Compressive Sensing Based Energy-Efficient Random Routing in Wireless Sensor Networks

Using random walk (RW) to collect data in wireless sensor networks (WSN) has been proven to be an energy-efficient method. The integration between compressive sensing (CS) and RW provides some different points of view about data collection in WSNs. Based on a small certain number of CS measurements (M), all raw sensor readings from N nodes can be recovered at the base-station (M << N). In this paper, we analyze RW based on theory and practice. We investigate the trade-off between exploring the measurement matrix and setting up the length for RWs to achieve the energy consumption smallest for WSNs. In addition, we formulate the total power consumption that contains the average consumed energy of each RW and the average consumed energy to send measurements in multi-hop through intermediate nodes from RWs to the base-station (BS). We analyze and suggest the optimal case for the networks to spend the lowest energy that helps sensors to prolong their lifetime or the network connection.