Tackling malfunctions caused by Radio Duty Cycling protocols that do not appear in simulation studies

Abstract

When a commercial Wireless Sensor Network designed for environmental control and security in buildings was tested before being launched, disappointing differences between the behaviour in simulations and in the real world were observed. Fortunately, these differences disappeared when the energy saving radio duty cycling protocols were deactivated. The company then decided to provide their customers with larger batteries. They also suggested the authors of this paper to start an in depth experimental analysis of radio duty cycling as implemented in the Contiki operating system for Zolertia Z1 motes. Many subtle issues that remained hidden in simulations appeared to have infrequent but disruptive consequences in the real world. These issues have been identified, understood and corrected. They were a source of inspiration for recommendations tending to improve the overall reliability of Contiki.