of healthcare is the use of natural language processing methods to build medical decision support systems based on electronic medical record (EMC) data. As a result of this study, a module for the extraction and pretreatment of patients’ EMC was developed. In addition, an approach was implemented to extract features from the unstructured textual information of patient admission protocols, with the formation of an appropriate vector representation of data. Predictive models for the diagnosis of groups of diseases based on the logistic regression model and BERT were developed. The highest efficiency in the experiments was shown by the logistic regression model, with a F1-score of 0.81 and Matthews correlation coefficient of 0.75. The obtained results have been posted for public access based on the django framework and can be used for preliminary assessment of patient health status, as well as integrated into existing medical decision support systems.
This paper is aimed in development of an attack detection model based on an analysis of the state chains of network nodes. In our research, we solve the problem of a typical attack profile in a network of telecommunication service providers formalization while constructing a sequence of network nodes state shiftings and a state change time of particulary investigated devices. The research includes the most popular types of attacks. We use the decision tree algorithm for the construction of the chain of security events. The ROC analysis is used for analyzing the results of security events binary classification, which is found using developed approach. The experimental study allowed evaluating the efficacy of the developed model for network attacks in the telecommunication service providers infrastructure recognition. Experimental results show high accuracy of determining the popular type of attacks. This gives an advantage to reduce the response time to security incidents in a larger network.