The Research On Application Of Compressed Sensing In Large Greenhouse Wireless Monitoring System

With the rapid development of science and technology, in order to improve the production efficiency of agriculture, the use of wireless monitoring system for greenhouse has been increased, which decreases the restriction of bad ecological environment to the development of agriculture and improves the level of agricultural production. There is a large number of sensor nodes in the wireless greenhouse monitoring system, and the sampling mechanism are based on the high sampling frequency of Nyquist sampling theorem, Therefore, the processing ability of the hardware is required to be very high, due to the limitation of hardware resources in wireless sensor networks, a large number of nodes are damaged, and the life cycle of the greenhouse monitoring system becomes shorter, which has a bad effect on the application and promotion of the intelligent Greenhouse in modern agricultural production. To reduce the number of data collection capacity and traffic, the author put forward the application of the compressed perception theory to the wireless greenhouse monitoring system, to replace high sampling frequency of Nyquist sampling theorem. In this paper, the main research contents is as follows:(1) The basic theory of compressed sensing is introduced. In particular, the key technologies are discussed in detail, including the sparse transform of signal, the measurement of signal compression and reconstruction of the signal.(2) The overall framework of greenhouse wireless monitoring system, hardware system and software system are designed. The greenhouse parameters by orthogonal matching pursuit algorithm (OMP) are compressed and refactored, and the best reconstruction property under the sparse degree of observed value by analysis is seeked.(3) As a single node has time correlation, and combines with the joint sparse model, a wireless sensor network in spatial and temporal correlation based on distributed compressed sensing model and the second of joint sparse model (JSM-2) applied in the greenhouse system is set up, and the coding and decoding scheme is put forward, more futher reduces the amount of data collection and traffic.(4) Finally, the above model and decoding algorithm are simulated by matlab experiments. First of all, The contrasting simulation experiments between One step greedy algorithm (OSGA) and Synchronous orthogonal matching pursuit algorithm (SOMP) based on JSM-2 are simulated and are compared to show SOMP is superior. By comparing OMP and SOMP under the condition of same reconstruction performance, SOMP needs less observed value. By energy consumption analysis and comparison, shows that compressed sensing needs only half of energy consumption compared with traditional transmission, and further shows the SOMP has more advantages in terms of energy saving...