- 자세한 설명 + 동영상 첨부 + Pytorch 버전 추가
Study on a Self-Driving Service through a MEC server (ICROS2023)
This paper is an experimental study that tests a self-driving service through an MEC system. This study analyzes the correlation between computing power, which is an important issue in the field of self-driving and proposes an efficient solution using wireless communication. To conduct the experiment, we imported two different self-driving service models (i.e., Nvidia and ResNet-50) on a Raspberry Pi model 3B+, model 4, and a laptop equipped with an Intel i5 processor. Then, we compare driving performance of these two models under the same environment. Through several comparative experiments, efficient wireless communication points could be identified.