BACKGROUND: In recent years, researchers have made significant strides in understanding the heterogeneity of breast cancer and its various subtypes. However, the wealth of genomic and proteomic data available today necessitates efficient frameworks, instruments, and computational tools for meaningful analysis. Despite its success as a prognostic tool, the PAM50 gene signature's reliance on many genes presents challenges in terms of cost and complexity. Consequently, there is a need for more efficient methods to classify breast cancer subtypes using a reduced gene set accurately. RESULTS: This study explores the potential of achieving precise breast cancer subtype categorization using a reduced gene set derived from the PAM50 gene signature. By employing a "Few-Shot Genes Selection" method, we randomly select smaller subsets from PAM50 and evaluate their performance using metrics and a linear model, specifically the Support Vector Machine (SVM) classifier. In addition, we aim to assess whether a more compact gene set can maintain performance while simplifying the classification process. Our findings demonstrate that certain reduced gene subsets can perform comparable or superior to the full PAM50 gene signature. CONCLUSIONS: The identified gene subsets, with 36 genes, have the potential to contribute to the development of more cost-effective and streamlined diagnostic tools in breast cancer research and clinical settings.

Breast cancer is the second most common cancer type and is the leading cause of cancer-related deaths worldwide. Since it is a heterogeneous disease, subtyping breast cancer plays an important role in performing a specific treatment. Gene expression data is a viable alternative to be employed on cancer subtype classification, as they represent the state of a cell at the molecular level, but generally has a relatively small number of samples compared to a large number of genes. Gene selection is a promising approach that addresses this uneven high-dimensional matrix of genes versus samples and plays an important role in the development of efficient cancer subtype classification. In this work, an innovative outlier-based gene selection (OGS) method is proposed to select relevant genes for efficiently and effectively classify breast cancer subtypes. Experiments show that our strategy presents an$F_{1}$score of 1.0 for basal and 0.86 for her 2, the two subtypes with the worst prognoses, respectively. Compared to other methods, our proposed method outperforms in the$F_{1}$score using 80% less genes. In general, our method selects only a few highly relevant genes, speeding up the classification, and significantly improving the classifier’s performance.

The Amazon Rainforest degradation is a worldwide concern. The rainforest has been endangered by the illegal wood extraction without control even in the preservation areas. Due to the large geography extension prevent these crimes with an unmanned aerial vehicle (UAV) is not always possible. The Wireless Acoustics Sensor Network (WASNs) technology can alleviate this problem. Here, we present an acoustical framework to detect the sounds produced by several chainsaws. Our framework was developed to be embedded in a sensor node, combining the Mel-Fourier Cepstral Coefficients (MFCCs) with One-class classification technique. This classification method, that is based on a kernel density approach, allows us to recognize only chainsaw sounds rejecting all the other possible environmental sounds, such as: animal's calls, weather noises or boat engines. In the experiments, we varied the number MFCCs coefficients and the Kernel bandwidth performing a leave-one-out cross validation to find the best combination. Finally, we found that the best parameter combination achieve 98% of accuracy showing a low FNR and a high TPR, fact that enhances the credibility of the system avoiding false alarms and making it an optimal choice for an WSN application.

In this work, we study the impact of data aggregation in WSNs with mobile sink nodes that can move at higher speeds. In these cases, propagated queries cannot be answered by using the sink's position when the query was sent, since the sink node will probably be elsewhere. Therefore, conventional data aggregation schemes may not be appropriate due to this high speed. Thus, we propose and evaluate the performance of three new algorithms for data routing and aggregation in WSNs when the sink is moving at a high speed. These algorithms explore the sink movement to create a routing graph composed of the union paths that intersect the sink's trajectory. At each hop, the sink speed and current network delays are used as metric to guide in-network data aggregation. Our results clearly show significant energy savings and efficient data delivery achieved by the proposed algorithms in different scenarios.

Localization systems play a major role in many applications for Vehicular Ad Hoc Networks (VANets). Although Data Fusion techniques can provide reliable localization information for most of the application requirements in VANets, enhancements on the localization systems are required and desirable. Unique characteristics of VANets such as mobility constraints, driver behavior, and high speed displacement nature of vehicles cause rapid and constant changes in the network topology, leading to the dissemination of outdated localization information. To circumvent this problem, an alternative is the use of predicted future locations of vehicles. The main idea of this approach is to use the localization prediction as an extension of a Data Fusion localization system. In such approach, a future position of a car is predicted for a given future time step and used to take advantage of a future time-space window of a vectorial trajectory rather than a static localization point. Thus, in this paper we further discuss this subject by analyzing the use of localization prediction as natural way to improve applications for VANets. We present a set of experiments that shows the results of such techniques when applied to a realistic VANet scenario.

In this work, we evaluate the performance of a distributed classification system in a Wireless Sensor Network for monitoring anurans. Our aim is to study how to take advantage of the collaborative nature of the sensor network to improve the recognition of anuran calls. To accomplish this, we evaluate four low-cost techniques (majority vote, weighted majority vote, arithmetic and geometric combinators) to combine three classifiers commonly used in sensor applications (Quadratic Discriminant Analysis, Naive Bayes, and Decision Trees) and trained to identify anuran calls. We investigate how the environment perceptions of the sensors can be used to discard confusing scenarios, i.e., scenarios in which there are multiple calls from different species at same time. Our best combination strategy achieved a gain of about 11% over a sensor taken in isolation. We also found that, by using the entropy of the species estimates, the sensor committee is able to effectively identify confusing scenarios, increasing gains over the isolated sensor to about 20%.

Data dissemination protocols suffer a critical drop in performance due to the topology changes and dynamic behavior of vehicular scenarios, which leads to increases in overhead rate, delivery delay and packet losses. Aiming at develop a protocol easily adaptable to dynamic vehicular scenarios, we propose TODD - an adaptive Traffic-Oriented Data Dissemination protocol. It leverages real-time traffic information to dynamically choose the best relay vehicles. This process is done by analyzing a metric calculated for each candidate vehicle, which may give preference to specific vehicle characteristics such as speed and distance to destination, based on the current traffic information. Besides that, a centralized version of TODD (CTODD) is proposed in order to deal with the lack of real-time traffic information stored in the vehicles. Simulation results show that both protocols can deliver up to 55% more packets with up to 85% less overhead than other important protocols from literature.

An addition to the target tracking application is the presence of a tracker that aims to intercept the target. The WSN provides the position of the target to the tracker, so that it can move towards the target. The objective is to reach the target in the shortest possible time, using the fewest number of nodes to save energy. In this paper, we propose and evaluate a Tracking Algorithm for Target Interception in face-structured sensor networks (TATI). TATI reduces the number of active nodes, activating only the faces that are feasible to be achieved by the target until the next sampling. Furthermore, the algorithm uses the nominal communication range of the nodes to shorten the route of the tracker. Simulation results show that TATI can save up to 15% in energy consumption, and it can keep the tracker around 10 meters closer to the target compared to the baseline.

In this work, we propose a new Greedy Forward algorithm to perform routing and data aggregation in WSNs. Differently from current Greedy Forward algorithms, our approach takes advantage of a sink node capable of long-range communication and uses the RSSI (Received Signal Strength Indicator) of exchanged packets to aggregate and forward data. Based on this, we propose the GRDA (Greedy Routing and Data Aggregation) algorithm with two different variations: GRDA Selection and GRDA Election. In the GRDA Selection, neighbors exchange packets with RSSI information and the next hop of the packet is then selected from a routing table at each step. In the GRDA Election, the next hop is dynamically elected at each step and no extra packets are required for the routing task. Our approach also takes advantage of RSSI values to create a time metric to guide in-network data aggregation. Our results clearly show significant energy savings and efficient data delivery achieved by the proposed algorithms in different scenarios with all the benefits of a Greedy Forward algorithm.

Wireless Sensor Networks (WSNs) are a special kind of ad hoc networks designed to comprise a high density of sensor nodes. These networks have high traffic of data and waste energy with an unnecessary number of active sensor nodes. In this paper we address the Density Control, Coverage and Connectivity Problem (DCCCP) in WSNs, that consists in activating a subset of sensor nodes, which assure the area coverage and the nodes connectivity, and minimize the energy consumption. We propose Multiperiod and Periodic approaches to solve the DCCCP. The Multiperiod approach divides the expected network lifetime in time periods and calculates, in a global way, a solution for each period, that minimizes the network energy consumption considering all time periods at once. The Multiperiod Approach has a global view of the nodes and the network expected lifetime. The Periodic Approach solves the problem as an static problem, updates the list of available nodes and repeats the procedure. The Periodic Approach has a global view of the nodes but not of the network lifetime, it finds local solutions (considering the periods) that together form a global solution, represented by the sum of all local solutions. These approaches are modeled through Integer Linear Programming (ILP). We compare our optimal solutions with Geographical Adaptive Fidelity (GAF) and Hierarchical Geographical Adaptive Fidelity (HGAF). Given the global aspects of our approaches we expect to derive lower bound for energy consumption for density control algorithms in WSNs

Map-based services have increasingly become more important in our daily life, especially mobile maps. However, the screen constraints on mobile devices significantly increases the mode-switching between panning and zooming when trying to retrieve map data which negatively affects the user experience. The Anchored Zoom (AZ) is one of the techniques that have been proposed to overcome this problem, but it has some limitations concerning the use of its anchor point, i.e. a point of reference that serves as a major tool to control zooming in this technique. This paper presents the Anchored Zoom with Anchor Management (AZAM), a new technique that extends AZ by adding an anchor management to it, i.e., by adding new features that better exploit the use of anchors, improving the choice of new anchors and access to the previous ones. We evaluated both techniques, AZ and AZAM, and we also investigated their relevance in comparison with the popular Pinch to Zoom (PZ) technique. The study reveals that AZAM is superior to AZ in perceived satisfaction and that it serves as a complement to PZ during visual search tasks.

Wildlife sounds provide relevant information for non-intrusive environmental monitoring when Wireless Sensor Networks (WSNs) are used. Thus, collecting such audio data, while maximizing the network lifetime, is a key challenge for WSNs. In this work, we propose a methodology that applies Compressive Sensing (CS) aiming at collecting as little data as possible to allow the signal reconstruction, so that the reconstructed signal is still representative. The key issue is to determine a sparse base that best represents the audio information used for identifying the target species. As a proof-of-concept, we focus on anuran (frogs and toads) calls, but the methodology can be applied for other animal families and species. The reason for that choice is that long-term anuran monitoring has been used by biologists as an early indicator for ecological stress. By using real wild anuran calls, we show that 98% classification rate can be achieved by using as little as 10% of the original data. We also use simulation to evaluate the impact of our solution on the network performance (energy consumption, delivery rate, and network delay).

Anurans (frogs or toads) are commonly used by biologists as early indicators of ecological stress. The reason is that anurans are closely related to the ecosystem. Although several sources of data may be used for monitoring these animals, anuran calls lead to a non-intrusive data acquisition strategy. Moreover, wireless sensor networks (WSNs) may be used for such a task, resulting in more accurate and autonomous system. However, it is essential save resources to extend the network lifetime. In this paper, we evaluate the impact of reducing data dimension for automatic classification of bioacoustic signals when a WSN is involved. Such a reduction is achieved through a wrapper-based feature subset selection strategy that uses genetic algorithm (GA). We use GA to find the subset of features that maximizes the cost-benefit ratio. In addition, we evaluate the impact of reducing the original feature space, when sampling frequencies are also reduced. Experimental results indicate that we can reduce the number of features, while increasing classification rates (even when smaller sampling frequencies of transmission are used).

Wireless Sensor Networks consist of a powerful technology for monitoring the physical world. Particularly, in-network data fusion techniques are very important to applications such as target classification and tracking to reduce the communication burden in these constrained networks. However, the efficiency of the solution can be affected by the data correlation among several sensor nodes. Thus, the application of value fusion (for clusters of nodes with correlated measurements) and decision fusion (combining the local decisions of the clusters) is a common strategy. In this work, we propose an algorithm for properly selecting the groups of nodes with correlated measurements. Experiments show that our algorithm is 30% better than a solution that considers only the spatial coherence regions.

Many applications in Wireless Sensor Networks (WSNs) consider remote and large scale monitoring. For those scenarios, the whole network is difficultly fully connected all the time. A possible vision is the union of WSNs and Disruptive Tolerant Network(DTNs) concepts, where mobile nodes make collect data in sparse networks and deliver them to a base station. This work presents a buffer management strategy, which is a basic principle in DTN networks. The proposed solution considers the global coverage to choose which messages are dropped, therefore, minimizing the impact on monitoring application. Such solutions are important for WSNs applications with limited resources. We show through simulation that the proposed Coverage-Based Drop-Policy in Wireless Sensor Network with Disruptive Connections (CBDP) improves the network coverage compared to traditional DTN drop policies like Drop Last Packet (DL) and Drop First Packet (DF).

Vehicular Networks routing protocols must deal with several issues such as high mobility, high speed and, consequently, high disconnection rate among the network nodes. In this paper, we propose VCARP, a geocast routing protocol for Vehicular Networks that takes into account context informations of the network (such as nodes location and destination) to make routing decisions. It consists of a shared cache mechanism used to avoid packets loss due to full caches, and a flow-based routing to reduce network overhead caused by unnecessary packet retransmissions. Simulations show that the proposed protocol can increase packet delivery rate by 40% and decrease overhead by up to 89%, compared to another geocast protocol of the literature. In addition, we have studied the influence of cache size and neighbors discovery period duration on the packet delivery rate.

Event detection is a topic widely discussed in the wireless sensor network (WSN) community. The goal of this process is to identify when the collected data represents an event occurrence. The lack of uniformity when approaching this problem, such as the modeling of the data collected by sensors, is an obstacle to the comparison among the different proposals. In this work, we present an extension of the Diffuse framework for scenarios regarding event detection, while considering settings with inaccurate and fault-susceptible measurements made by non-ideal sensing devices in a noisy environment. Based on that framework, we compare a new proposed algorithm with a method based on the literature, obtaining hit rates of event detection above 72% in settings with 30% of sensor faults.

Wireless Sensor Networks (WSNs) are commonly employed for environmental and wildlife monitoring. In these scenarios, mobile robots with specialized sensing, processing and actuation abilities may be employed to investigate relevant events in place. However, driving the robot to the event place is not a trivial task in the typical case where sensor nodes do not have positioning sensors such as GPS. In this work we propose a novel navigation algorithm for the mobile robot based solely on the Received Signal Strength Indication (RSSI) of communication packets. The proposed algorithm builds on a probabilistic signal propagation model recovered from real signal decay data, unlike alternative solutions found in literature. Simulations show a superior performance in comparison to similar related work.

6LoWPAN opens a great opportunity for wireless sensor networks (WSNs) to be operated from the Internet. However, 6LoWPAN requires a gateway in order to make the use of IPv6 feasible inside the WSN. In this paper, a gateway is designed and constructed to interconnect 6LoWPAN wireless sensor networks with IPv6 clients, enabling them to receive sensor nodes readings or send commands to the 6LoWPAN WSN at anytime through direct communication with the sensor nodes. Furthermore, this solution allows IPv4 clients to exchange information with the WSN through a bridge architecture implemented at the gateway. The experimental results show that the communication between 6LoWPAN WSN and IPv6/IPv4 clients is successful.

In this paper, it is presented the design of a Universal Plug and Play (UPnP) gateway for wireless sensor networks (WSNs), capable of discovering all the services of a WSN, and creating a unique UPnP device to advertise them to interested UPnP control points. Besides, an application protocol for WSN is designed to exchange UPnP-related data between the WSN and the gateway. Experiments in an indoor environment with restricted sensor nodes confirm the feasibility of this solution.

Target tracking is an important application of sensor networks, particularly interesting in applications for Ecology for monitoring and tracking animals. In this context, understanding the movement pattern and the territorial occupation of animals are fundamental for understanding their habits. In practice, target tracking often operates on quantized areas (divided into cells). In this work, we propose and evaluate a quantized target tracking approach in such a way that the network is organized in a grid, where each cell is a region occupied by the target (animal). The cell size is determined according to the desired granularity. The computation of the target's position obeys a voting scheme, so the technique is simple and low cost. To estimate the target's position, we use the Kalman or Particle filters. Results show that position computation errors are close to two cells, depending on the scenario.

Sink nodes are responsible for aggregating data gathered by a Wireless Sensor Network (WSN) and transmitting them to a monitoring facility. Usually, these are static nodes that serve as gateways to infrastructured networks. In some applications, these sink nodes move around the monitoring area, usually in lower speeds (e.g., robots, dirigibles). In this work, we go further and study WSN scenarios in which sink nodes can move at higher speeds, such as an Unmanned Aerial Vehicle (UAV) or even an airplane. In these cases, propagated queries cannot be answered by using the sink's position when the query was sent, since the sink will be elsewhere due to its high speed. Thus, we propose and evaluate the performance of three algorithms for data query in WSNs when the sink is moving at a high speed. Our results show clearly the need for new algorithms for these scenarios as well as the good performance of our proposed algorithms.

Location-Based Services have become quite popular in mobile computing. However, these services are often not effective when devices are in indoor environments. This paper presents an indoor room-level Bluetooth-based localization system. Fixed stations distributed throughout the environment get the RSSI from user devices. With these data, we compute the distance between the device and the stations, based on the signal propagation. In this paper, we propose a room localization method, based on a voting scheme, that eliminates cells close to a base station detecting devices with a weak signal. Another contribution of this paper is the empirical evaluation of the methods within a real environment.

The majority of event-based solutions for wireless sensor networks (WSNs) are evaluated considering a given event-detection model, such as a fixed radius. Such models are used to validate algorithms for different problems in WSNs, such as coverage, routing, density control, and target tracking. However, current models are not suitable for every scenario. Usually, the detection efficacy depends on several factors, such as environmental conditions, sensor reliability, measurement methods, and event characteristics. In this work, we propose a probabilistic detection-model that aims at modeling different scenarios, and can be configured to properly approximate other detection models adopted in WSNs. We simulate the detection capability of our model in comparison to the popular fixed radius.

Wireless Sensor Networks (WSN) are often designed to detect events based on measurements provided by sensor nodes. In this work, we propose and evaluate two algorithms, based on information fusion methods, for fire detection in rainforests when we have only light and temperature sensors. The idea is to have acceptable detection methods even when we do not have (or cannot afford) other sensor units that would lead to more accurate inferences, such as humidity sensors, smoke detectors or video cameras. In this sense, we collect and analyze real environmental data in situations that are similar to those found in real scenarios. Then, we create some controlled fire (events) of small proportion, close to sensor nodes, to force event occurrences. Experimental results indicate that forest fire can be detected at night by using just light measurements. In addition, light data can be used to distinguish heat resulting from sunbeam and fire. The first method is based on the moving average filter and thresholds (Threshold method), while the second uses the moving average filter and the Dempster-Shafer theory. Both methods present similar results regarding the detection efficacy considering the generated instances. However, the Threshold method is more stable, due to the use of state machine to represent the environment conditions.

In wireless sensor networks (WSNs), target tracking algorithms are often based on geographical information provided by a localization algorithm. However, the errors introduced by such algorithms may affect the performance of tasks that depend on such information. In this paper, we evaluate how errors resulting from actual localization algorithms affect two classical target tracking algorithms: the Kalman filter and the Particle filter. As a proof-of-concept, we choose two relevant localization algorithms for WSNs. The first is the RPE (Recursive Position Estimation), a pioneer iterative solution, and the second is the DPE (Directed Position Estimation), another iterative solution that evolved from the RPE reducing errors and the cost of its predecessor. Our results clearly indicate that the tracking algorithms successfully filter the noise introduced by the target-node distance estimation. However, they fail to eliminate the errors introduced by localization algorithms.

The establishment of a localization system is an important task in wireless sensor networks. Due to the geographical correlation between sensed data, location information is commonly used to name the gathered data and address nodes and regions in data dissemination protocols. In general, to estimate its location, a node needs the position information of at least three reference points (neighbors that know their positions). In this work, we propose a different scheme in which only two reference points are required in order to estimate a position. To choose between the two possible solutions of an estimate, we use the known direction of the recursion. This approach leads to a recursive localization system that works with low-density networks (increasing by 40 percent the number of nodes with estimates in some cases), reduces the position error by almost 30 percent, requires 37 percent less processor resources to estimate a position, uses fewer beacon nodes, and also indicates the node position error based on its distance to the recursion origin. No GPS-enabled node is required, since the recursion origin can be used as a relative coordinate system. The algorithm's evaluation is performed by comparing it with a similar localization system; also, experiments are made to evaluate the impact of both systems in geographic algorithms.

In this work, we propose a new Greedy Forward algorithm to be used in Wireless Sensor Networks (WSNs) routing protocols. Differently from current Greedy Forward algorithms, which depend on the position information of nodes, our approach uses only the Received Signal Strength Indicator (RSSI) of exchanged packets. Since nodes' position are usually computed based on three or more distance estimations (commonly by means of RSSI measurements) and the positions of beacon nodes (which have inaccuracies by themselves), by using only RSSI measurements, we should achieve better results. Based on this observation, we propose the RSSR (Received Signal Strength Routing) algorithm with two variants: RSSR Election and RSSR Selection. In the RSSR Election, the next hop is dynamically elected at each step and no packets are required for the routing task. In the RSSR Selection, neighbors exchange packets with RSSI information and the next hop of the packet is then selected from a routing table at each step. Our results indicate clearly that the proposed algorithms have all the benefits of a Greedy Forward algorithm but with better performance and without requiring positions information.

In wireless sensor networks (WSNs), detected events need to be localized in both time and space. As a result, sensor nodes need to have precisely synchronized clocks as well as to be localized in a common spatial reference system. While synchronization and localization algorithms have been proposed to solve these problems independently, in this paper we propose to combine both synchronization and localization into a single problem that we refer to as the time-space localization problem. We then propose the enlightness algorithm (enhanced and lightweight time and space localization), a new and enhanced time-space localization algorithm for wireless sensor networks. Our proposed algorithm, based on a previously proposed time-space localization algorithm called Lightness, uses mathematical ranking and weighting techniques to enhance the time computation process. We present an extensive set of experiments to evaluate the performance of our new algorithm. Our results indicate clearly that the proposed scheme can scale to thousands of nodes (over 2048) while reducing the synchronization error by half when compared to the original lightness algorithm. In comparison to the FTSP algorithm (flooding time synchronization protocol), a well known synchronization algorithm, our proposed enlightness algorithm is able to synchronize the clocks of the sensor nodes with a four times greater precision in small networks and about a 10 times greater precision in large networks.

A typical problem in wireless sensor networks is how to collect and send historical information from all sensor nodes to the base station. As the available energy of these nodes is a critical resource, it is typically not feasible to transmit a full-resolution data feed from each sensor node to the base station. Therefore, it is important to apply techniques of data reduction so that fewer bits can be transmitted. In this work, we propose an algorithm for differentiated codification such that the sensor nodes send only the differences of their readings to a common data base. Simulation results reveal that the algorithm performs well in applications where the sensor nodes collect similar readings.

Wireless sensor networks (WSNs) are commonly treated as a distributed database system that is accessed by means of a query language. However, the computation of such queries are usually performed by information fusion techniques. Information fusion has been used by applications to detect/classify events, track targets, and filter noisy measurements. In this tutorial, we discuss some of the information fusion techniques that are currently used in WSNs.

Localization in time and space can be defined as the problem of solving both synchronization and positioning problems at the same time. This is a key problem for wireless sensor networks that need to determine the timing and location information of detected phenomena, especially for tracking applications. In this paper, we discuss the relationship between these two problems and propose the Mobilis (Mobile Beacon for Localization and Synchronization) algorithm, a new time-space localization algorithm for wireless sensor networks. The main aspect of the Mobilis algorithm is the use of a mobile beacon for both localization and synchronization. A mobile beacon is a node that is aware of its time and position (e.g. equipped with a GPS receiver) and that has the ability to move around the sensor field. In our algorithm, the synchronization component uses the packets required by the positioning component to improve its performance. Similarly, the positioning component benefits from the communication required by the synchronization component to decrease errors. We also present an extensive set of experiments and simulations to evaluate the performance of our algorithm.

A wireless sensor network (WSN) is energy constrained, and the extension of its lifetime is one of the most important issues in its design. Usually, a WSN collects a large amount of data from the environment. In contrast to the conventional remote sensing - based on satellites that collect large images, sound files, or specific scientific data - sensor networks tend to generate a large amount of sequential small and tuple- oriented data from several nodes, which constitutes data streams. In this work, we propose and evaluate two algorithms based on data stream, which use sampling and sketch techniques, to reduce data traffic in a WSN and, consequently, decrease the delay and energy consumption. Specifically, the sampling solution, provides a sample of only log n items to represent the original data of n elements. Despite of the reduction, the sampling solution keeps a good data quality. Simulation results reveal the efficiency of the proposed methods by extending the network lifetime and reducing the delay without loosing data representativeness. Such a technique can be very useful to design energy-efficient and time-constrained sensor networks if the application is not so dependent on the data precision or the network operates in an exception situation (e.g., there are few resources remaining or there is an urgent situation).

Monitoring applications define an important class of wireless sensor networks (WSNs). In these applications, the network perceives the environment and searches for event occurrences (phenomena) by sensing different physical properties, such as temperature, humidity, pressure, ambient light, and movement. In such cases, location in time and space of both phenomenon and nodes are usually required for tracking and correlation purposes. In the former case, nodes' clocks are synchronized, while in the latter, nodes' physical location (e.g., latitude, longitude) are discovered. The need for such information in WSNs and the proximity and similarity of both problems suggest they can be addressed as a single problem, which we call the localization in time-space. In this paper, we discuss the relationship between these problems and also propose the Synapse (synchronization and positioning for sensor networks), a time-space localization system for sensor networks. Simulation results show that the proposed algorithm is able to find the nodes' location while synchronizing their clocks with a precision of a few microseconds. A key advantage of Synapse is that it reduces the communication cost compared to the situation wherein localization and synchronization are separately solved.

This work presents a sampling data stream algorithm for wireless sensor networks (WSNs). The proposed algorithm is based on sampling techniques applied to data histograms created from original data streams acquired by sensor nodes. As a result, the algorithm provides a sample of only log n items to represent the original data of n elements. We show that by using our algorithm, we can save energy and reduce delay in WSN applications in different scenarios while keeping a good data quality.

A fundamental goal of a wireless sensor network (WSN) is to collect and deliver data to external applications. Due to the strong constraints of these networks, routing algorithms usually consider application-specific characteristics and, consequently, there is no self-contained algorithm appropriate for every case. In particular, many WSN applications are event- driven. In such scenarios, the behavior of the network may vary a lot, which favors different algorithms at different instants. In these cases, hybrid adaptive solutions are more suitable by allowing the adoption of a better routing strategy in response to the variation of the network conditions. In this work, we evolve this approach by considering the temporal characteristic of event occurrence and detection through an estimation model. The proposed solution, called multi, predicts the need of routing infrastructure creation and maintenance for the adaptation between a reactive and a proactive strategy. To show the advantages of the proposed approach, an instance of multi is evaluated through simulations. Comparisons with its independent reactive and proactive components and other proposed solutions show improvements on energy consumption.

Localization systems have been identified as a key issue to the development and operation of the Wireless Sensor Networks (WSN). A DV-Hop localization system works by transforming the distance to all beacon nodes from hops to units of length measurement (e.g., meters, feet) using the average size of a hop as a correction factor. Despite its advantages, a DV-Hop algorithm has some disadvantages, such as its large communication cost that limits its scalability, and its mapping from hops to distance units that introduces errors that are propagated to the computation of a node location. This last issue has been solved by some recent works, but the scalability problem still is an open problem that limits this technique to small or medium sized networks. In this work, we propose a novel approach that uses Voronoi diagrams in order to scale a DV-Hop localization algorithm while mantaining or even reducing its localization error. Two types of localization can result from the proposed algorithm: the physical location of the node (e.g., latitude, longitude), or a region limited by the node's Voronoi cell. The algorithm evaluation is performed by comparison with similar algorithms. We show how the proposed algorithm can scale in different aspects such as communication and processing costs when increasing the number of nodes and beacons.

This work presents the design of real-time applications for wireless sensor networks (WSNs) by using an algorithm based on data stream to process the sensor data. The proposed algorithm is based on sampling techniques applied to data histograms created from original data streams acquired by sensor nodes. As a result, the algorithm provides a sample of log n items to represent the original data stream of n elements. In this work, we show how to use the data reduction algorithm in real-time WSN design.

In this work, we propose an integrated solution to jointly solve the localization and data routing problems in sensor networks. In our solution, the localization system divides the network into Voronoi cells to reduce communication costs, and as a result, the system allows two ways for specifying a node location: the nodes ' physical location or the Voronoi cell it belongs to. This localization system yields a natural hierarchy based on Voronoi cells that is used by the integrated routing algorithm we propose. The proposed geographic routing algorithm may use either the nodes ' coordinates or the Voronoi cells to forward data packets. A key advantage of this routing algorithm is that no additional communication is required for route discovery, leading to a cost-free and very scalable routing algorithm. Such a dual solution, called Cellular (Cell Unified Localization And Routing) algorithm, to solve the localization and routing problems, represents a new trend of algorithms for sensor networks, where multiple problems are jointly solved to provide cheaper and integrated solutions.

Localization in time and space can be defined as the problem of solving both synchronization and positioning problems at the same time. This is a key problem for wireless sensor networks that need to determine timing and location information of detected phenomena, especially in tracking applications. In this paper, we discuss the relationship between these two problems and propose the Lightness algorithm (Lightweight Time and Space localization), a new time-space localization algorithm for wireless sensor networks. Our proposed algorithm not only takes advantage of the additional hardware resources required by a positioning system to improve the performance and scalability of the synchronization component, but also benefits from the additional communication needed by the synchronization component to decrease positioning errors. We also present an extensive set of experiments to evaluate the performance of our algorithm. Our results indicate clearly that the proposed scheme scales to thousands of nodes (over 2048) while keeping a synchronization error of a few microseconds (nearly 10µs) with an O(n) communication cost. The additional packets needed to compute clocks' drift can also be used to decrease the positioning errors to almost one third of the initial positioning.

Wireless sensor networks are ad hoc networks comprised mainly of small sensor nodes with limited resources and one or more base stations, which are much more powerful laptopclass nodes that connect the sensor nodes to the rest of the world. The advent of this tecnology over the last decade enables largescale deployment of such sensors. On the other hand, this poses the challenging of how the great amount of information generated by these sensor networks will be handled at the base station. In this paper, we propose a network processor architecture tailored specifically to handling information at sensor network base stations. Our approach optimize information processing by implementing tasks in hardware. We show that the novel architecture is one order of magnitude faster than an architecture based on traditional RISC processor.

In resource-constrained wireless sensor networks (WSNs), communication protocols need to be energyefficient. In this context, information-fusion techniques have been used to reduce packet transmissions, and role assignment algorithms have been proposed to define which nodes should perform information fusion. However, most of the role assignment algorithms for WSNs are proactive in the sense that roles are assigned even when no event is being detected. In this work, we propose a reactive algorithm that assigns roles only when an event is detected, therefore saving energy during the network inactivity. In addition, to save energy during event notification, our algorithm searches for the shortest paths that maximize data aggregation.

The establishment of a localization system is an important task in wireless sensor networks. Due to the geographical correlation of the sensed data, location information is commonly used to name the gathered data and address nodes and regions in data dissemination protocols. In general, to estimate its location, a node needs the position information of, at least, three reference points (neighbors that know their positions). In this work, we propose a different scheme in which only two reference points are required to estimate a position. To choose between the two possible solutions of an estimate, we use the known direction of the recursion. This approach leads to a recursive localization system that works with low density networks (increasing in 40% the number of nodes with estimates in some cases), reduces the position error in almost 30%, requires 37% less processor resources to estimate a position, uses less beacon nodes, and also indicates the node position error based on its distance to the recursion origin. No GPS-enabled node is required, as the recursion origin can be used as the relative coordinate system.

In this paper we present a new tool called Bean-Watcher. This tool allows the semi-automatic generation of monitoring and management applications for wireless networks such as WLANs and WPANs. The architecture of the tool is based on a component model flexible enough to allow the creation of new components and the optimization of the components currently provided. BeanWatcher was designed to offer a development environment suitable for both expert and beginner users allowing them to choose the programming language that better fits the application requirements.