eLoBaPS: Towards Energy Load Balancing with Wake-Up Radios for IoT

In an effort to extend the lifetime and reliability of multi-hop wireless sensor networks we recently presented LoBaPS, a protocol to select opportunistic parents and achieve load balancing. This algorithm takes advantage of the wake-up radio for its ultra-low power consumption and always-on feature. Moreover, it overcomes an open problem in the routing layer: achieving both stability and efficient parent selection at the same time. However, the random load balancing strategy and the energy wastage in listening mode still limits the network lifetime. In this article, we present eLoBaPS, a significant modification of LoBaPS that distributes better the energy among the parents improving around 17% the lifetime of the network towards the ideal case. In a nutshell, the next hop is selected in a decentralized way and it is the parent that issues a shorter back-off period before attempting to retransmit. In addition, the nodes overhear all the traffic in the wake-up radio channel and adapt the protocol parameters to the current state of the battery of the neighbors. We perform simulations with a network of ContikiOS nodes running eLoBaPS, LoBaPS and W-MAC, a reference protocol that uses the wake-up radio.