Wireless sensor networks have been widely used to collect information due to their flexibility in agricultural applications. However, the methodology of placing nodes in corn fields is ambiguous due to the complicated wireless communication and tough agricultural environments. This work provided a reference for other applications of WSN in crop fields.;Experiments of 2.4GHz radio propagation were designed and conducted as corn plants grew. Antenna height and growth stage of corn were considered as important factors. Three antenna height settings strategies were designed. The first height setting strategy was corresponding to a variable Fresnel zone as the changes of plant heights. The second height setting strategy kept 60% Fresnel zone clear during the development of corn plants. The third height setting strategy was set to close to corn heights corresponding to zero Fresnel zone. Growth stages included V6, V8, V10, V12, VT and R5. The first experiment was 3*5 factorial structure from V6 to VT. The second experiment was 2*2 factorial structure including the first two height setting strategies and two growth stages: VT and R5. Fundamental path loss modeling was conducted under each designed condition. There were three repetitions for each designed condition. Each repetition was to training path loss model along a transmission direction in a corn field. Three repetitions were corresponding to three transmission directions: angle, column, and row. Path loss models were established when antenna was set by height setting strategy two and corn plant was at VT.;The results indicated that the corn plant did not significant affect radio propagation when the size of clearance zone between a dominant path of radio and corn plants was consistent and the area has a significant effect on mean of path loss exponent from V6 to VT. This implied that plant did not have significant effect on the rate of energy decay when radio avoided obstacles at line of sight condition, but the area of Fresnel zone affected the rate of energy decay. The comparison between developed models with free space model showed the developed models were better to predict path loss than free space model.
展开▼
