Data sampling control, compression and query in sensor networks
Jinbao Li, Jianzhong Li · International Journal of Sensor Networks · 2007
Nodes in wireless sensor networks have very limited storage capacity, computing ability and battery power. Node failure and communication link disconnection occur frequently, which means weak services of the network layer. Sensed data is inaccurate which often has errors. Focusing on inaccuracy of the observed data and power limitation of sensors, this paper proposes a sampling frequency control algorithm and a data compression algorithm. Based on features of the sensed data, these two algorithms are combined together. First, it adjusts the sampling frequency on sensed data dynamically. When the sampling frequency cannot be controlled, data compression algorithm is adopted to reduce the amount of transmitted data to save energy of sensors. Based on the compressed data, we also propose an approximate query processing algorithm, which reduces query processing time dramatically. Experiments and analysis show that the proposed algorithms can decrease sampling times reduce the amount of transmitted data, save energy of sensors and improve the query efficiency.