A Low-Energy Security Solution for IoT-Based Smart Farms

This work proposes a novel configuration of the Transport Layer Security protocol (TLS),
suitable for low energy Internet of Things (IoT), applications. The motivation behind
the redesign of TLS is energy consumption minimisation and sustainable farming, as
exemplified by an application domain of aquaponic smart farms. The work therefore considers
decentralisation of a formerly centralised security model, with a focus on reducing energy
consumption for battery powered devices. The research presents a four-part investigation
into the security solution, composed of a risk assessment, energy analysis of authentication
and data exchange functions, and finally the design and verification of a novel consensus
authorisation mechanism. The first investigation considered traditional risk-driven threat
assessment, but to include energy reduction, working towards device longevity within a
content-oriented framework. Since the aquaponics environments include limited but specific
data exchanges, a content-oriented approach produced valuable insights into security and
privacy requirements that would later be tested by implementing a variety of mechanisms
available on the ESP32.
The second and third investigations featured the energy analysis of authentication
and data exchange functions respectively, where the results of the risk assessment were
implemented to compare the re-configurations of TLS mechanisms and domain content.
Results concluded that selective confidentiality and persistent secure sessions between paired
devices enabled considerable improvements for energy consumptions, and were a good
reflection of the possibilities suggested by the risk assessment.
The fourth and final investigation proposed a granular authorisation design to increase
the safety of access control that would otherwise be binary in TLS. The motivation was
for damage mitigation from inside attacks or network faults. The approach involved an
automated, hierarchy-based, decentralised network topology to reduce data duplication whilst
still providing robustness beyond the vulnerability of central governance. Formal verification
using model-checking indicated a safe design model, using four automated back-ends.
The research concludes that lower energy IoT solutions for the smart farm application
domain are possible.