To ensure reliability and service availability, next-generation networks are expected to rely on automated anomaly detection systems powered by advanced machine learning methods with the capability of handling multi-dimensional data. Such multi-dimensional, heterogeneous data occurs mostly in today’s Industrial Internet of Things (IIoT), where real-time detection of anomalies is critical to prevent impending failures and resolve them in a timely manner. However, existing anomaly detection methods often fall short of effectively coping with the complexity and dynamism of multi-dimensional data streams in IIoT. In this paper, we propose an adaptive method for detecting anomalies in IIoT streaming data utilizing a multi-source prediction model and concept drift adaptation. The proposed anomaly detection algorithm merges a prediction model into a novel drift adaptation method resulting in accurate and efficient anomaly detection that exhibits improved scalability. Our trace-driven evaluations indicate that the proposed method outperforms the state-of-the-art anomaly detection methods by achieving up to an 89.71% accuracy (in terms of Area under the Curve (AUC)) while meeting the given efficiency and scalability requirements.

The growth of streaming data originating from Internet of Things (IoT)-based Industry 4.0 opens doors to real-time analytics of time-sensitive services. However, this ever-increasing amount of data inevitably leads to anomalies, resulting in considerable risks for time-sensitive applications. Thus, real-time detection of anomalies is critical to prevent impending failures and resolve them in time. Given that the problem is to detect application-level anomalies in real time, we develop a deep learning-based technique, which integrates time-series data inference with a Long-Short Term Memory (LSTM)-based prediction model. Our proposed method relies on a novel metric called Sequence Inconsistency Distance (SID), which determines the abnormality likelihood of a target record in real time. Our trace-driven evaluations indicate that the proposed method achieves up to a 92.6% performance gain compared to the current state-of-the-art anomaly detection methods in terms of true positive and false positive rate while meeting the essential efficiency requirements.

The Industrial Internet of Things (IIoT) applications is usually associated with stringent latency requirements. An anomaly in an IIoT edge network deteriorates the performance and thus needs a real-time Root Cause Analysis (RCA) to identify the anomalous node and provide robust network infrastructure. In this paper, we present an automated, real-time RCA technique to identify the network-level root cause nodes. We use a deep learning-based approach, exploiting a Graph Neural Network (GNN) to identify root cause nodes. In GNN-RCA, we used a sampling technique and optimized aggregator function to reduce detection time. We have shown that the proposed GNNRCA method outperforms the existing benchmarks in terms of classification score and execution time.

This paper investigates the precise identification of physical phenomena in the Internet of Things (IoT) context, which is one of the main challenges when dealing with the massive scale of IoT data. For this, we use information theory quantifiers in the characterization and classification of physical phenomena to minimize the effects of the lack of proper descriptions and the high heterogeneity of IoT sensors. Thus, by understanding the dynamics behind physical phenomena, we perform the classification of sensor data based on their expected behavior, not their data points. By using a simple classification algorithm, we show that the behavioral dynamics of some physical phenomena are more affected by different geographical regions than others. This gives a classification accuracy of 75% when all phenomena are considered and of 93% when considering only the invariant ones, with a worst case of false positives of 12%. This result indicates the high potential of our technique to correctly identify physical phenomena from sensor data, a fundamental issue for several applications, even in an unreliable IoT environment.

Vehicular ad-hoc networks are a promising type of networks that allows the communication between vehicles with the goal to promote safe and efficient traffic. During the trips, vehicles can communicate with each other and with other networks by interacting with them and with the cell phone network. In VANETs, data dissemination is a general task required by many services, which consists of the delivery of data messages to a group of vehicles. In this type of communication, the knowledge of the interactions among vehicles can help to improve the performance of data dissemination. Thus, in this work, we propose the TBD (Trajectory Based Dissemination), a solution that transmits the information on the network according to the density of vehicles in the path from the source to the region of interest. Simulation results evaluated in scenarios of various densities have shown that it is possible to reduce the number of messages transmitted with a high delivery ratio.

The broadcast storm problem is characterized by unwanted packet retransmission, causing frequent flooding, contention, and collisions that compromise network performance. This paper presents a new vehicular message dissemination protocol named CN-vP (Complex Network- vehicular Protocol). It is based on the observation and analysis of complex network metrics. In order to mitigate the sending of unnecessary messages, the present work combines probabilistic and delay approaches together with a better knowledge of the network, in particular, the neighboring vehicles of a certain transmitter. Once the three best relays have been chosen by probabilistic calculation, the waiting intervals are estimated, for each retransmission, in order to avoid packet flooding. The analysis of the simulation results shows that the solution developed allows a more assertive decision, thus mitigating the problem of broadcast storm.

In D2D opportunistic networks, nodes need to cooperate acting as relays for transmitting messages to other nodes according to an opportunistic routing algorithm. To store these messages until they are propagated, each node uses a buffer with limited capacity. However, when multiple messages are forwarded in the network, the number of incoming messages may exceed the nodes' capacity, causing a buffer overflow. In this scenario, message dropping policies are very important to this problem, because when a message is dropped, there is a chance that other copies of this message still exist in the network. In this work, we propose a new buffer management algorithm for opportunistic routing in D2D networks named ST-Drop (Space-Time-Drop). We have evaluated our solution in three different types of opportunistic routing algorithms: epidemic-based, probabilistic, and social-aware. We have conducted simulations using two different publicly available data sources and considered different network traffic loads. Compared to other message drop policies, ST-Drop obtained the highest message delivery ratio in all considered scenarios and the lowest overhead when applied to the state-of-art social-aware and probabilistic routing algorithms, namely, Bubble Rap and Prophet.

Vehicular network applications are emerging to bring many benefits to the users during their journey. However, the design and deployment of such applications require studies that provide insightful information about the network formed by vehicles. To this end, in the last years researchers have been characterizing real mobility traces collected from taxis equipped with GPS devices. However, these traces present temporal and spatial gaps, as it was demonstrated in other studies. In this work, we compare the real original traces with their calibrated version (i.e., with no gaps) to show that the existing gaps expressively affect key metrics such as contact duration, inter-contact time, and network capacity. As result, we contribute to the research community by presenting important analysis of real vehicular network topology and by showing the importance of calibrating the traces before evaluating vehicular network applications.

The Medium Access Control layer in Wireless Sensor Networks handles the failures that occur when a node sends a frame to a dormant neighbor. Some protocols work in such a way that when a node shares its working schedule with its neighbors, each sender node schedules its transmission to the moment when the next node is awake. In others, the nodes do not share information and when one needs to send a frame, it first discovers the time when the next node will be awake. Both approaches are limited due to additional transmissions. This work proposes a hybrid policy that combines the existing ones in order to minimize their costs. Simulation results reveal that our solution often reduces the number of transmissions and energy consumption of the existing policies in the evaluated scenarios.

Overtaking vehicles is a risky task and could cause serious accidents, especially on two-lane highways. There are various efforts in order to make this a safer task. An alternative to this is the use of communication between vehicles and advanced techniques to decide the safest time for overtaking. Thus, in this paper we propose a driver assistance service that uses real-time information transmitted among vehicles and formal methods to calculate the trajectories and to assign the optimal behavior to overtake. For this, we used Probabilistic Model Checking (PMC), which explores all the possibilities of the system indicating the correct configuration to the vehicles involved. In the evaluated study case, the results showed that it is possible to find a secure configuration in a scenario with three vehicles and a collision's probability of 98%.

Everyday, vehicles transit in a city and along their trajectories, they encounter other vehicles. The frequency of these encounters is influenced by many factors, such as: vehicle speed, destinations, traffic conditions, and the period of the day. However, these factors are justified by the public roads limits and the driver's behavior. The people present daily routines and similar behaviors that have a great impact in the daily traffic evolution. In this work, we present a numerical analysis of real and realistic data sets that describe the mobility of a set of vehicles. Social metrics are computed, and the results obtained are compared to random graphs in the direction to verify if vehicular network presents a social behavior. Finally, we discuss new social perspectives in vehicular networks.

People have routines and their mobility patterns vary during the day, which have a direct impact on vehicular mobility. Therefore, proto- cols and applications designed for Vehicular Ad Hoc Networks need to adapt to these routines in order to provide better services. With this issue in mind, in this work, we propose a data dissemination solution for these networks that considers the daily road traffic variation of large cities and the relationship among vehicles. The focus of our approach is to select the best vehicles to rebroadcast data messages according to social metrics, in particular, the clustering coefficient and the node degree. Moreover, our solution is designed in such a way that it is completely independent of the perceived road traffic density. Simulation results show that, when compared to related protocols, our proposal provides better delivery guarantees, reduces the network overhead and possesses an acceptable delay.

A wireless sensor network can be used to collect and process environmental data, which is often of multivariate nature. This work proposes a multivariate sampling algorithm based on component analysis techniques in wireless sensor networks. To improve the sampling, the algorithm uses component analysis techniques to rank the data. Once ranked, the most representative data is retained. Simulation results show that our technique reduces the data keeping its representativeness. In addition, the energy consumption and delay to deliver the data on the network are reduced.

Vehicular Mobility is strongly influenced by the speed limits and direction of the public roads. At the same time, the driver's behavior produces great influences in vehicular mobility. People tend to go to the same places, at the same day period, through the same trajectories, which le ad them to the appearance of driver's daily routines. These routines lead us to the study of mobility in VANETs under a social perspective and to investigate how effective is to explore social interactions in this kind of network. The work herein proposed presents the characterization and evaluation of a realistic vehicular trace found in literature. Our aim is to study the vehicles' mobility in accordance to social behaviors. With our analysis is possible to verify the existence of regularity and common interests among the drivers in vehicular networks. Finally, we discuss how the social metrics may be used to improve the performance of protocols and services in Vehicular Networks.

Wireless sensor networks are designed to be deployed in the physical environment to monitor a wide variety of real-world phenomena. These networks are composed of various small sensor nodes deposited in a given area. In these networks, the battery replacement is unfeasible and thus the energy is a critical resource that must be considered by all protocols. In order to reduce the energy consumption, protocols for MAC layer use techniques that turn off the radio of sensor nodes. However, in many protocols, all sensor nodes operate on the same static schedule in which they wake up periodically at the same fixed interval of time. This scheme might not be appropriate in dynamic scenarios in which the traffic varies. In these situations, it is desirable to adjust the operation mode of sensor nodes according with the network traffic. In this work, we present the Sensor-MAC with Dynamic Duty Cycle (SMAC-DDC), a new MAC protocol that adjusts the duty cycle of sensor nodes according with the network traffic. Through a cross layer interaction, SMAC-DDC uses the information of the network layer to adjust the value of the duty cycle of sensor nodes. Simulations results show that the proposed protocol reduces the latency and the energy consumption of a network, keeping the delivery ratio.

In this work we propose a framework based on the small world features to design Heterogeneous Sensor Network (HSN) topologies with QoS. The framework creates three different topologies and each topology has its goals related to the latency and energy consumption during data communication. We also propose a routing algorithm that uses the created topologies in order to provide QoS in HSN. Simulation results showed that the three topologies can provide different levels of QoS that can be used in many sensor network applications. We also compare our framework to the literature topology to provide QoS in HSN and our framework presents better tradeoff between energy consumption and latency during data communication.

A typical Wireless Sensor Network (WSN) assumes a homogeneous set of nodes in terms of capabilities. However, this kind of network suffer from poor fundamental limits of latency during the data communication. Another model of WSN assumes a heterogeneous set of nodes with different capabilities (especially in terms of communication range and energy reserves) called Heterogeneous Sensor Networks (HSNs). In this work, we propose a tree-based approach based on small world concepts to design HSNs. The proposed model introduces an adjustable search space in order to find H-sensors to create shortcuts between them. The endpoints of these shortcuts are nodes with more powerful communication range and energy reserves to support the long communication range. The model was designed in such a way that the created shortcuts among the powerful nodes create a connected topology among them. As a consequence of this, it is not necessary that the nodes with lower capabilities forward messages as a bridge to interconnect the shortcuts. Simulation results showed that with just a few powerful nodes, a wireless sensor network can be tuned into a HSN with small world features. Also, when the shortcut addition creates a connected topology among the H-sensors, the data communication latency is significantly reduced.

In this work, we propose a small world model to design Heterogeneous Sensor Networks (HSNs). The proposed model takes into account the communication pattern of this network to create shortcuts directed to the monitoring node, decreasing data communication latency. The endpoints of these shortcuts are nodes with more powerful communication range and energy reserves to support the long communication range. However, this kind of communication causes a high interference in the wireless channel. For this, the proposed model uses the multi-interface and multi-channel capability of the MAC layer to reduce the number of collisions during data communication. Simulation results showed that with just a few powerful nodes, a wireless sensor network can be tuned into a HSN with small world features. Also, when the shortcuts are directed to the monitoring node and assigned to a different wireless channel, the number of collisions and the data communication latency are reduced.

In wireless sensor networks, there are some network functions and applications that need to know the local time of a node in order to timestamp a sensed data. In general, it may be unfeasible to have a synchronization hardware in each sensor. In this work, we propose a heterogeneous wireless sensor network model based on small world concepts to improve synchronization algorithms for wireless sensor networks. This improvement has two goals: (i) reduce the synchronization error for large wireless sensor networks, and (ii) increase the resilience of synchronization algorithms in the presence of node failures. Therefore, we modified and evaluated the Flooding Time Synchronization Protocol in the proposed model. Simulation results show that the use of small world based models in wireless sensor networks can significantly reduce the synchronization error and also increase the resilience in the synchronization problem.

Data communication represents the greatest energy cost of wireless sensor networks. Both MAC and network layers of the protocol stack are responsible for data communication in these networks. This work proposes the Receiver- based Medium Access Control (Rb-MAC) protocol to be used in a combined solution with a receiver-based routing scheme. The main advantage of combining receiver-based MAC and routing protocols is that this solution does not use neighbor tables and, consequently, it saves energy when compared with MAC and routing protocols that use these tables. Simulation results reveal that receiver-based data communication has a great data delivery ratio, and reduces energy consumption, number of transmissions, and collisions when compared with a sender-based data communication. Moreover, we use a model checking technique to perform the formal verification of Rb-MAC.

In this work, we propose an on-line model to design heterogeneous sensor network topologies with small world features. The proposed model takes into account the data communication flow in this kind of network to create network shortcuts toward the sink node in a way that the communication between the sink and the sensor nodes is optimized. The end-points of these shortcuts are nodes with more powerful hardware, leading to a heterogeneous sensor network. We evaluate the on-line model and show that it presents the same small world features observed in the theoretical model. When the shortcuts are created toward the sink node, with a few number of powerful sensors, the network presents better small world features and interesting tradeoffs between energy and latency in the data communication when compared with the original model proposed in the literature.

In wireless sensor networks, energy consumption is generally associated with the amount of sent data once communication is the activity of the network that consumes more energy. This work proposes an algorithm based on “Principal Component Analysis” to perform multivariate data reduction. It is considered air quality monitoring scenario as case study. The results show that, using the proposed technique, we can reduce the data sent preserving its representativeness. Moreover, we show that the energy consumption and delay are reduced proportionally to the amount of reduced data.

Data communication may have a strong impact on the design of a wireless sensor network (WSN) since the energy cost related to the data transmission is typically much higher than the energy cost to perform data processing. Typically, data communication in a WSN tends to be different from other "traditional" data networks such as the Internet. In a WSN, there is a special node called sink that is either the origin or the destination of a message whereas, in the other networks, data communication happens between arbitrary communicating entities. In this scenario, the latency and the energy consumption during communication between the sink node and the other nodes can be optimized using Heterogeneous Sensor Networks (HSNs) created by applying the small world concepts. In order to generate a network with small world features, one should add a small number of long-range links, called shortcuts, whose endpoints are equipped with more powerful hardware, mainly the energy resources. In this work, we propose and evaluate two small world models that can be used in the design of HSNs. In particular, the directed angulation toward the sink node model exhibits the most interesting tradeoff between energy consumption and latency and the sink node as source/destination model presents the best results related to energy consumption and latency in the nodes near to the sink.

An important issue in the design of a wireless sensor network (WSN) is to devise techniques to make efficient use of its energy, and thus, extend its lifetime. When two WSNs are deployed at the same place and their sensors cooperate with the other networks forwarding their packets, the distance of the transmissions decreases and, therefore, the power consumption as well. The goal of this work is to examine the extent to which different WSNs can cooperate and save their energy. Simulation results reveal that different densities and data collecting rates among WSNs, the routing algorithm and the path loss exponent have major impact in the establishment of cooperation.

In a wireless sensor network, data communication may have a strong impact on its design since the energy cost related to the transmission is typically much higher than the energy cost to perform data processing. Typically, data communication in a WSN tends to be different from other ldquotraditionalrdquo data networks such as the Internet. In a WSN, there is a special node called sink node that is either the origin or the destination of a message whereas in the other networks data communication happens between arbitrary communicating entities. In this scenario, the theory of complex networks is being employed in the design of wireless sensor networks, which have certain non-trivial topological features. One of the most well-known examples of complex networks is small-world network. In this work, we will use the small-world concept as a modeling technique to build efficient data dissemination in a wireless sensor network. We propose and evaluate two small-world models that can be used in the design of a WSN. In particular the Sink Node as Source/Destination model (SSD) exhibits the most interesting tradeoff between energy and latency allowing the design of strict applications that demand a small latency and energy consumption.

A gossip protocol randomly decides the set of nodes that will forward a packet it received. Gossiping was proposed to be used in dynamic topology networks such as Wireless Sensor Networks (WSNs). This work proposes Gossiping using the Energy Map (GEM), a new gossiping-based protocol to perform energy-aware broadcasting in WSNs. The key idea is to change the random selection of neighbors in a way that the selection process uses the energy map. In our protocol, the routing flow is directed to the nodes with the greatest energy reserves, balancing the network energy and preserving nodes localized inside low energy regions to perform sensing tasks.

Low-power devices require energy-aware design and dynamic power management techniques to reduce the power consumption. Communication intensive devices make the decision of shutting down idle components very difficult, specially in wireless communication. This paper presents a new system-level dynamic power management technique to reduce the power consumption of communication intensive sensor nodes in wireless sensor networks. It analyzes the application-level information to determine the need for communication and to selectively shutdown hardware components, specially the radio. The hybrid automata framework is used to dynamically change duty cycles, making the best trade-off between energy conservation and quality of service, according to the application requirements. Simulation results reveal that application-level information is crucial to reduce power consumption and to increase the sensor node lifetime.

In this paper, a new data dissemination algorithm for wireless sensor networks is presented. The key idea of the proposed solution is to combine concepts presented in trajectory-based forwarding with the information provided by the energy map of the network to determine routes in a dynamic fashion, according to the energy level of the sensor nodes. This is an important feature of an autonomic system, which must have the capacity of adapting its behavior according to its available resources. Simulation results revealed that the energy spent with the data dissemination activity can be concentrated on nodes with high-energy reserves, whereas low-energy nodes can use their energy only to perform sensing activity or to receive information addressed to them. In this manner, partitions of the network due to nodes that ran out of energy can be significantly delayed and the network lifetime extended.

A fundamental issue in the design of a wireless sensor network is to devise mechanisms to make efficient use of its energy, and thus, extend its lifetime. Due to the importance of energy conservation, it is highly desirable to define the amount of energy each protocol can spend to perform its goal. Using this idea, we can associate a finite energy budget with each network activity, and ask this activity to achieve its best performance using only its budget. This should be considered a new paradigm to design algorithms for networks that are battery powered, specially for wireless sensor networks. In this paper, we present this new paradigm, and show how it can be used to construct the energy map of a wireless sensor network. Our goal is to construct the best energy map using only a defined amount of energy. Results show that we can approach these performance limits using the proposed finite energy budget model.