Research On Compressed Sensing Based Scheduling Algorithms In Low Power Body Area Networks

In recent years, the rapid development of E-healthcare systems ofWireless Body Area Networks (WBAN) have shown tremendous potentialto promise enormous change in future health-care industry. The systemgenerally consists of multiple sensor nodes that monitor various medicalsignals and deliver data to a network coordinator for further processing.One of the major issues of such systems is the energy eiffciency, owingto the limited power supply and diiffcult replacement of batteries in someimplanted sensors.According to the network hierarchy of Body Area Network, the Energy-saving mechanisms should be considered in both physical layer and datalink layer. The emerging compressed sensing (CS) technology enablessenor nodes to employ a much lower sampling rate than Nyquist while en?suring the accuracy of reconstruct signals. Meanwhile, CS is also easierto implement than the conventional compression method. In this paper,we use compressed sensing in the acquisition of biomedical signals, andcompare the performance of different measurement matrices and sparsebases using ECG as an example. An experimental platform was built tovalidate the results. We also analyze the energy eiffciency of compres?sive sampling which includes the energy-saving effects under different re?quirements of reconstructed signal and energy consumption of computing compressed sensing. Under different circumstances, with the result of us-ing compressed sensing or not, we present an energy efficient cross-layer design and propose a slot allocation algorithm based on the design, which effectively improves the slot utilization. On the other hand, we investi-gate the scheduling problem with Lyapunov optimization technique and propose a dynamic scheduling policy including sleep scheduling and op-portunistic transmission, which takes into account time-varying channel condition. A tradeoff between the energy optimality and the average net-work congestion can be obtained by appropriately adjusting the weighting factor. We demonstrate that our policy can push the energy consumption arbitrarily close to the global minimum solution while sustaining the net-work stability.