By exploiting novel HW concepts and HW/SW cross optimizations, we are developing solutions for a new generation of sensor nodes equipped with radio triggering circuits and multi-source energy harvesting. Collectively, these new technologies have the potential to change the time horizon for WSNs autonomous operations, posing the basis for energy neutral systems.
We have designed and developed solutions for structural health monitoring, mechanisms for energy prediction and management, protocols for harvesting-aware resource allocation, and harvesting-aware schemes that enable robust and reliable security support. We have also developed GreenCastalia, an open-source energy-harvesting simulation framework that allows to simulate networks of embedded devices with heterogeneous harvesting capabilities.
Green WSNs for structural health monitoring
This activity has been performed within the framework of the FP7 EC project GENESI, coordinated by Prof. Chiara Petrioli. The first version of GENESI system has been used to monitor structural health monitoring parameters during the construction of the Rome underground B1 line. As a part of this activity, we used wind-energy-harvesting Telos B motes to instrument 220 meters of underground tunnel, collecting data about the air-flow generated by passing trains for more than a month. An analysis of the energy availability in such scenario, quantified in terms of typical WSN operations, such as communication, storage and sensing, is available here.
We have also presented a power management technique for improving the efficiency of harvesting energy from air-flows. The proposed architecture consists of a two-stage energy conversion circuit: an AC-DC converter followed by a DC-DC buck-boost regulator with Maximum Power Point Tracking (MPPT) capability. By using the adaptive AC-DC converter combined with power prediction algorithms, nodes deployed in an underground tunnel of the Metro B1 line in Rome can harvest up to 22% more energy with respect to previous methods.
Energy prediction models
Energy prediction methods can be employed to alleviate the problem of the variable behavior of ambient energy sources, which results in different amounts and rates of energy available to power-scavenging devices over time. Such models forecast the source availability and estimate the expected energy intake, allowing the system to take critical decisions about the utilization of the available energy. We contribute to this topic by developing a general framework for multi-source (i.e., solar and wind) energy-harvesting systems, which accurately predicts the energy intake within forecasting horizons that are dynamically chosen based on the application needs. The key component of our solution is Pro-Energy (PROfile Energy prediction model), a novel energy prediction model that leverage past energy observations to provide estimations of future energy availability. Extensive performance evaluations, performed by using real-life traces of harvested energy, confirmed that Pro-Energy significantly outperforms energy predictors previously proposed in the literature, such as EWMA and WCMA.
We also presented a further enhancement of the Pro-Energy prediction algorithm, named Pro-Energy-VLT, which adapts to the source dynamics to improve the accuracy of energy predictions, while reducing memory and energy overhead.
Task allocation and selective activation
Sensor mission assignment involves matching the sensing resources of a WSN to appropriate tasks (missions), which may come to the network dynamically. Although solutions for WSNs with battery-operated nodes have been proposed for this problem, no attention has been given to networks whose nodes have energy harvesting capabilities, which impose quite a different energy model. We have addressed such problem by providing both an analytical model and a distributed heuristic, called EN-MASSE, for energy harvesting WSNs. The objective of both our model and EN-MASSE is to maximize the profit of the network, fully exploiting the harvesting technologies, while ensuring the execution of the most critical missions within a given target WSN lifetime. By comparing mission assignment schemes in several different scenarios we have demonstrated that traditional assignment algorithms cannot harness the full potential provided by the harvesting technology, which is instead taken into account efficiently by our proposed scheme. Moreover, using our analytical model as a benchmark, we also show that the profit earned by EN-MASSE is close to the optimum. Finally, we have implemented our proposed solution in TinyOS and experimentally validated its performance.
Energy harvesting capabilities embedded in modern sensor nodes permit to take design choices, for a variety of wireless sensor network services, which would be hardly viable in traditional battery-powered motes. Such new opportunities are expecially attractive in the context of security in WSNs, as many solutions are deemed as being too energy demanding to be practically implemented on embedded devices. Our contribution in this topic includes the development of a context-aware decentralized data Access control for GREEn WSNs, called AGREE and of techniques to make ECDSA signatures energetically low cost and implementable on resource-constrained devices.
We have proposed several optimizations for dealing with resource and energy constrained embedded systems, as well as a caching mechanisms to exploit the characteristics of Green WSNs. Our implementation on MagoNode++, Telos B and Mica2 motes have confirmed that such technique are able to efficiently operate based on the energy harvested in excess.
Simulations of energy-harvesting WSNs
The emergence of energy-scavenging techniques for powering networks of embedded devices is raising the need for dedicated simulation frameworks that can support researchers and developers in the design and performance evaluation of harvesting-aware protocols and algorithms. Our contribution to this topic is GreenCastalia
, an open-source energy-harvesting simulation framework we have developed for the popular Castalia simulator. GreenCastalia supports multi-source and multi-storage energy harvesting architectures, it is highly modular and easily customizable. In addition, it allows to simulate networks of embedded devices with heterogeneous harvesting capabilities.
Radio-triggering techniques and wake-up-enabled communication stacks for Energy-Harvesting WSNs
Emerging low-power radio triggering techniques for wireless motes are a promising approach to prolong the lifetime of WSNs. By allowing nodes to activate their main transceiver only when data need to be transmitted or received, wake-up-enabled solutions virtually eliminate the need for idle listening, thus drastically reducing the energy toll of communication.
Within the framework of the FP7 EC project GENESI
, in collaboration with the EE group lead by Prof. Trifiletti, we have designed and developed a novel wake-up receiver architecture based on an innovative pass-band filter bank with high selectivity capability. The proposed concept combines both frequency-domain and time-domain addressing space to allow selective addressing of nodes. To take advantage of the functionalities of the proposed receiver, as well as of energy-harvesting capabilities modern sensor nodes are equipped with, we have designed a novel wake-up-enabled harvesting-aware communication stack that supports both interest dissemination and convergecasting primitives. This stack builds on the ability of the developed WuR to support dynamic address assignment, which is exploited to optimize system performance. Comparison against traditional WSN protocols shows that the proposed concept allows to optimize performance tradeoffs with respect to existing low-power communication stacks.