An Adaptive Approach to Time Synchronization for Wireless Sensors Under Extreme Conditions

In an attempt to reduce the mass and design complexity of cable harnesses, the space industry is eager to deploy reliable wireless sensor networks (WSNs) in spacecraft and launcher systems. Low latency sensor data acquisition in these systems depends on synchronized local clocks between the nodes of the network, making time synchronization a vital part of their operation. The particular problem of controlling the local clock on a node between the synchronization points provided by a time distribution protocol is critical. A key problem in the space application domain is the harsh changes in environmental conditions like temperature or electromagnetic interference (EMI) that sensors are subjected to which causes the local clocks of network nodes to rapidly drift out of sync. Common approaches to local clock control are not able to sufficiently mitigate these large drift rates and cope poorly with rapidly changing amounts of drift. This article proposes a novel control approach to local clock discipline in the extreme environments these sensors are deployed. The main contribution of this article is the adaptive local clock control (ALCC) algorithm that can minimize the deviation from a reference clock under extreme conditions while performing nearly an order of magnitude better than the common approaches used so far. A formal model for this algorithm is presented that is compared to the existing solutions in a simulation environment that allows to replicate real-world application scenarios. The final solution is subsequently verified on a hardware test platform.