Green Edge Servers Placement for Intelligent Transport Systems

Abstract

Edge computing empowers service providers to deploy smart vehicles applications that require high throughput and extremely low latency. In this context, optimal Edge servers' placement becomes more difficult since it requires addressing several interrelated requirements at the same time, such as delay, deployment cost, and energy consumption. This paper studies optimal Edge server placement for energy efficiency. The proposed approach, called Green Optimal Edge Server Placement (GOESP), models the placement problem using integer linear programming to address the trade-off between latency, energy, and deployment cost while considering Edge servers' capacity and expected vehicle's traffic on the road. GOESP minimizes the energy consumption by minimizing the number of deployed Edge servers while meeting end-to-end communication latency and avoiding servers' overloading. We evaluate the efficiency of our approach mathematically and through simulations utilizing real-world traffic extracted from open data of Bordeaux city, France. The results demonstrate that our technique outperforms other methods in terms of energy efficiency and guarantees latency and workload balancing requirements.