An Efficient Resource Scheduling Strategy for V2X Microservice Deployment in Edge Servers

Abstract

The fast development of connected vehicles with support for various V2X (vehicle-to-everything) applications carries high demand for quality of edge services, which concerns microservice deployment and edge computing. We herein propose an efficient resource scheduling strategy to containerize microservice deployment for better performance. Firstly, we quantify three crucial factors (resource utilization, resource utilization balancing, and microservice dependencies) in resource scheduling. Then, we propose a multi-objective model to achieve equilibrium in these factors and a multiple fitness genetic algorithm (MFGA) for the balance between resource utilization, resource utilization balancing, and calling distance, where a container dynamic migration strategy in the crossover and mutation process of the algorithm is provided. The simulated results from Container-CloudSim showed the effectiveness of our MFGA.

Evaluation and Challenges of IoT Simulators for Intelligent Transportation System Applications

“…In their work [18], the authors proposed an efficient resource scheduling strategy to optimize the deployment of V2X (Vehicle-to-Everything) microservices in edge servers. They addressed the challenge of resource allocation in edge computing environments by introducing a dynamic resource allocation algorithm that took into consideration service requirements and resource availability.…”

Microservices Scheduling Algorithms: A Survey Study

“…Other researchers [10] suggest a model for multi-objective resource scheduling for vehicle-to-everything (V2X) microservices based on the edge container cloud architecture. The scheduling model that the authors worked on is the multiple fitness genetic algorithm (MFGA).…”

“…Using those three factors, the authors introduced the goal of this study under the formula of an objective function, as shown in equation (10).…”

“…Next, the authors defined a fitness function, which has a value that indicates if the solution to the problem is weak or not. The function is calculated using weight parameters, which are microservice dependencies, resource utilization, and resource utilization balancing [10].…”

“…The final step is to re-add the containers that are missing because of gene exchange using the fitness function and the gene evaluation function. The authors have grouped all the steps mentioned above to formulate the MFGA algorithm [10].…”