r18 - 13 Nov 2009 - 10:58:28 - DanieleApilettiYou are here: TWiki >  Public Web > ResearchActivities > SereneFramework

Analysis of sensor network data

Sn-DB.jpg

Motivation

Smart sensors are small-scale mobile devices which integrate sensing, processing, storage and communication capabilities. They can programmatically measure physical quantities, perform simple computations, store, receive and transmit data. The lattice built by a set of cooperating smart sensors is called a sensor network. Because of the ambivalent role of each device, which acts simultaneously as a data producer and as a data forwarder, sensor networks provide a powerful infrastructure for large scale monitoring applications (e.g., habitat monitoring). Hence, nowadays Wireless Sensor Networks (WSN) are being used for a fast-growing number of different application fields which can be classified in: Habitat-monitoring applications (e.g., environment monitoring) and surveillance applications (e.g., health care monitoring, condition maintenance in industrial plants and process compliance in food and drug manufacturing). While in the first case a given environment is continuously monitored, in the second a control system is alerted when a critical event occurs in a given environment. In both cases, the monitored area is densely deployed with sensing devices.

Querying the network entails the frequent acquisition from sensors of measurements describing the state of the monitored environments. To transmit the required information, sensors consume energy. Since sensors are battery-powered, network querying needs to be driven by three factors: (i)~Power management, (ii) limited resources, and (iii) real-time constraints. While CPU overheads are very small (i.e., no significant processing takes place on the nodes), the main contributors to energy cost are communication and data acquisition from sensors. Thus, when querying a sensor network, the challenge is to reduce the data collection cost, in terms of both energy and bandwidth consumption. An important issue in this context is the reduction of energy consumption to maximize the longevity of the network.

Serene Framework

Serene is a flexible framework which provides high quality models for sensor networks to efficiently acquire sensor data. Given sensor readings, the goal of Serene is to find and understand the relationships, both in the space and time dimensions, among sensors and sensor readings to select a subset of good quality representatives of the whole network. The Serene framework exploits clustering techniques to select representative sensors, which will be queried instead of the whole network to reduce communication and computation costs and balance energy consumption among sensors. This approach is particularly suitable for monitoring applications.

Publications

[1] Baralis E., Cerquitelli T., D’Elia V., Intelligent Acquisition Techniques for Sensor Network Data. In: Alfredo Cuzzocrea. Intelligent Techniques for Warehousing and Mining Sensor Network Data. Accepted on September 2008. In press.

[2] Baralis E., Cerquitelli T., D'Elia V., Modeling a Sensor Network by means of Clustering, In: DEXA '07, IEEE Computer Society, 18th International Conference on Database and Expert Systems Applications, Regensburg, Germany 3-7 September 2007, pp. 177-181, 2007

[3] Baralis E., Cerquitelli T., Selecting Representatives in a Sensor Network, In: SEBD 2006, SEBD 2006, Portonovo Ancona Italy June 18-21, 2006, pp. 351-360, 2006, ISBN: 88-6068-018-2

[4] Apiletti D.; Baralis E.; Cerquitelli T.; Cheema A. A., Energy-aware models for sensor network data acquisition. Proceedings of the International Conference on Distributed Computing in Sensor Systems (DCOSS 2009), Marina Del Rey, California, USA, June 8-10, 2009

Links

  • WiFi4Energy project at Politecnico di Torino.
    • Specific WP3 activities

 
Home
Copyright by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
If not differently specified, this work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 License.

Creative Commons License