Task Scheduling for Multi-Access Edge Computing in IRS-Aided Vehicular Networks

Nei Kato, Professor, Fellow of Engineering Academy of Japan, IEEE, IEICE

Graduate School of Information Sciences, Tohoku University

Nei Kato is a full professor and the Dean with Graduate School of Information Sciences, Tohoku University. He has researched on computer networking, wireless mobile communications, satellite communications, ad hoc & sensor & mesh networks, UAV networks, smart grid, AI, IoT, Big Data, and pattern recognition. He has published more than 500 papers in prestigious peer-reviewed journals and conferences. He is the Director of Magazine of IEEE Communications Society. He served as the Vice-President (Member & Global Activities) of IEEE Communications Society (2018-2021), the Editor-in-Chief of IEEE Network Magazine(2015-2017), and the Editor-in-Chief of IEEE Transactions on Vehicular Technology (2017-2021). He is a Distinguished Lecturer of IEEE Communications Society and Vehicular Technology Society, a Fellow of the Engineering Academy of Japan, a Fellow of IEEE, and a Fellow of IEICE.

Abstract: Multi-access Edge Computing (MEC) has played an important role in realizing intelligent beyond 5G (B5G) vehicular networks. The computation tasks of intelligent applications can be offloaded to and processed by near-end-user MEC servers to meet strict latency requirements. However, the latency of provided services is dependent on MEC processor scheduling and millimeter wave (mmWave) transmission conditions for the urban B5G vehicular networks. To alleviate the mmWave signal attenuation caused by buildings, Intelligent Reflecting Surface (IRS) has been regarded as efficient and prospective infrastructure. In this paper, we study the IRS-aided MEC-served vehicular networks and analyze the relationship between computation resource allocation and offloading policy at an intersection. Considering the vehicle mobility patterns, transmission conditions, and task sizes, we optimize the task scheduling by improving the allocation of limited processors and IRS resource. Moreover, the mutual interference among concurrent transmissions is taken into account. In this presentation, by assuming the moving directions available, a dynamic task scheduling algorithm is presented which considers both the communications and computations. The simulation results illustrate that the new scheme outperforms benchmark methods in terms of task offloading rate, computing rate, and finish rate for the IRS-aided MEC-served vehicular networks.