Reconfigurable Intelligent Surface assisted Computation Offloading for autonomous systems in Mobile Edge Computing

Abstract

Mobile Edge Computing (MEC) is a new paradigm that utilizes edge infrastructure to bring computation power closer to end-users. This reduces latency and improves performance. With the advancement of self-driving technology, real time traffic monitoring, and on-board entertainment services, vehicular networks have made significant progress. Roadside units (RSUs), or roadside edge servers, are used by MEC and strategically placed along highways to bring computing resources and services closer to the vehicle. Through optimized performance, vehicular services can meet the high standards of computation and precision necessary for efficient and reliable performance. However, a problem arises when the vehicle and roadside unit (RSU) are outside the line of sight (LOS) communication range of each other. Reconfigurable intelligent surfaces (RIS) have become a potential solution to solve this problem. These intelligently reflect the signal towards the receiver in mm Wave and THz communication when there is a blockage between the transmitter and receiver. In this thesis, we propose an RIS-assisted latency-aware computational offloading strategy for autonomous systems in a mobile edge computing environment. This strategy enables an autonomous vehicle to offload its task to an RSU even when the LOS view between the autonomous vehicle and RSU is blocked. We place an RIS at the center of this environment to enable line-of-sight communication between the vehicle and RIS, and between the RIS and RSU. Our simulations show that our proposed approach works well in a dynamic environment where the conditions are constantly changing, in terms of received signal strength and time delay. We also compared our results to the existing schemes, and our approach showed 10 dBm increase in receive power at RSU. The proposed solution achieved 5-7 seconds reduction in MES execution delay compared to local execution delay. The simulation results demonstrated a clear correlation between RTT and the number of states in the system.