Carlos Giovanni Nunes de Carvalho

Improving Prediction Accuracy for WSN Data Reduction by Applying Multivariate Spatio-Temporal Correlation

This paper proposes a method based on multivariate spatial and temporal correlation to improve prediction accuracy in data reduction for Wireless Sensor Networks (WSN). Prediction of data not sent to the sink node is a technique used to save energy in WSNs by reducing the amount of data traffic. However, it may not be very accurate. Simulations were made involving simple linear regression and multiple linear regression functions to assess the performance of the proposed method. The results show a higher correlation between gathered inputs when compared to time, which is an independent variable widely used for prediction and forecasting. Prediction accuracy is lower when simple linear regression is used, whereas multiple linear regression is the most accurate one. In addition to that, our proposal outperforms some current solutions by about 50% in humidity prediction and 21% in light prediction. To the best of our knowledge, we believe that we are probably the first to address prediction based on multivariate correlation for WSN data reduction.

Application of wireless sensor networks for beehive monitoring and in-hive thermal patterns detection

A beehive monitoring able of detecting stress condition of the bees in high temperature.Overheating can lead to absconding.Monitoring and analysis of the microclimate on hives alert of overheating.Use of clustering for similarities to detect thermal patterns that indicate overheating.Due to the patterning of the collected thermal data, we avoid sending redundant data, reducing the energy cost. As cold-blooded animals, bees seek to control the environment thermal variation to live and work in their hives. In semi-arid regions, such as in Northeast Brazil, bees lead a natural thermoregulation mechanism inside their hives so that they can deal with high temperatures. However, when thermoregulation is not fully accomplished, all bees can leave the nest in a process known as colony absconding. In such a process, absconding is due to a thermal stress stimulus. In this context, here we propose a proactive monitoring of hives using a wireless sensor network which detects atypical heating. Through thermal patterns obtained on a daily basis, we developed a mechanism for detecting the temperature rise inside the hive (microclimate). Our results show various thermal patterns related to hive conditions, and highlight the temperature as a key factor to detect potential absconding conditions.

Multiple linear regression to improve prediction accuracy in WSN data reduction

Simple linear regression is usually used for WSN data reduction. The mechanism is concerned about energy consumption, but neglects the prediction accuracy. The prediction error from it is often ignored and inconsistencies are forwarded to the user application. This paper proposes to use a method based on multiple linear regression to improve prediction accuracy. The improvement is achieved by multivariate correlation of readings gathered by sensor nodes in field. Tests show that our solution outperforms some current solutions adopted in the literature.

A predictive algorithm for mitigate swarming bees through proactive monitoring via wireless sensor networks

Swarming is the massive outflow of the bees in a hive, whose most common causes are high temperatures, lack of food, stress and humidity changes. Among the types of swarming, one in which the complete abandonment of the hive occurs, has created large losses to Brazilian beekeepers, especially the Northeast. In an attempt to mitigate this problem, we propose in this paper a system for monitoring hive, via a wireless sensors network capable of identifying the preswarming colony behavior. Through a pattern of collections obtained from the cyclical behavior daily temperatures, we developed a predictive algorithm based on pattern recognition techniques, able to detect the increase in temperature in the hive (microclimate) responsible for the typical stress of bees that culminates in swarming. This mechanism is also able to recognize and avoid sending redundant information over the network in order to reduce radio communication, thereby reducing costs of data transmission and energy.

A Quality-Aware and Energy-Efficient Context Management Framework for Ubiquitous Systems

Sensor-rich Context Management Frameworks (CMF) for Ubiquitous Systems should be able to continuously gather raw data from observed entities in order to characterize the current situation. However, the death of independent sensors and monitoring platform reduce the ability of CMF for detecting the current situation, which directly affects the availability of context-aware applications/services. This paper proposes a quality-aware data reduction approach to minimize the amount of sensed raw data sent to CMF, reducing the energy consumption and network traffic. The proposed approach, based on Adaptative Simple Linear Regression (ASLR), rebuilds the gathered raw data that was not intentionally sent to CMF by prediction. Quality requirements defined on gathered data (Quality of Context) are respected by the reduction approach, avoiding the loss of precision (QoCI precision) and timeliness (QoCI up-to-dateness). The proposed data reduction approach has been integrated into our Context Management Framework (CxtMF), which provides context information for two context-sensitive services: beehive and ECG monitoring services. Experimental results indicate that the proposed approach reduces the amount of packets sent over network to 3% for the ECG monitoring service, and 12.15% for the beehive monitoring service, respectively.

An Energy-Efficient Context Management Framework for Ubiquitous Systems

Sensor-rich Context Management Frameworks (CMF) for Ubiquitous Systems should be able to continuosly gather raw data from observed entities (e.g., people, surround enviroment) in order to characterize the current situation (i.e., context). However, the energy of sensors can end up, which reduce the ability of CMF for detecting the current situation, directly affecting the availability of context-aware applications/services. This paper propose a data reduction approach to lower the amount of data sent to CMF over the network, minimising the energy consumption and the network traffic of sensor-rich CMF. The proposed data reduction approach rebuilds data that are not intentionally sent from sensors by prediction based on simple linear regression. The gathered raw data is modeled by linear equations and its parameters are sent to the CMF, instead of a set of readings. Thus, it reduces the communication overhead between sensors and CMF, enhancing the lifetime of sensors. Experimental results show that is possible to reduce the amount of packets sent over the network to 3% in ECG monitoring service, and 12.15% in beehive monitoring service with mean square error of 0.0009 and 0.0981, respectively.

Reducing Energy Consumption in Provisioning of Virtual Sensors by Similarity of Heterogenous Sensors

In the context of sensor clouds, the provisioningprocess is essential since it is responsible for selecting physicalsensors that will be allocated to compose virtual sensors. Inliterature, most works consider the allocation of all sensorswithin the region of interest. Such an approach, however, cancause serious problems such as wasted energy consumption. The objective of this paper is to present an approach toreducing energy consumption in provisioning of virtual sensorsby similarity of heterogenous sensors. The approach minimizesthe number of selected nodes and accordingly reduces the totalenergy consumption of sensor nodes that make up the cloud. Results from initial experiments show that the approach reducesenergy consumption by 73.97%, providing a solution to beconsidered in sensor cloud scenarios.

A pervasive energy-efficient ECG monitoring approach for detecting abnormal cardiac situations

Mobile and pervasive ECG monitoring systems require continuous connectivity with server-side ECG analyser for instantaneously detecting abnormal cardiac situations. Normally, these systems generate a large amount of data, resulting in a high energy expenditure with data transmission on pervasive ECG platform. In this context, data reduction mechanisms can be applied for saving transmission energy of pervasive ECG monitoring devices, maximizing the availability and confiability of ECG monitoring systems. This paper proposes an pervasive energy-efficient ECG monitoring approach for detecting abnormal cardiac situations for ubiquitous health systems. The data reduction approach based on error prediction maximize the life time of pervasive ECG monitoring device by gathering and reducing heart signal before sending it to server-side ECG analyzer application. Moreover, Pearsons Coefficient (correlation rate) is applied on the proposed data reduction approach, enhancing the quality of monitored heart signal.

An Energy-Efficient Approach to Enhance Virtual Sensors Provisioning in Sensor Clouds Environments

Virtual sensors provisioning is a central issue for sensors cloud middleware since it is responsible for selecting physical nodes, usually from Wireless Sensor Networks (WSN) of different owners, to handle user’s queries or applications. Recent works perform provisioning by clustering sensor nodes based on the correlation measurements and then selecting as few nodes as possible to preserve WSN energy. However, such works consider only homogeneous nodes (same set of sensors). Therefore, those works are not entirely appropriate for sensor clouds, which in most cases comprises heterogeneous sensor nodes. In this paper, we propose ACxSIMv2, an approach to enhance the provisioning task by considering heterogeneous environments. Two main algorithms form ACxSIMv2. The first one, ACASIMv1, creates multi-dimensional clusters of sensor nodes, taking into account the measurements correlations instead of the physical distance between nodes like most works on literature. Then, the second algorithm, ACOSIMv2, based on an Ant Colony Optimization system, selects an optimal set of sensors nodes from to respond user’s queries while attending all parameters and preserving the overall energy consumption. Results from initial experiments show that the approach reduces significantly the sensor cloud energy consumption compared to traditional works, providing a solution to be considered in sensor cloud scenarios.

Using QoC for improving energy-efficient context management in U-Health Systems

Context Management Framework (CMF) for Ubiquitous Health (U-Health) Systems should be able to continuously gather raw data from observed entities to characterize their current situation (context). However, the death of battery-dependent sensors reduce their ability for detecting the context, which directly affects the availability of context-aware u-health services. This paper proposes the use of Quality of Context (QoC) integrated with a data reduction approach to minimize the amount of sensed raw data sent to CMF, reducing the energy consumption and maximizing the lifetime of sensor-based CMF. The proposed approach rebuilds the gathered raw data taking into account QoC requirements, avoiding the loss of precision (QoC Indicator precision) and timeliness (QoC Indicator up-to-dateness), which has been integrated into our Context Management Framework (CxtMF). Experimental results demonstrate the effectiveness of our approach by reducing the amount of packets sent over network to 3% for the ECG monitoring service.