Energy-efficient Data Aggregation Scheme For Wireless Sensor Networks

A wireless sensor network (WSN) is an ad hoc network consisting of a large number of inexpensive sensor nodes. The nodes in a WSN cooperatively sense and monitor environmental information, such as temperature, humidity, pressure, concentration of pollutants, etc. This information is delivered to the sink node, through which it is delivered to users via WAN. Sensors are usually powered by batteries so that they are extremely energy-constrained it is not convenient or even infeasible to replace their batteries. As a result, energy conservation is of great concern for WSNs to run longer.In a WSN, some nodes need to aggregate the sensed data, in order to reduce the amount of data transmission, reduce energy consumption, and prolong network lifetime. This paper proposes a spatial-temporal data aggregation scheme based on compressive sensing (CS). The main contributions are as follows:(1) Encodes the sensed data using CS, so as to limit the amount of transmitted data within a threshold, and reduce the energy of data transmission.(2) Proposes a routing algorithm based on inter-cluster chaining (ICC) to build the aggregation tree, which approximately optimizes the energy consumption of the "hotspot" area, and prolongs network lifetime.(3) In order to build the spatial-temporal sparsity basis for the process of CS decoding, uses lifting scheme wavelet transform for space domain, which involves a polynomial predictor, for sparsifying the spatial signal of randomly-deployed sensor nodes; and uses discrete cosine transform for time domain, which further sparsifies the transformed data in space domain. The sparsity basis guarantees the accuracy of signal reconstruction. Simulation results show that, in comparison to the traditional data gathering scheme, the proposed scheme can significantly reduce energy consumption, mitigate the "hotspot" problem, and prolong network lifetime on the premise of guaranteeing the accuracy of signal reconstruction.