Reliable Edge Federation Formation for Smart Agriculture using Machine Learning

Abstract

In the domain of smart agriculture, the integration of edge computing has revolutionized data processing by bringing computational services in close proximity to sensors and movable equipment. The transition to edge computing reshapes the security landscape, where robust safeguards are essential not only for protecting data but also for ensuring reliable service delivery. In agriculture, reliability is critical, as insecure or untrustworthy ESPs can cause interruptions and downtime that undermine both federation performance and provider reputation. The overall challenge is to form a resilient edge federation that enables the seamless sharing of services from trustworthy providers to clients, such as farm equipment and monitoring systems. To address this, a new algorithm Edge Nearest Neighbor (ENN) as an extension of the k-nearest neighbor (kNN) machine learning approach has been presented. The proposed algorithm selects highly reliable ESPs based on user requirements to form a federation, ensuring smart services are delivered without network interruption. This enables precision agriculture systems to reliably monitor and maintain large arable lands effectively. Experiments demonstrate that ENN quickly forms a federation of reliable ESPs with high accuracy, providing a secure and dependable foundation for smart agriculture.