A new wireless sensor network wave propagation model based on Zigbee protocol for protected mango greenhouse environment

The wireless sensor network (WSN) is the promising technology and it is widely used for monitoring and controlling the environmental conditions of precision agriculture. The deployment of the WSN nodes in real environments faces hard challenges of proper communication links and network coverage especially in the case of deployment of wireless nodes near the ground and the existence of dense vegetation which may impair the propagating signals. Modeling of wireless communication channel is important to achieve a successful implementation of WSN system in agricultural environment. In WSN, accurate propagation path loss models help for realization appropriate evaluation of the WSN performance, achieving more reliable communication, improving the power efficiency of the network nodes and decreasing the overall cost of the wireless network. There are many propagation path loss models used for modeling wireless communication channels, but most of them might not be suitable for the WSN applications due to propagation medium and the IEEE 802.15.4 standard. In this research, the WSN signal propagation path losses inside the mango greenhouse environment are investigated by using WSN based on the ZigBee standards. Various empirical measurements were conducted to examine the effect of each part of a tree on path loss with different transceivers’ heights to select the best antenna heights that adopted in all experiments for deriving the new path loss model. Indeed, a new propagation path loss model for greenhouse environment (Greenhouse Propagation Path Loss Model - GHPLM) is derived based on a regression technique. This new model is used for computing the total propagation path losses and for deployment the wireless sensor nodes in the real field based on the maximum separation distance measurements. The outcomes from this work proved that the antenna heights and the vegetation depth are the two most important factors in channel modeling. The empirical results emphasize that the Plane Earth (PE) model is inaccurate for predicting path loss in real environments due to it is based on simplistic approaches and considered to be very optimistic in real propagation scenarios as the case in mango greenhouse environment. Thus, the combination of this model with the vegetation path loss model contribute more convincing results and can best describe the behavior of actual WSN systems when deployed in a real environment. The empirical results proved that the GHPLM model is the best candidate compared to other existing empirical propagation path loss models. The Mean Absolute Percentage Error (MAPE) that measured the difference between the actual and prediction path loss was 3.96% for the new GHPLM model compared to other vegetation path losses which were 44.55%, 41.07%, 31.82% and 15.48% for Weissberger, ITU-R, FITU-R and COST235 models respectively.