Wireless Sensor Networks

In recent years Wireless Sensor Networks (WSNs) have been deployed in wide range of applications from the health and environment monitoring to building and industrial control. However, the pace of prevalence of WSN is slower than anticipated by the research community due to several reasons including required embedded systems expertise for developing and deploying WSNs; use of proprietary protocols; and limits in scalability and reliability. In this paper we propose PyFUNS (Python-based Framework for Ubiquitous Networked Sensors) to address these challenges. PyFUNS handles low level and networking functionalities, using the services provided by Contiki, and leaves to the user only the task of application development in the form of Python scripts. This approach reduces required expertise in embedded systems to develop WSN based applications. PyFUNS also uses 6LoWPAN and CoAP standard protocols to enable interoperability and ease of integration with other systems, pursuing the Internet of Things vision. Through a real implementation of PyFUNS in two constrained platforms we proved its feasibility in mote devices, as well as its performance in terms of control delay, energy consumption and network traffic in several network topologies. As it is possible with PyFUNS to easily compare performance of different deployments of distributed application, PyFUNS can be used to identify optimal design of distributed application.

Counting and tracking multiple targets by binary proximity sensors (BPS) is known difficult because a BPS in “on” state cannot distinguish how many targets are presenting in its sensing range. Existing approaches investigated target counting by utilizing joint readings of a network of BPSs, called a snapshot [2,11]. A recent work [14] presented a snapshot-based target counting lower bound. But counting by individual snapshot has not fully utilized the information between the sequential readings of BPSs. This paper exploits the spatial and temporal dependency introduced by a sequence of snapshots to improve the counting bounds and resolution. In particular, a dynamic counting scheme which considers the dependency among the snapshots were developed. It leads to a dynamic lower bound and a dynamic upper bound respectively. Based on them, an improved precisely counting condition was presented. Simulations were conducted to verify the improved counting limits, which showed the improvements than the snapshot-based methods.

Sensor networks operating in the 2.4 GHz band often face cross-technology interference from co-located WiFi and Bluetooth devices. To enable effective interference mitigation, a sensor network needs to know the type of interference it is exposed to. However, existing approaches to interference detection are not able to handle multiple concurrent sources of interference. In this paper, we address the problem of identifying multiple channel activities impairing a sensor network’s communication, such as simultaneous WiFi traffic and Bluetooth data transfers. We present SpeckSense, an interference detector that distinguishes between different types of interference using a unsupervised learning technique. Additionally, SpeckSense features a classifier that distinguishes between moderate and heavy channel traffic, and also identifies WiFi beacons. In doing so, it facilitates interference avoidance through channel blacklisting. We evaluate SpeckSense on common mote hardware and show how it classifies concurrent interference under real-world settings. We also show how SpeckSense improves the performance of an existing multichannel data collection protocol by 30%.

In order to provide useful energy saving recommendations, energy management systems need a deep insight in the context of energy consumption. Getting those insights is rather difficult. Either exhaustive user questionnaires or the installation of hundreds of sensors are required in order to acquire this data. Measuring the energy consumption of a household is however required in order to find and realize saving potentials. Thus, we show how to gain insights in the context of energy consumption directly from the energy consumption profile. Our proposed methods are capable of determining the user’s current activity with an accuracy up to 98% as well as the user’s current place in a house with an accuracy up to 97%. Furthermore, our solution is capable of detecting anomalies in the energy consumption behavior. All this is mainly achieved with the energy consumption profile.

We present SocialSense , a collaborative smartphone based speaker and mood identification and reporting system that uses a user’s voice to detect and log his/her speaking and mood episodes. SocialSense collaboratively works with other phones that are running the app present in the vicinity to periodically send/receive speaking and mood vectors to/from other users present in a social interaction setting, thus keeping track of the global speaking episodes of all users with their mood. In addition, it utilizes a novel event-adaptive dynamic classification scheme for speaker identification which updates the speaker classification model every time one or more users enter or leave the scenario, ensuring a most updated classifier based on user presence. Evaluation of using dynamic classifiers shows that SocialSense improves speaker identification accuracy by 30% compared to traditional static speaker identification systems, and a 10% to 43% performance boost under various noisy environments. SocialSense also improves the mood classification accuracy by 4% to 20% compared to the baseline approaches. Energy consumption experiments show that its device daily lifetime is between 10-14 hours.

Human mobility is a complex pattern of movements and activities that are based on some underlying semantics of human behavior. In order to construct accurate models of human mobility, this semantic behavior needs to be unearthed from the data sensed as a human being moves around and visits certain classes of locations such as home, work, mall, theater, restaurant etc. The ideal data for understanding the semantics of mobility would constitute timestamped mobility traces with detailed geographic locations with annotations about the type of each location. One way of achieving this is by following a hybrid strategy of participatory sensing (with each person carrying a wireless sensor device) and deploying static sensors at each location of interest – the contacts between the mobile and (annotated) static sensors can be logged at each location, and then collated to form an appropriate mobility traces. For example, a person can connect with his mobile phone over Bluetooth or WiFi to a local hotspot while checking into FourSquare at a restaurant. In the absence of static sensors, a person may manually annotate the places he visits on his device over time. However, most mobility traces consist of network connectivity data from cell phones (e.g., contact with towers) which lack detailed geographic locations and are ambiguous, noisy and unlabeled. Thus, it is important to extract the semantics of mobility that is latent in the available contact traces. To this end, we propose in this paper the concept of Probabilistic Latent Semantic Trajectories (PLST), an unsupervised approach to extract semantically different locations and sequential patterns of mobility from such traces. PLST extracts semantic locations as contextually co-occurring network elements (cell towers and Bluetooth devices) and models the behavior of their sequence. PLST extracts distinct locations with spatial, temporal and semantic coherency and can be used for accurate prediction of the next place a user visits. PLST also analyzes the complexity of mobility traces using information theoretic metrics to study the underlying structure and semantic content in mobility traces. This semantic content can be extracted allowing us to investigate mobility patterns in a completely unsupervised manner.

Limited memory capacity is one of the major constraints in Delay Tolerant Wireless Sensor Networks. Efficient management of the memory is critical to the performance of the network. This paper proposes a novel buffer management algorithm, SmartGap, a Quality of Information (QoI) targeted buffer management algorithm. That is, in a wireless sensor network that continuously measures a parameter which changes over time, such as temperature, the value of a single packet is governed by an estimation of its contribution to the recreation of the original signal. Attractive features of SmartGap include a low computational complexity and a simplified reconstruction of the original signal. An analysis and simulations in which the performance of SmartGap is compared with the performance of several commonly used buffer management algorithms in wireless sensor networks are provided in the paper. The simulations suggest that SmartGap indeed provides significantly improved QoI compared the other evaluated algorithms.

Because of the dramatic changes in topology and frequently interrupted connections between nodes, messages in delay-tolerant networks are forwarded in the store-carry-forward approach. Routing methods in such an environment tend to increase the number of messages to improve the delivery ratio. However, excessive message copies lead to buffer overflows because of limited storage space. Therefore, an efficient message-scheduling and drop strategy is vital to maximizing network resources, especially when bandwidth is limited and message sizes differ. We developed a theoretical framework called the knapsack-based message scheduling and drop strategy in theory (KMSDT) based on epidemic message dissemination. To improve the delivery ratio, this strategy sorts message copies by utility per unit and, if buffer overflows occur, it decides which messages to drop based on the solution to the knapsack problem. Furthermore, we developed a practical framework called the knapsack-based message scheduling and drop strategy in practice (KMSDP). Rather than collecting global statistics as done in the KMSDT, KMSDP estimates all parameters by using locally collected statistics. Simulations based on synthetic trace are done in ONE. Results show that, without affecting the average delay or overhead ratio, KMSDP and KMSDT achieve a better delivery ratio than other congestion-control strategies.

In the last years the Low Power and Lossy Networks (LLNs), have become more and more popular. LLNs are inherently dynamic - nodes move, associate, disassociate or experience link perturbations. In order to meet the specific requirements for LLNs, the IETF has developed a new routing protocol - IPv6 Routing Protocol for Low-Power and Lossy Networks (RPL) that routes packets inside LLNs. RPL has to work in such dynamic environment and mechanisms that can mitigate such conditions are suggested in the standard such as Neighbor Unreachability Detection or Bidirectional Forwarding Detection. In this article, we show that such mechanisms fail to prevent serious node disconnection, which significantly increases the packet loss and leads to severe underachievements. To provide RPL the ability to mitigate network dynamics generated by node disconnection, we therefore propose a new cross-layer protocol operating at layers 2 and 3 known as Mobility-Triggered RPL (MT-RPL). MT-RPL benefits from the X-Machiavel MAC protocol that favors medium access to mobile devices. X-Machiavel has been extended to trigger RPL operations in order to maintain efficient connectivity with the network. MT-RPL is evaluated together with Neighbor Unreachability Detection and Bidirectional Forwarding Detection through an extensive simulation campaign. Results show that MT-RPL significantly reduces the disconnection time, which increases the packet delivery ratio and reduces energy consumption per data packet.

We study the feasibility of leveraging the sensors embedded on mobile devices to enable a user authentication mechanism that is easy for users to perform, but hard for attackers to bypass. The proposed approach lies on the fact that users perform gestures in a unique way that depends on how they hold the phone, and on their hand’s geometry, size, and flexibility. Based on this observation, we introduce two new unlock gestures that have been designed to enable the phone’s embedded sensors to properly capture the geometry and biokinetics of the user’s hand during the gesture. The touch sensor extracts the geometry and timing of the user hand, while the accelerometer and gyro sensors record the displacement and rotation of the mobile device during the gesture. When combined, a sensor fingerprint for the user is generated. In this approach, potential attackers need to simultaneously reproduce the touch, accelerometer, and gyro sensor signatures to falsely authenticate. Using 5000 gestures recorded over two user studies involving a total of 70 subjects, our results indicate that sensor fingerprints can accurately differentiate users while achieving less than 2.5% false accept and false reject rates. Attackers that directly observe the true user authenticating on a device, can successfully bypass authentication only 3% of the time.

Trickle is a polite gossip algorithm for managing communication traffic. It is of particular interest in low-power wireless networks for reducing the amount of control traffic, as in routing protocols (RPL), or reducing network congestion, as in multicast protocols (MPL). Trickle is used at the network or application level, and relies on up-to-date information on the activity of neighbors. This makes it vulnerable to interference from the media access control layer, which we explore in this paper. We present several scenarios how the MAC layer in low-power radios violates Trickle timing. As a case study, we analyze the impact of CSMA/CA with ContikiMAC on Trickle’s performance. Additionally, we propose a solution called Cleansing that resolves these issues.

We introduce the concept of Featurecast with addressing and routing based on node features defined as predicates. For instance, we can send a packet to the address composed of features {temperature and Room D} to reach all nodes with a temperature sensor located in Room D. Each node constructs its address from the set of its features and disseminates it in the network so that intermediate nodes can build routing tables. In this way, a node can send a packet to a set of nodes matching given features. Our experiments and evaluation of this scheme show very good performance compared to Logical Neighborhoods (LN) and IP multicast with respect to the memory footprint and message overhead.

Many data collection protocols have been proposed to cater for the energy-efficient flow of sensor data from distributed sources to a sink node. However, the transmission of control commands from the sink to one or only a small set of nodes in the network is generally unsupported by these protocols. Supplementary protocols for packet routing and data dissemination have been developed to this end, although their energy requirements commonly thwart the low-power nature of data collection protocols. We tackle this challenge by presenting RoCoCo in this paper. It combines data collection and dissemination by extending the low-energy ORiNoCo collection protocol by means to reconfigure subsets of nodes during runtime. Synergistically leveraging existing message types, RoCoCo allows for the definition of multicast recipient groups and forwards commands to these groups in an opportunistic fashion. Relying on Bloom filters to define the recipient addresses, RoCoCo only incurs small memory and energy overheads. We confirm its feasibility by evaluating the introduced delays, command success rates, and its energy overhead in comparison to existing collection/dissemination protocols.

Wireless sensor-actuator networks (WSANs) offer an appealing communication technology for process automation applications. However, such networks pose unique challenges due to their critical demands on reliability and real-time performance. While industrial WSANs have received attention in the research community, most published results to date focused on the theoretical aspects and were evaluated based on simulations. There is a critical need for experimental research on this important class of WSANs. We developed an experimental testbed by implementing several key network protocols of WirelessHART, an open standard for WSANs widely adopted in the process industries, including multi-channel TDMA with shared slots at the MAC layer and reliable graph routing supporting path redundancy. We then performed a comparative study of the two alternative routing approaches adopted by WirelessHART, namely source routing and graph routing. Our study shows that graph routing leads to significant improvement over source routing in term of worst-case reliability, at the cost of longer latency and higher energy consumption. It is therefore important to employ graph routing algorithms specifically designed to optimize latency and energy efficiency.

We propose a Corrupt Packet Recycling (CPR) approach for WSN that processes and forwards partially-corrupt packets over multiple hops without necessitating their complete recovery. We motivate this approach with two insights: address-agnostic routing in WSN can forgive header errors since intermediate nodes know the next hop and the destination; and that payload errors can be either interpolated, due to error-tolerant nature of information in WSN applications, or rectified using spatio-temporal redundancies. CPR, without introducing any transmission overhead, improves information delivery rate by up to 4×.

Bug hunting in sensor networks is challenging: Bugs are often prompted by a particular, complex concatenation of events. Moreover, dynamic interactions between nodes and with the environment make it time-consuming to track and reproduce a bug. We introduce LibReplay to ease bug hunting in sensor networks: it provides (1) lightweight and flexible logging and (2) deterministic replay. LibReplay logs function calls to and from the application or another code of interest. It enables deterministic replay of execution traces in a controlled environment such as a full-system simulator. This allows the user to benefit from well-established debugging tools such as stepping through code, breakpoints, or watchpoints. We show that the lightweight architecture of LibReplay provides the benefits of replay debugging at an efficiency that is comparable to traditional logging tools, which commonly do not allow replay debugging.

Low-power wireless networks, especially in outdoor deployments, are exposed to a wide range of temperatures. The detrimental effect of high temperatures on communication quality is well known. In this paper, we use a testbed with self-made temperature control devices to investigate the effects of temperature on several communication-relevant metrics. The analyses both confirm some previously published results and demonstrate deviations from others. Based on these results, we propose a Reed–Solomon-based FEC scheme to mitigate the negative effects of temperature and provide results suggesting that such a scheme is both feasible and advantageous.

In this paper, we present a novel approach to study and reveal network and protocol information from energy instrumentation in wireless sensor network. Unlike prior approaches which focused on analyzing the aggregate statistics of energy efficiency of a network or a protocol, our approach aims at revealing network protocols, application workloads, and topology information by fine-grained energy instrumentation on the nodes. We design a set of features based on various aspects of energy data and use those features to classify and reveal network activity. Results from experiments on three testbeds indicate that our approach can achieve 97% accuracy to identify the routing protocols, and infer the network topology with 98% accuracy.

Low-power wireless actuation is attracting interest in many domains, yet it is significantly less investigated than its sensing counterpart, especially in large-scale scenarios. As a consequence, guidelines about which protocol, among the few existing ones, is best suited to a given scenario are generally lacking.

In this paper, we investigate the relative performance of simple dissemination-based solutions against the standard, state-of-the-art RPL protocol. These choices of protocols are motivated concretely by our involvement in the deployment of a large-scale infrastructure for smart city applications, which directly informs our evaluation, where we use the actual network topology.

Our findings, albeit in a specific scenario, suggest that RPL still leaves much to be desired w.r.t. actuation. Two out of the three RPL implementations we considered exhibited unacceptable performance when used out-of-the-box. Even after some tuning and debugging, simple, dissemination-based solutions perform surprisingly better under several conditions. These findings motivate further research on the topic of large-scale low-power wireless actuation.

Conventional transmission in IPv6 over Networks of Resource Constrained Nodes (6lo) favours fixed-size packets and results in low network performance when bulk data transmission is required by applications, for example firmware updating. To tackle this problem, we first investigate performance of bulk data transmission through large packets and make two important observations. Then we propose an adaptive mechanism at IP layer to dynamically adjust packet size in 6lo. We implemented the mechanism on Contiki OS and evaluated it through a series of experiments in Cooja. Experimental results demonstrate that our mechanism outperforms Contiki standard implementation significantly from both reliability and goodput under various network conditions.