A lightweight decentralized service placement policy for performance optimization in fog computing

TL;DR

The paper addresses the scalability and latency challenges of centralized fog service placement by proposing a fully decentralized, local-optimization policy that runs on each fog device to place the most popular services closer to clients. It migrates under resource constraints along the shortest path to the cloud and coherently moves interoperated services to avoid execution loops, with the objective quantified by the Weighted Average Hop Count $\mathrm{WAHC}$. Key contributions include a survey of FSPP approaches, a low-overhead decentralized placement policy, and an experimental validation in iFogSim showing reduced hop distances and network usage, with substantial latency improvements for high-demand services at some cost to low-demand ones. The approach demonstrates feasibility and potential scalability benefits of fully decentralized fog orchestration for IoT workloads, though it incurs more migrations and may degrade latency for less popular services under heavy workload.

Abstract

A decentralized optimization policy for service placement in fog computing is presented. The optimization is addressed to place most popular services as closer to the users as possible. The experimental validation is done in the iFogSim simulator and by comparing our algorithm with the simulator's built-in policy. The simulation is characterized by modeling a microservice-based application for different experiment sizes. Results showed that our decentralized algorithm places most popular services closer to users, improving network usage and service latency of the most requested applications, at the expense of a latency increment for the less requested services and a greater number of service migrations.

Figures (13)

• Figure 1: Fog computing architecture.
• Figure 2: Example of network delay benefits for a service migration scheme within the shortest path.
• Figure 3: Example of network delay benefits for a service migrating scheme with migration of interoperated services.
• Figure 4: Decentralized service placement manager.
• Figure 5: Example of transitive closures for each service of an application.