Compressive sensing-based sequential data gathering in WSNs

For the temperature monitoring in Wireless Sensor Networks (WSNs), the paper puts forward a Compressive Sensing-based sequential data gathering framework. The framework first utilizes the covariance matrix to generate the sparsifying basis of sensory data. It then introduces the numerical sparsity to estimate the sparsity performance. The measurement matrix adopts sparse binary matrix and the number of measurements is bounded by the numerical sparsity. For each measurement, only parts of sensor nodes gather sensory data and transmit these data to the sink node for data recovery. The real temperature experiments demonstrate that the constructed sparsifying basis can make real temperature data approximately sparse. Compared with other types of the sparsifying bases, the constructed sparsifying basis can make the numerical sparsity of real temperature data be smaller and the recovery performance of sequential temperature data be better. Furthermore, total energy consumption of the proposed framework is less than that of other compressive data gathering algorithms.