Energy-efficient Data Acquisition Method For Wireless Sensor Network

Wireless sensor networks are composed of a large quantity of sensor nodesdistributed in a certain area, and data sensed by nodes have highly spatial and temporalcorrelation. Based on the spatial-temporal correlation theory and taking sensingaccuracy and energy optimization as a prerequisite, this dissertation performs researchon energy-efficient data sampling algorithm. The algorithms proposed can reduce thevolume of data collected by nodes, and lower energy consumption for processing andtransmitting. Also, the algorithm is capable of balancing energy consumption of thewhole networks as well as avoiding the ‘energy hole’, and thus prolongs the networks’life effectively. The main work is as follows:Firstly, a novel sensor nodes clustering algorithm, the SCA (the sensing clusteringalgorithm) is proposed. Being different to all the conventional methods applied tocommunication, the algorithm clusters nodes depending on the nodes’ sensing ability,and thus forms a comprehensive covered and fully connected network. Moreover, theclustering algorithm balances the energy load by scheduling the cluster heads’duty-cycle. The clustering algorithm is the basis of the subsequent two samplingalgorithms.Secondly, an adaptive sampling algorithm for wireless sensor networks is designed.Based on the temporal correlation of the sensed data, the algorithm analyses theminimum mean square error of the data and adjusts the sampling frequency adaptively.Compared with the fixed-frequency sampling algorithms, the algorithm adjusts thesampling frequency according to the state of the target signal effectively, and thusreduces the redundancy and achieves the goal of energy conservation.Last. An asynchronous sampling algorithm for wireless sensor networks isproposed. The algorithm aimed at the situation that multi-nodes sample a single pointsource, uses the traditional over-sampling theory, and achieves the high-precisionsampling based on the distribute method. The algorithm can reduce and balance theenergy consumption of sampling among nodes, and thus prolong the networks’ lifetime.