Compressed Sensing Based Data Retrieval And Gathering In Low Duty Cycle Wireless Sensor Networks

Wireless sensor networks are gaining their popularity in practical industry and academical research. Many WSNs are widely deployed for forest, ocean and mountain monitoring and surveillance. A key issue constraint the usage of WSNs that they are limited by the bottleneck of battery energy. To deal with the energy drain caused by idle listening, many WSNs work in low duty cycle mode. But the low duty cycle model also decreases the transmission efficiency of WSNs and makes the network more complex. In our research, we proposed new method to gather global data efficiently by exploiting compressive sensing. In our approach, global data can be retrieved from any node instead of a fixed sink. We designed a method of measurement merging to improve the efficiency of the network. The method of sequential observation in compressed sensing is also improved to fit the low duty cycle network. With extensive simulations, our approach is approved to be efficient, accurate and robust.