Billions of IoT devices are deployed worldwide and batteries are their main power source. However, these batteries are bulky, short-lived and full of hazardous chemicals that damage our environment. Relying on batteries is not a sustainable solution for the future IoT. As an alternative, battery-less devices run on long-lived capacitors charged using energy harvesters. The small energy storage capacity of capacitors results in an intermittent on-off behaviour. LoRaWAN is a popular Low Power Wide Area Network technology used in many IoT devices and can be used in these new scenarios. In this work, we present a Markov model to characterize the performance of battery-less LoRaWAN devices for uplink and downlink transmissions and we evaluate their performance in terms of the parameters that define the model (i.e., device configuration, application behaviour and environmental conditions). Results show that LoRaWAN battery-less communications are feasible if choosing the proper configuration (i.e., capacitor size, turn-on voltage threshold) for different application behaviour (i.e., transmission interval, UL/DL packet sizes) and environmental conditions (i.e., energy harvesting rate). Since downlink in the second reception window highly affects the performance, only small DL packet sizes should be considered for these devices. Besides, a 47 mF capacitor can support 1 Byte SF7 transmissions every 60 s at an energy harvesting rate of 1 mW. However, if no DL is expected, a 4.7 mF capacitor could support 1 Byte SF7 transmissions every 9 s.

Billions of Internet of Things (IoT) devices rely on batteries as the main power source. These batteries are short-lived, bulky and harmful to the environment. Battery-less devices provide a promising alternative for a sustainable IoT, where energy harvested from the environment is stored in small capacitors. This constrained energy storage and the unpredictable energy harvested result in intermittent on–off behavior of the device. Measuring and understanding the current consumption and execution time of different tasks of IoT applications is crucial to properly operate these battery-less devices. In this paper, we study how to properly schedule sensing and transmission tasks on a battery-less LoRaWAN device. We analyze the trade-off between sleeping and allowing the device to turn off between the execution of application tasks. This study allows us to properly define the device configuration (i.e., capacitor size) based on the application tasks (i.e., sensing and sending) and environmental conditions (i.e., harvesting rate). We define an optimization problem that determines the optimal capacitor voltage at which the device should start performing its tasks. Our results show that a device using LoRaWAN Class A can measure the temperature and transmit its data at least once every 5 s if it can harvest at least 10 mA of current and uses a relatively small capacitor of 10 mF or less. At harvesting rates below 3 mA, it is necessary to turn off the device between application cycles and use a larger supercapacitor of at least 140 mF. In this case, the device can transmit a temperature measurement once every 60–100 s.

Today’s IoT devices rely on batteries, which offer stable energy storage but contain harmful chemicals. Having billions of IoT devices powered by batteries is not sustainable for the future. As an alternative, batteryless devices run on long-lived capacitors charged using energy harvesters. The small energy storage capacity of capacitors results in intermittent on-off behaviour. Traditional computing schedulers can not handle this intermittency, and in this article we propose a first step towards an energy-aware task scheduler for constrained batteryless devices. We present a new energy-aware task scheduling algorithm that is able to optimally schedule application tasks to avoid power failures, and that will allow us to provide insights on the optimal look-ahead time for energy prediction. Our insights can be used as a basis for practical energy-aware scheduling and energy a vailability prediction algorithms. We formulate the scheduling problem as a Mixed Integer Linear Program. We evaluate its performance improvement when comparing it with state-of-the-art schedulers for batteryless IoT devices. Our results show that making the task scheduler energy aware avoids power failures and allows more tasks to successfully execute. Moreover, we conclude that a relatively short look-ahead energy prediction time of 8 future task executions is enough to achieve optimality.

We focus on the problem of managing a shared physical wireless sensor network where a single network infrastructure provider leases the physical resources of the networks to application providers to run/deploy specific applications/services. In this scenario, we solve jointly the problems of Application Admission Control (AAC), that is, whether to admit the application/service to the physical network, and wireless Sensor Network Slicing (SNS), that is, to allocate the required physical resources to the admitted applications in a transparent and effective way. We propose a mathematical programming framework to model the joint AAC-SNS problem which is then leveraged to design effective solution algorithms. The proposed framework is thoroughly evaluated on realistic wireless sensor networks infrastructures.

Tiny batteryless Internet of Things (IoT) devices that depend on the harvested energy from their environment provide a promising alternative for a sustainable IoT vision. These devices use small capacitors as energy storage, which together with the unpredictable and dynamic harvesting environment results in intermittent on–off behavior of the device. The crucial issue to effectively use batteryless IoT devices is to find a way of enabling the successful execution of application tasks in face of this intermittency. As the conventional computing models cannot handle this behavior, in this article, we present an energy-aware task scheduler for batteryless IoT devices based on dependencies and priorities, which can intelligently schedule the application tasks avoiding power failures and maintaining forward progress. With the properly defined voltage thresholds for each application task, using our energy-aware task scheduler a safer execution can be ensured. We evaluate our approach based on emulated and real experiments and validate it using two types of power management units (PMUs) (environment emulator and intelligent PMU based on the AEM10941 chip). Our results show that the energy-aware task scheduler is able to react and adapt the execution to environmental changes, avoiding power failures. Comparing to the state-of-the-art scheduling approaches, which are mostly not aware of the energy, we show that our energy-aware task scheduler can keep the device on during the full time of the experiment, executing more tasks when a relatively small capacitor of 10 mF or less is used at harvesting currents as low as $40 ~\mu \text{A}$ .

Sensor network virtualization is a promising paradigm to move away from highly-customized, application-specific wireless sensor network deployment by opening up to the possibility of dynamically assigning general purpose physical resources to multiple stakeholder applications. In this field, this paper introduces an optimization framework to perform the allocation of physical shared resources of wireless sensor networks to multiple requesting applications. The proposed optimization framework aims to maximize the total number of applications which can share a common physical network, while accounting for the distinguishing characteristics and limitations of the wireless sensor environment (limited storage, limited processing power, limited bandwidth, tight energy consumption requirements). Due to the complexity of the optimization problem, a heuristic algorithm is also proposed. The proposed framework is finally evaluated by simulation considering realistic parameters from actual sensor nodes and deployed applications to provide a detailed performance evaluation and to assess the gain involved in letting multiple applications share a common physical network with respect to one-application, one-network vertical design approaches.