Novel backscatter communication techniques enable battery-free sensor tags to interoperate with unmodified standard IoT devices, extending a sensor network’s capabilities in a scalable manner. Without requiring additional dedicated infrastructure, the battery-free tags harvest energy from the environment, while the IoT devices provide them with the unmodulated carrier they need to communicate. A schedule coordinates the provision of carriers for the communications of battery-free devices with IoT nodes. Optimal carrier scheduling is an NP-hard problem that limits the scalability of network deployments. Thus, existing solutions waste energy and other valuable resources by scheduling the carriers suboptimally. We present DeepGANTT, a deep learning scheduler that leverages graph neural networks to efficiently provide near-optimal carrier scheduling. We train our scheduler with optimal schedules of relatively small networks obtained from a constraint optimization solver, achieving a performance within 3% of the optimum. Without the need to retrain, our scheduler generalizes to networks 6 × larger in the number of nodes and 10 × larger in the number of tags than those used for training. DeepGANTT breaks the scalability limitations of the optimal scheduler and reduces carrier utilization by up to compared to the state-of-the-art heuristic. As a consequence, our scheduler efficiently reduces energy and spectrum utilization in backscatter networks.

Fuzz testing (“fuzzing”) is a widely-used and effective dynamic technique to discover crashes and security vulnerabilities in software, supported by numerous tools, which keep improving in terms of their detection capabilities and speed of execution. In this paper, we report our findings from using state-of-the-art mutation-based and hybrid fuzzers (AFL, Angora, Honggfuzz, Intriguer, MOpt-AFL, QSym, and SymCC) on a non-trivial code base, that of Contiki-NG, to expose and fix serious vulnerabilities in various layers of its network stack, during a period of more than three years. As a by-product, we provide a Git-based platform which allowed us to create and apply a new, quite challenging, open-source bug suite for evaluating fuzzers on real-world software vulnerabilities. Using this bug suite, we present an impartial and extensive evaluation of the effectiveness of these fuzzers, and measure the impact that sanitizers have on it. Finally, we offer our experiences and opinions on how fuzzing tools should be used and evaluated in the future.

In this paper, we address the support of Time-Slotted Channel Hopping (TSCH) on multiple frequency bands within a single TSCH network. This allows to simultaneously run applications with different requirements on link characteristics and to increase resilience against interference. To this end, we first enable sub-GHz communication in TSCH, which has been primarily defined for the 2.4 GHz band. Thereafter, we propose two designs to support multiple physical layers in TSCH on the same nodes. Our experimental evaluation shows that TSCH is applicable in a wide range of data rates between 1.2 kbps and 1000 kbps. We find that data rates of 50 kbps and below have a long communication range and a nearly perfect link symmetry, but also have a 20x higher channel utilization compared to higher data rates, increasing the risk of collisions. Using these findings, we show the advantages of the multi-band support on the example of synchronization accuracy when exchanging TSCH beacons with a low data rate and application data at a high data rate.

Standard protocols for wireless Internet of Things (IoT) communication must be energy-efficient in order to prolong the lifetimes of IoT devices. Two energy-saving strategies for wireless communication are prevalent within the IoT domain: 1) sleepy devices and 2) radio duty cycling. In this paper, we conduct a comprehensive evaluation as to what types of application scenarios benefit the most from either type of energy-saving strategy. We select the lightweight machine to machine (LwM2M) protocol for this purpose because it operates atop the standard constrained application protocol, and has support for sleepy devices through its Queue Mode. We implement the Queue Mode at both the server side and client side, and design enhancements of Queue Mode to further improve the performance. In our experimental evaluation, we compare the performance and characteristics of Queue Mode with that of running LwM2M in a network stack with the standard time-slotted channel hopping as the duty cycling medium access control protocol. By analyzing the results with the support of an empirical model, we find that each energy-saving strategy has different advantages and disadvantages depending on the scenario and traffic pattern. Hence, we also produce guidelines that can help developers to select the appropriate energy-saving strategy based on the application scenario.

The Internet of Things (IoT) is projected to soon interconnect tens of billions of new devices, in large part also connected to the Internet. IoT devices include both high-end devices which can use traditional go-to operating systems (OSs) such as Linux, and low-end devices which cannot, due to stringent resource constraints, e.g., very limited memory, computational power, and power supply. However, large-scale IoT software development, deployment, and maintenance requires an appropriate OS to build upon. In this paper, we thus analyze in detail the specific requirements that an OS should satisfy to run on low-end IoT devices, and we survey applicable OSs, focusing on candidates that could become an equivalent of Linux for such devices, i.e., a one-size-fits-most, open source OS for low-end IoT devices.

Research and industrial installations have shown that the on-board temperature of wireless sensor nodes deployed outdoors can experience high fluctuations over time with a large variability across the network. These variations can have a strong impact on the efficiency of low-power radios and can significantly affect the operation of communication protocols, often compromising network connectivity. In this paper, we show the adverse effects of temperature on communication protocols and propose techniques to increase their resilience. First, we experimentally show that fluctuations of the on-board temperature of sensor nodes reduce the efficiency of data link layer protocols, leading to a substantial decrease in packet reception rate and to a considerable increase in energy consumption. Second, we investigate the reasons for such performance degradation, and show that high on-board temperatures reduce the effectiveness of clear channel assessment, compromising the ability of a node to avoid collisions and to successfully wake-up from low-power mode. After modelling the behaviour of radio transceivers as a function of temperature, we propose two mechanisms to dynamically adapt the clear channel assessment threshold to temperature changes, thus making data link layer protocols temperature-aware. An extensive experimental evaluation shows that our approaches considerably increase the performance of a network in the presence of temperature variations commonly found in real-world outdoor deployments, with up to 42% lower radio duty-cycle and 87% higher packet reception rate.

Real-world experiments have shown that the transmission power and the received signal strength of low-power radio transceivers used in sensornets decrease when temperature increases. We analyze how this phenomenon affects the network layer, and find that temperature fluctuations may cause undesirable behavior by sensornet routing protocols such as CTP and RPL. Furthermore, we present an approach to make these protocols robust to temperature fluctuations by augmenting the ETX link metric with temperature hints.

Low-power wireless networks must leverage radio duty cycling to reduce energy consumption, but duty cycling drastically increases the risk of radio collisions, resulting in power-expensive retransmissions or data loss. We present Strawman, a contention resolution mechanism designed for low-power duty-cycled networks that experience traffic bursts. Strawman efficiently resolves network contention, mitigates the hidden terminal problem, and has zero overhead unless activated to resolve data collisions. Our testbed experiments show that Strawman instantaneously provides increased network capacity when needed, allocates the available bandwidth evenly among contenders, and increases energy efficiency in multi-hop collection networks compared to the traditionally used random backoff.

Standardization has driven interoperability at multiple layers of the stack, such as the routing and application layers, standardization of radio duty cycling mechanisms have not yet reached the same maturity. In this work, we pitch the two de facto standard flavors of sender-initiated radio duty cycling mechanisms against each other: ContikiMAC and TinyOS LPL. Our aim is to explore pragmatic interoperability mechanisms at the radio duty cycling layer. This will lead to better understanding of interoperability problems moving forward, as radio duty cycling mechanisms get standardized. Our results show that the two flavors can be configured to operate together but that parameter configuration may severely hurt performance.

Interoperability is essential for the commercial adoption of wireless sensor networks. However, existing sensor network architectures have been developed in isolation and thus interoperability has not been a concern. Recently, IP has been proposed as a solution to the interoperability problem of low-power and lossy networks (LLNs), considering its open and standards-based architecture at the network, transport, and application layers. We present two complete and interoperable implementations of the IPv6 protocol stack for LLNs, one for Contiki and one for TinyOS, and show that the cost of interoperability is low: their performance and overhead is on par with state-of-the-art protocol stacks custom built for the two platforms. At the same time, extensive testbed results show that the ensemble performance of a mixed network with nodes running the two interoperable stacks depends heavily on implementation decisions and parameters set at multiple protocol layers. In turn, these results argue that the current industry practice of interoperability testing does not cover the crucial topic of the performance and motivate the need for generic techniques that quantify the performance of such networks and configure their run-time behavior.

Interoperability is key to widespread adoption of sensor network technology, but interoperable systems have traditionally been difficult to develop and test. We demonstrate an interoperable system development and performance diagnosis environment in which different systems, different software, and different hardware can be simulated in a single network configuration. This allows both development, verification, and performance diagnosis of interoperable systems. Estimating the performance is important since even when systems interoperate, the performance can be sub-optimal, as shown in our companion paper that has been conditionally accepted for SenSys 2011.

We make the case for a sensor network model in which each mote stores sensor data locally, and provides a database query interface to the data. Unlike TinyDB and Cougar, in which a sink node provides a database-like front end for filtering the current sensor values from a data collection network, we propose that each sensor device should run its own database system. We present Antelope, a database management system for resource-constrained sensors. Antelope provides a dynamic database system that enables run-time creation and deletion of databases and indexes. Antelope uses energy-efficient indexing techniques that significantly improve the performance of queries. The energy cost of a query that selects 100 tuples is less than the cost of a single packet transmission. Moving forward, we believe that database techniques will be increasingly important in many emerging applications.

RPL is the IETF candidate standard for IPv6 routing in low-power wireless sensor networks. We present the first experimental results of RPL which we have obtained with our ContikiRPL implementation. Our results show that Tmote Sky motes running IPv6 with RPL routing have a battery lifetime of years, while delivering 0.6 packets per second to a sink node.

We demonstrate low-power IP-based sensor networks by showing a system that interacts with the sensor network using a RESTful web service interface. The sensor data is displayed with overlaid 3D graphics on top of a live camera feed, so-called augmented reality. The augmented reality application is built with off-the-shelf components with no sensor network-specific code. The IP-based sensor network runs the Contiki operating system.

Low-power networked devices, such as sensors and actuators, are becoming a vital part of our everyday infrastructure. Being networked, the continued development of these systems needs involvement from the networking community. We present a framework for simulation, experimentation, and evaluation of routing mechanisms for low-power IPv6 networking. The framework provides a detailed simulation environment for low-power routing mechanisms, and allows the system to be directly uploaded to a physical testbed for experimental measurements.

Wireless sensor networks are being considered for use in industrial process and control environments. Unlike traditional deployment scenarios for sensor networks, in which energy preservation is the main design principle, industrial environments stress worker safety and uninterrupted production. To fulfill these requirements, sensor networks must be able to provide performance guarantees for radio communication. In this paper, we consider as a case study the deployment of a sensornet in an oil refinery in Portugal, where sensor nodes are deployed outdoors and might experience high temperature fluctuations. We investigate how the variations of ambient temperature influence data delivery performance and link quality in low-power radio communications. We also study the impact that specific implementation requirements, such as the ATEX fire-safety regulations, can have on the design of the overall network. Our experiments show that temperature directly affects the communication between sensor nodes, and that significantly less transmission power is required at low temperatures. We further illustrate that it is possible to save up to 16% energy during nights and cold periods of the year, while still ensuring reliable communication among sensor nodes. In view of these experimental results, we elaborate on how the temperature influences both the design and the deployment of wireless sensor networks in industrial environments.

Power consumption is of utmost concern in sensor networks. Researchers have several ways of measuring the power consumption of a complete sensor network, but they are typically either impractical or inaccurate. To meet the need for practical and scalable measurement of power consumption of sensor networks, we have developed a cycle-accurate simulator, called COOJA/MSPsim, that enables live power estimation of systems running on MSP430 processors. This demonstration shows the ease of use and the power measurement accuracy of COOJA/MSPsim. The demo setup consists of a small sensor network and a laptop. Beside gathering software-based power measurements from the motes, the laptop runs COOJA/MSPsim to simulate the same network.We visualize the power consumption of both the simulated and the real sensor network, and show that the simulator produces matching results.

Software in deployed sensor networks needs to be updated to introduce new functionality or to fix bugs. Reducing dissemination time is important because the dissemination disturbs the regular operation of the network. We present a method for reducing the dissemination time and energy consumption based on compression of native code modules. Code compression reduces the size of the software update, but the decompression on the sensor nodes requires processing time and energy. We quantify these trade-offs for seven different compression algorithms. Our results show that GZIP has the most favorable trade-offs, saving on average 67% of the dissemination time and 69% of the energy in a multi-hop wireless sensor network.

With emerging IPv6-based standards such as 6LowPAN and ISA100a, full IPv6 sensor networks are the next major step. With millions of deployed embedded IPv6 devices, interoperability is of major importance, both within the sensor networks and between the sensors and the Internet hosts. We present uIPv6, the first IPv6 stack for memory-constrained devices that passes all Phase-1 IPv6 Ready certification tests. This is an important step for end-to-end interoperability between IPv6 sensors and any IPv6 capable device. To allow widespread community adoption, we release uIPv6 under a permissive open source license that allows both commercial and non-commercial use.

In this paper we present the design and implemen tation of a small-scale marine sensor network. The network monitors the temperature in the Baltic Sea on different heights from the water surface down to the bottom. Unlike many other wireless sensor networks, this network contains both a wired and a wireless part. One of the major challenges is that the network is hard to access after its deployment and hence both hard- and software must be robust and reliable. We also present the design of an advanced buoy system featuring a diving unit that achieves a better vertical resolution and discuss remaining challenges of sensor networking in aquatic environments.

Being able to estimate the energy consumption of sensor nodes is essential both for evaluating existing sensor network mechanisms and for constructing new energy-aware mechanisms. We present a software-based mechanism for estimating the energy consumption of sensor node at run-time. Unlike previous energy estimation mechanisms, our mechanism does not require any additional hardware components or add-ons.