No abstract

In this paper, one solution for an end-to-end deep neural network for autonomous driving is presented. The main objective of our work was to achieve autonomous driving with a light deep neural network suitable for deployment on embedded automotive platforms. There are several end-to-end deep neural networks used for autonomous driving, where the input to the machine learning algorithm are camera images and the output is the steering angle prediction, but those convolutional neural networks are significantly more complex than the network architecture we are proposing. The network architecture, computational complexity, and performance evaluation during autonomous driving using our network are compared with two other convolutional neural networks that we re-implemented with the aim to have an objective evaluation of the proposed network. The trained model of the proposed network is four times smaller than the PilotNet model and about 250 times smaller than AlexNet model. While complexity and size of the novel network are reduced in comparison to other models, which leads to lower latency and higher frame rate during inference, our network maintained the performance, achieving successful autonomous driving with similar efficiency compared to autonomous driving using two other models. Moreover, the proposed deep neural network downsized the needs for real-time inference hardware in terms of computational power, cost, and size.

Low Power Wide Area Networks (LPWANs) are gaining attention in both academia and industry by offering the possibility of connecting a large number of nodes over extended distances. LoRa is one of the technologies used as a physical layer in such networks. This paper investigates the LoRa links over seawater in two typical scenarios: clear Line-of-Sight (LOS) and obstructed path in two different Industrial, Scientific and Medical (ISM) radio bands: 8680.266667emMHz and 4340.266667emMHz. We used three different LoRa devices in the experiments: the Own Developed LoRa Transceiver (ODT) and two commercial transceivers. Firstly we investigated transceivers’ Receive Signal Strength Indicator (RSSI) and Signal-to-Noise (SNR) measurement chain linearity and provided correction factors for RSSI to correlate it with actual signal levels received at transceivers’ inputs. Next, we carried out field experiments for three different LoRa Spreading Factors, SF∈[7,10,12], within a bandwidth of BW=1250.266667emkHz and Coding Rate CR=4/6. The experiments showed that LoRa links are fully feasible over seawater at distances at least 220.266667emkm long, using only low-cost off-the-shelf rubber duck antennas in LOS path condition in both ISM bands. In addition, we showed that LoRa links can be established over 280.266667emkm obstructed LOS oversea path in ISM 4340.266667emMHz band, but using costly, higher gain antennas. Furthermore, the laboratory experiments revealed that RSSI is linear in a wide range, up to −500.266667emdBm, whereas the SNR measurement chain goes into saturation for Received Signal Strength (RSS) values higher than −1000.266667emdBm. These findings enabled accurate interpretation of the results obtained in field experiments.

No abstract

No abstract