Mean Field Analysis of Serverless Edge Computing for Smart Vehicle Applications

Smart cities and vehicles generate numerous time sensitive tasks, making edge computing a key solution for reducing latency. However, using serverless models at the edge faces challenges like dynamic city traffic and resource constraints. This study analyzes serverless edge computing for smart vehicles through modeling and experiments. we develop a mean field model captures steady-state system behavior under stochastic job arrivals, enabling efficient analysis of metrics like mean sojourn time and service degradation probability. Through experiments and simulations, we demonstrate the feasibility of serverless edge servers for low-latency offloading, providing insights for optimizing these systems in urban environments.