Data Reconstruction Accuracy of Compressive Sleeping Scheme with Modified S-MAC for Body Sensor Networks

Today the main difficulty in Body Sensor Network (BSN) is the limited battery power and accuracy of the data. Some cases in health monitoring requires a long sensor life time such as health monitoring of critical patients, so that the data is needed by doctors or hospitals can be fulfilled. The Compressive Sleeping algorithm and the scheduling scheme using the Medium Access Control Sensor (S-MAC) are proposed in this research, to reduce power consumption. First, Compressive Sleeping algorithm is applied to select a several of the sensor to be activated and a several of it will go into sleeping mode, the selection is based on the sensor type and remaining battery of each sensor. The output of this algorithm is several sensors that are suitable for active. Furthermore, the active sensor will be scheduling data transmission using S-MAC, scheduling is based on the sensor priority and remaining batteries of each priority. Sensors that do not transmit data will go into temporary sleep mode, then the sensor will be reactivated if it gets a turn to transmit data to the fusion center (FC). The calculation of energy consumption is carried out on each process block. We calculated the accuracy of all reconstructed data in the FC using the Orthogonal Matching pursuit algorithm (OMP). The results of this research produce a good energy efficiency, that is, for the sensor selection ratio of 40%, the energy efficiency is 67.03 % and data accuracy is 95.5%.