Comparative analysis of rule-based heuristic algorithms for microservice chain placement in fog computing

The rapid proliferation of IoT devices has created unprecedented demands for efficient, low-latency data processing in networks. Fog computing emerges as a critical solution by leveraging decentralized computational resources, yet faces fundamental challenges in resource management, particularly in solving the Fog Application Placement Problem (FAPP) efficiently while maintaining performance in dynamic scenarios. While recent works have gravitated toward sophisticated approaches with many dependencies, the often-overlooked rule-based heuristics offer fast, one-shot placement solutions. However, their performance characteristics remain inadequately analyzed, which is crucial given their one-shot nature without gradual improvement capabilities. This research addresses this gap through comprehensive empirical evaluation of rule-based approaches using the YAFS simulator, focusing on examining response time and network usage metrics. Our results demonstrate that rule-based heuristics consistently outperform genetic algorithms (GA) across all tested scenarios, achieving faster execution times while maintaining superior performance metrics. Through systematic analysis of placement patterns, network transmission patterns, and response time components, we provide a complete performance characterization of rule-based methods for FAPP. These findings offer critical insights for implementing lightweight placement solutions in emerging fog deployments. The study establishes a foundation for next-generation fog resource management through the development of dynamic heuristic algorithms, security-based algorithms, energy-aware solutions, and their integration with machine learning techniques for dynamic evaluation.