Reduction of the energy consumption of wireless sensor nodes is crucial for the successful deployment of Internet of Things (IoT) services and applications. This paper investigates the trade-off between energy and information freshness from the perspective of the transmission module of IoT sensor nodes. In particular, timer-based sleep-wake scheduling is investigated for a single sensor node for the purpose of reducing the energy consumption of the transmission module for which average Age of Incorrect Information (AoII) is used as the performance metric for information freshness. We propose two scheduling strategies, namely Idling and Zero-idle, to control the transitions between Sleep and Active modes of operation. To the best of our knowledge, this is the first study considering the tradeoff between energy and average AoII for sleep-wake servers. We first provide closed-form expressions for both average AoII and average power when the updates take place according to a Poisson process and service times are either deterministic or phase-type distributed. Subsequently, we obtain the optimal timer parameters in closed form for the proposed strategies that minimize the energy consumption while the average AoII is kept below a pre-defined threshold. Moreover, we also propose a practical variation of both strategies that adaptively tunes the timer parameters of the sleep-wake scheduler when the update (or arrival) rate changes in time. We validate our analysis and findings with simulations.

Frame coalescing is a well-established technique which manages the low power idle (LPI) mode supported by energy efficient Ethernet (EEE) interfaces. Generally, this technique enables EEE interfaces to remain in the LPI mode for a certain amount of time upon the arrival of the first frame (timer-based coalescing), or until a predefined amount of traffic accumulates in the transmission buffer (size-based coalescing). In this paper, we propose a novel open-loop dynamic coalescing technique that is based on model predictive control (MPC) and queuing theory. In contrast with the conventional timer-based coalescing, the proposed method enables the update of the timer parameter repeatedly throughout the duration of the LPI mode of a single coalescing cycle by taking into account the arrival instant and size of the frames waiting in the buffer. Two different methods, namely MPC-mean and MPC-tail, are proposed which attempt to minimize the energy consumption of the Ethernet link under constraints on mean and tail of the queue waiting time, respectively. The effectiveness of the proposed dynamic MPC-based coalescing algorithms are validated using simulations with synthetic and actual traffic traces.