A self-configuration framework for balancing services in the fog of things

Abstract

Fog Computing has been playing a pivotal role in the Internet of Things (IoT) ecosystem, offering benefits such as local availability, access facilities, and enhanced communication among devices. However, managing numerous gateways in an IoT network poses service distribution and network management challenges, leading to imbalances and inefficiencies. Within this context, this paper presents a novel self-organizing environment based on the Fog of Things approach, designed to address these challenges. Our key contributions include developing the FoT Balance Management service, which dynamically configures and optimizes the distribution of services across the network. This service utilizes advanced load-balancing algorithms to ensure the workload is evenly distributed among the available gateways, preventing any single node from becoming a bottleneck for the service distributions. Additionally, we integrate Apache Karaf Cellar for real-time monitoring and adaptive reconfiguration. This integration allows the system to continuously monitor the network state and automatically reconfigure the service distribution in response to changes, such as adding or removing nodes. This approach ensures seamless adaptation to network changes, maintaining high performance and load balancing. We validate our solution through planned experiments using ANOVA and a 2^k factorial design. The experimental results demonstrate significant improvements in network performance, response time, and load balancing. Specifically, in scenarios with ten fog nodes, our approach increases average availability by 10 ​%–20 ​% and achieves 70 ​%–80 ​% load balancing. The analysis reveals that the absence of a balancing strategy can reduce availability by approximately 30 ​%. Our proposed solution effectively prevents infrastructure overload, balancing computation costs and node availability, thereby enhancing the efficiency and responsiveness of the IoT ecosystem.