Application research of compressed sensing in wireless sensor monitoring networks on coal mine

In wireless sensor monitoring networks on coal mine, vast data processing in nodes will result in the situation that some nodes lose energy at a higher rate and die much faster than others. To solve this problem, this paper proposes a novel method of data acquisition and transmission with high efficiency, in which theory of compressed sensing (CS) is applied to wireless sensor nodes. First, Gaussian random matrix is used as the measurement matrix to sample sparse signal. Then the signal is transmitted to the monitoring center via Wi-Fi and wired optical fiber. Finally, ROMP algorithm whose reconstruction error is less than others is chosen to reconstruct the original signal. In this paper, simulation experiments on the temperature and humidity are carried out through MATLAB. Theoretical analysis and simulation experimental results show that the reconstructed signal in monitoring center is near perfect which reduces the channel bandwidth pressure. Meanwhile, the method saves energy, balances energy consumption and prolongs the lifetime of network significantly. © 2015 ISSN.