iFogSim-Placement: A Simulation Framework for Edge Service Placement
TL;DR
iFogSim-placement, a framework designed for experimenting with self-adaptive service placement, is introduced, demonstrating its ability to capture performance metrics, including application latency, resource usage, and energy consumption.
Abstract
Service placement in edge computing is a complex multi-objective optimisation problem defined by conflicting requirements and dynamic environments. While self-adaptive approaches are essential for handling this dynamism, the community needs specific frameworks to evaluate them. We introduce iFogSim-placement, a framework designed for experimenting with self-adaptive service placement. Unlike existing simulators, iFogSim-placement decouples decision-making from execution for modelling information staleness, mobility, and stochastic demand. The artefact enables implementing self-adaptive approaches as Java plug-ins and evaluating them under heterogeneous configurations. Researchers can define experiments to evaluate self-adaptive approaches under variable configurations, including network topology, application complexity, request patterns, and mobility scenarios. We present the architecture, usage instructions, and validation of the artefact, demonstrating its ability to capture performance metrics, including application latency, resource usage, and energy consumption. The artefact is open-source (https://github.com/DawnSpider96/iFogSim-Placement), aiming to lower the barrier to developing and fairly benchmarking self-adaptive edge systems.