A Scheduler for Time Slotted Channel Hopping Networks supporting QoS Differentiated Services

Abstract

In monitoring applications for industry, collected data are heterogeneous and data services often require real-time communication, strictly constrained in delay and jitter. In industrial Wireless Sensor Networks, multiple RPL instances with different objective functions are established to provide QoS differentiated data services. The Time Slotted Channel Hopping (TSCH) mode of the Medium Access Control (MAC) layer in IEEE.802.15.4e has allowed more reliable, more deterministic and timely data delivery. An open issue is that the allocation of physical resources (timeslots) at the TSCH MAC layer should comply with the predefined objective function of each RPL instance. In this work, we proposed a learning-based scheduler for TSCH in which the delay-sensitive data is ensured a minimum delay delivery. Extensive simulation studies show that the scheduler obtains very low latency in data delivery under low energy consumption at each sensor node.