Adaptive Filter Based Strategy for Data Collection in Wireless Sensor Networks

Data collection is a fundamental task in many wireless sensor networks applications. It is impracticable to send all sensed data to base station for each sensor node, due to the constraints in communication cost and the bandwidth. Filter can provide sensed data estimation with the error bound guaran...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2016 IEEE International Conferences on Big Data and Cloud Computing (BDCloud), Social Computing and Networking (SocialCom), Sustainable Computing and Communications (SustainCom) (BDCloud-SocialCom-SustainCom) S. 317 - 324
Hauptverfasser: Ran Bi, Liang Sun, Xu Zheng, Guozhen Tan
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.10.2016
Schlagworte:
Adaptation models
approximate algorithms
Base stations
Data collection
Data models
Estimation
Filter
Optimization
Wireless sensor networks
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Data collection is a fundamental task in many wireless sensor networks applications. It is impracticable to send all sensed data to base station for each sensor node, due to the constraints in communication cost and the bandwidth. Filter can provide sensed data estimation with the error bound guarantee. For given filter [l i , u i ], node i sends data if and only if the sensed data is beyond the range of [l i , u i ]. The main idea of the filter based approach is to maintain the filters of each node at both sensor node and base station. In this paper, we investigate the adaptive filter based strategy for data collection. The variation of sensed data is modeled as a one-dimensional random walk and the formulas for model parameter estimation are provided. The problem of filter assignment with error bound guarantee is formalized as an optimization problem. A greedy heuristic based algorithm for filter assignment subject to the error bound constraints is proposed, whose time and space complexities are O(nτ/α min ) and O(n) respectively. And a light-weight filter update strategy is provided, when a filter is failure. Experimental results show that our algorithms have better performance in terms of communication cost and expected time of valid filters.
DOI:10.1109/BDCloud-SocialCom-SustainCom.2016.55