Dynamic Task Offloading for Mobile Edge Computing in Urban Rail Transit
Lin He, Junhui Zhao, Xiaoke Sun, Danyang Zhang · 2021 13th International Conference on Wireless Communications and Signal Processing (WCSP) · 2021
Due to the high-speed mobility of rail transit, the low delay of vehicle-to-ground communication cannot be guaranteed. In order to improve the low delay and stability of information transmission, this paper studies the dynamic data offloading problem based on MEC in urban rail transit. As the key technologies in the 5th generation mobile networks, SDN and MEC can meet the computing needs of mobile users and carry out dynamic management. According to the autonomous learning of reinforcement learning algorithm, SDN controller enables mobile users to select MEC servers to offload their data. In order to determine the offloading of mobile users' data to a specific MEC server, a non-cooperative game is played between mobile users, and the unique existence of Nash Equilibrium point is demonstrated. The optimization problem of the user utility function is formulated to determine the optimal offloading data received by the MEC servers and the maximum user utility function. The simulation results show that the dynamic offloading algorithm proposed in this paper can effectively improve the user's utility function and achieve the optimal offloading.