No abstract

Driver distraction is one of the major causes of traffic accidents. In recent years, given the advance in connectivity and social networks, the use of smartphones while driving has become more frequent and a serious problem for safety. Texting, calling, and reading while driving are types of distractions caused by the use of smartphones. In this paper, we propose a non-intrusive technique that uses only data from smartphone sensors and machine learning to automatically distinguish between drivers and passengers while reading a message in a vehicle. We model and evaluate seven cutting-edge machine-learning techniques in different scenarios. The Convolutional Neural Network and Gradient Boosting were the models with the best results in our experiments. Results show accuracy, precision, recall, F1-score, and kappa metrics superior to 0.95.

Unmanned aerial vehicles bases stations (UAVBS) have many applications in telecommunications. Enables integration into systems in order to provide network signals for users on the ground. The electromagnetic signal from the UAV is characterized by air-to-ground propagation. At different altitudes, the signal suffers losses along the way, thus facing several problems related to transmissions, such as attenuation, fading, and distortion. This paper studies UAV air-to-ground path loss at different altitudes of the UAV. To this, implement a field measurement campaign, which collects and analyzes the signal strength in wireless networks. Finally, it proposes the use of recurrent neural networks to predict the propagation loss in the network. The results were found to show good accuracy in the chosen scenario.

This research presents a study on the identification of post-quantum cryptography algorithms through machine learning techniques. Plain text files were encoded by four post-quantum algorithms, participating in NIST's post-quantum cryptography standardization contest, in ECB mode. The resulting cryptograms were submitted to the NIST Statistical Test Suite to enable the creation of metadata files. These files provide information for six data mining algorithms to identify the cryptographic algorithm used for encryption. Identification performance was evaluated in samples of different sizes. The successful identification of each machine learning algorithm is higher than a probabilistic bid, with hit rates ranging between 73 and 100%.