Volume 22 (2021): Issue 4 (November 2021)

The main objective of this work is to develop a perception algorithm for self-driving cars which is based on pure vision data or camera data. The work is divided into two major parts. In part one of the work, we develop a powerful and robust lane detection algorithm which can determine the safely drive-able region in front of the car. In part two we develop and end to end driving model based on CNNs to learn from the drivers driving data and can drive the car with only the camera data from on-board cameras. Performance of the proposed system is observed by the implementation of the autonomous car that can be able to detect and classify the stop signs and other vehicles.

Internet of Vehicles (IoV) is a network of vehicles communicating with each other by exchanging road traffic information via radio access technologies. Two potential technologies of V2X that have gained attention over the past years are DSRC and cellular networks such as 4G LTE and 5G. DSRC is suitable for low latency communications, however provides a shorter coverage range whereas, 4G LTE offers a wide coverage range but has high transmission time intervals. In contrast, 5G offers higher data rates, low latencies but prone to blockages. Single technology might not fully accommodate the requirements of vehicular communications. Hence, it is required to interwork with more than one radio access network to satisfy the requirements of safety vehicular applications. One issue identified when working with multiple radio access networks is the selection of the most appropriate network for vertical handover. Usually, in the previous works, the network is selected directly or will be connected to the available network due to which the handover had to take place frequently resulting in unnecessary handovers. Hence, in the existing state-of-the-art, the need for handover is not validated. In this paper, we have proposed a dynamic Q-learning algorithm to validate the need for handover, and then, appropriate selection of network would take place by using a fuzzy convolutional neural network. Besides, a modified jellyfish optimization algorithm is proposed to select the shortest paths by forming V2V pairs that take into account channel metrics, vehicle metrics, and vehicle performance metrics. The proposed algorithms are then evaluated using OMNET++ and compared with the existing state-of-the-art concerning mean handover, HO failure, throughput, delay, and packet loss as the performance metrics.

The article describes a project to use PLC technology in railway light signs. The proposal presents a solution for the separation of supplying circuits with the use of filters. The necessity of separation is a result of long distance data transmission between sign controllers and LED railway signs using PLC modems which work on the same frequency. Problems in the selection of filter parameters are presented due to the fact that there is no specification for the characteristic of the separation filter, there are only requirements for the reliability communication system. To meet these requirements, the circuits separation efficiency must be assessed. The paper presents a quick method to evaluate separation efficiency. Structure of the network, as well as the set of devices which realize this specific kind of wired sensor network for supervising railway LED sign network and maintenance parameters are also presented in this article.

The paper considers an approach to building various mathematical models for homogeneous groups of intersections manifested through the use of clustering methods. This is because of a significant spread in their traffic capacity, as well as the influence of several random factors. The initial data on the traffic flow of many intersections was obtained from real-time recorders of the convolutional neural network. As a result of the analysis, we revealed statistically significant differences between the groups of intersections and compiled their linear regression models as a basis for the subsequent formation of generic management decisions. To demonstrate visually the influence of random factors on the traffic capacity of intersections, we built distribution fields based on the fuzzy logic methods for one of the clusters consisting of 14 homogeneous intersections. Modeling was based on the Gaussian type of membership functions as it most fully reflects the random nature of the pedestrian flow and its discontinuity.

Unlike literature and studies coming from high-income or Western countries, the existing conducted on the Middle East and North Africa fail to draw a nearly complete image of the characteristics of passenger travel behaviors in the urban areas of the region. This gap necessitates a holistic review of the previous studies and comparing their results of those of the international findings. This paper summarizes the status of urban travel behavior studies on the MENA region under eight categories of socioeconomics, land use, perceptions and attitudes, urban sprawl, neighborhood design, public transportation use, active mobility, and new technologies and concepts. Descriptive literature review and desk research depicts both lack of research results or data and differences between the behaviors in the MENA region and the Western countries. Moreover, based on the background review, this paper provides a list of recommendations for having more sustainable mobility in the MENA region.

The application of SON algorithms for automating the processes of operating fourth-generation mobile networks based on the networks of operation, administration and management of OAM (Operation and Maintenance) is considered. The features of the tasks at the stages of self-optimization and self-configuration of the network for the various stages of 4G mobile network life cycle are shown. Criteria and approaches to assessing the effectiveness of solving problems by the SON network are proposed. The technical requirements are also formulated for SON algorithms. The experimentally achieved values of the selected performance exponents depending on the duration of the test cluster self-optimization time of the 4G network are shown.

The technical conditions and service life of steering elements of vehicles are an important factor directly affecting road safety. Therefore, high reliability of such kind’s components is required. In the paper, on the basis of the stand test, the fatigue durability of a ball joint of a steering tie rod is determined. It is elaborated together with a prediction for the further number of cycles, enabling to determine the technical state of the tested component containing its service life. The aim of the article is to select an appropriate mathematical model with respect to describing the relationship between the moment of force and the fatigue cycles performed for the ball joint of a steering rod of a vehicle with a GVW above 3.5 tonnes, and identifying the model’s parameters. As a result, the limit number of loading cycles after which the examined joint does not meet safety requirements is estimated.

Autonomous navigation is important not only in autonomous cars but also in other transportation systems. In many applications, an autonomous vehicle has to follow the curvature of a real or artificial road or in other words lane lines. In those application, the key is the lane detection. In this paper, we present a real-time lane line tracking algorithm mainly designed to mini vehicles with relatively low computation capacity and single camera sensor. The proposed algorithm exploits computer vision techniques in combination with digital filtering. To demonstrate the performance of the method, experiments are conducted in an indoor, self-made test track where the effect of several external influencing factors can be observed. Experimental results show that the proposed algorithm works well independently of shadows, bends, reflection and lighting changes.

This paper proposes a method of automatic speaker-independent recognition of human psycho-emotional states by analyzing the speech signal based on Deep Learning technology to solve the problems of aviation profiling. For this purpose, an algorithm to classify seven human psycho-emotional states, including anger, joy, fear, surprise, disgust, sadness, and neutral state was developed. The algorithm is based on the use of Mel-frequency cepstral coefficients and Mel spectrograms as informative features of speech signals audio recordings. These informative features are used to train two deep convolutional neural networks on the generated dataset. The developed classifier testing on a delayed verification dataset showed that the metric for the multiclass fraction of correct answers’ accuracy is 0.93. The solution proposed in the paper can be in demand in human-machine interfaces creation, medicine, marketing, and in the field of air transportation.