A Vulnerable-by-Design IoT Sensor Framework for Cybersecurity in Smart Agriculture

Emil Marian Pasca (), Daniela Delinschi, Rudolf Erdei, Iulia Baraian and Oliviu Dorin Matei
Additional contact information
Emil Marian Pasca: Department of Electrical, Electronics and Computer Engineering, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
Daniela Delinschi: Department of Electrical, Electronics and Computer Engineering, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
Rudolf Erdei: Department of Electrical, Electronics and Computer Engineering, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
Iulia Baraian: Department of Electrical, Electronics and Computer Engineering, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania
Oliviu Dorin Matei: Department of Electrical, Electronics and Computer Engineering, Technical University of Cluj-Napoca, 400114 Cluj-Napoca, Romania

Agriculture, 2025, vol. 15, issue 12, 1-36

Abstract: Agricultural Internet of Things (IoT) deployments face unique cybersecurity challenges due to resource constraints and direct impact on food production. This paper introduces a vulnerable-by-design, containerized IoT framework simulating both cybersecurity vulnerabilities and sensor health anomalies in agricultural settings. We demonstrate its agricultural relevance through a tomato greenhouse case study where combined DDoS attacks and sensor faults masked critical temperature increases to 43 °C, potentially reducing yields by up to 30%. Our masking analysis revealed counter-intuitive relationships between sensor faults and attack detectability: spike faults enhanced BOLA attack detectability by up to 95.9%, while dropout faults masked command injection attacks by 18.0%. We identified distinctive temporal signatures for each attack type and quantified these relationships through a composite detectability score. Our LSTM-based validation achieved moderate recall (0.5473 average) with significant variation across fault conditions (0.3194–0.8145), while maintaining strong precision (0.8285). The LSTM model performed best with drift fault conditions (0.9749 accuracy), while DDoS attacks were most consistently detectable (avg. score: 0.6886) and resource exhaustion attacks the most difficult (0.3056). These findings challenge conventional approaches that treat sensor health and security as separate domains. Our open-source implementation with systematic dataset generation capabilities addresses reproducibility challenges in agricultural IoT security while demonstrating that integrated health-security monitoring could significantly improve threat detection in smart agriculture deployments.

Keywords: agricultural cybersecurity; IoT vulnerabilities; containerized testbed; sensor health simulation; LLM usage in cybersecurity; reproducible security testing (search for similar items in EconPapers)
JEL-codes: Q1 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 (search for similar items in EconPapers)
Date: 2025
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2077-0472/15/12/1253/pdf (application/pdf)
https://www.mdpi.com/2077-0472/15/12/1253/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jagris:v:15:y:2025:i:12:p:1253-:d:1675523

Access Statistics for this article

Agriculture is currently edited by Ms. Leda Xuan

More articles in Agriculture from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().