SlotSwapper: A Schedule Randomization protocol for Real-Time WirelessHART Networks

TL;DR

This paper introduces SlotSwapper, a schedule randomization protocol for WirelessHART networks that enhances security by making communication schedules unpredictable while maintaining real-time constraints.

Contribution

The paper proposes a novel schedule randomization method, SlotSwapper, that preserves feasibility in WirelessHART networks and increases schedule entropy for improved security.

Findings

Schedule feasibility maintained in simulations

Increased schedule entropy indicates higher unpredictability

Effective in random topologies with 100 motes

Abstract

Industrial process control systems are time-critical systems where reliable communications between sensors and actuators need to be guaranteed within strict deadlines to maintain safe operation of all the components of the system. WirelessHART is the most widely adopted standard which serve as the medium of communication in industrial setups due to its support for Time Division Multiple Access (TDMA)based communication, multiple channels, channel hopping, centralized architecture, redundant routes and avoidance of spatial re-use of channels. However, the communication schedule in WirelessHART network is decided by a centralized network manager at the time of network initialization and the same communication schedule repeats every hyper-period. Due to predictability in the time slots of the communication schedule, these systems are vulnerable to timing attacks which eventually can disrupt the safety of the system. In this work, we present a moving target defense mechanism, the SlotSwapper, which uses schedule randomization techniques to randomize the time slots over a hyper-period schedule, while still preserving all the feasibility constraints of a real-time WirelessHART network and makes the schedule uncertain every hyper-period. We tested the feasibility of the generated schedules on random topologies with 100 simulated motes in Cooja simulator. We use schedule entropy to measure the confidentiality of our algorithm in terms of randomness in the time slots of the generated schedules.