Blind vs. Feedback-Aided Retransmission of High-Priority Data in Asynchronous IoT Networks

We consider IoT networks in which devices sending high-priority messages (to be delivered with high reliability to a gateway) coexist with senders of lower-priority messages. Our focus is on low-cost IoT networks that lack features such as time synchronization, carrier sensing, and resource reservations. For such networks, we compare two approaches, namely blind replications and feedback-aided retransmissions, to improve reliability levels for high-priority traffic. We demonstrate that both approaches increase relia-bility, but their relative performance gains and the impact on the coexisting low-priority messages vary greatly depending on the traffic intensity, density of high-priority nodes, and gateway type (half-duplex vs. full-duplex). An analysis of the message-delivery performance under each scheme is presented in order to facilitate appropriate scheme selection.