Joint RAN Slicing and Computation Offloading for 5G-Assisted Autonomous Vehicular Networks Seminar
Thursday, July 8, 2-3 p.m.
Dr. Qiang Ye, Minnesota State University
Faculty candidate seminar
In this seminar, I will first give an introduction on fifth generation (5G) communication networks and network slicing as a new resource orchestration technology. A two-timescale radio access network (RAN) slicing and computing task offloading framework is then presented for a 5G-assisted autonomous vehicular network with layered edge computing. The proposed framework is to jointly maximize the communication and computing resource utilization with diverse quality-of-service (QoS) guarantee for autonomous driving tasks. Specifically, to capture the small-timescale network dynamics, a computing task offloading problem is formulated as a stochastic optimization program, for maximizing the long-term network-wide computation load balancing with minimum task offloading variations. To deal with the problem complexity and information uncertainty, a cooperative multi-agent deep Q-learning (MA-DQL) with fingerprint is employed to solve the problem by learning a set of stationary task offloading policies. Given the offloading decisions, a RAN slicing problem is further studied in a large timescale, which is formulated as a convex optimization program, to maximize the network utility with statistical QoS provisioning. Due to the correlation between the problems of two timescales, a hierarchical optimization framework is established to iteratively determine the optimal radio resource slicing ratios for computation load balancing. Future research directions will also be discussed.
Presented by Department of Computer Science