From perception to reaction: an RLCC-engine-driven cybersecurity framework for smart agricultural ecosystems

Abstract Agriculture 4.0 makes significant use of interconnected sensors, autonomous farm equipment, cloud-based farm management systems, and data-driven decision-making. Even though these tools make farm operations more productive and sustainable, they also expose farm equipment, Internet of Things (IoT) devices, supply-chain management systems, and farm data to Cybersecurity attacks. Cybersecurity incidents in the agricultural sector may result in operational disruption, financial loss, and the compromise of data sovereignty. To address these risks, this study adopts a three-stage analytical approach consisting of challenge identification through literature and infrastructure analysis, prioritisation of the identified challenges, and systematic mapping onto a four-layer Agricultural Data-sharing Ecosystem (ADE) architecture comprising perception, network, edge, and cloud/application layers. The analysis identifies six structural challenges driving Cybersecurity risks in agriculture: data management, data governance, integration and interoperability, data privacy and security, knowledge gaps, and infrastructure limitations. Based on these findings, a conceptual multi-layer Cybersecurity framework is proposed that introduces a dedicated Reaction Layer enabling coordinated prevention, detection, and mitigation across ADE layers. The operational logic of the Reaction Layer is formalised through the Reaction Layer Cross-Layer Coordination (RLCC) algorithm, which specifies the sequential coordination of prevention, detection, mitigation, and audit actions across the ADE infrastructure layers, and links each response action to the identified challenges. To support the transition from conceptual design to practical deployment, an initial prototype architecture is outlined. It identifies which components can use existing technologies and which require new development. It also defines the scenarios and metrics that will be used to evaluate the framework.