Optimally sensors nodes selection for adaptive heterogeneous precision agriculture using wireless sensor networks based on genetic algorithm and extended kalman filter.

Precision agriculture (PA) has recently employed as an innovative agriculture to increase the profitability through enhancing the crop yields. Wireless sensor networks (WSNs) is widely used in PA to measure the agricultural parameters. However, the main challenges of using WSNs in PA are to improve the measurement accuracy and to increase the network lifetime of WSNs. A novel scheme is developed in this paper to measure soil moisture and temperature in heterogeneous farms called Adaptive heterogeneous precision agriculture (AHPA). As well as the Extended kalman filter (EKF), AHPA contains two invented algorithms, named Optimally sensor selection algorithm (OSSA) and Adaptive sampling interval algorithm (ASIA). As part of the EKF, agricultural state is proactively predicted and then it is updated through measuring it using WSN so that the noises associated with sensor measurements are eliminated. OSSA employs the Genetic algorithm (GA) for selecting a group of sensor nodes that are used to sense the agricultural state, in order to enhance network lifetime and crop yields. The root system is modeled and used in the OSSA to select the nodes inside the area covered by the root system. In the prediction stage, the ASIA is used to calculate an adaptive sampling interval based on the predicted state of agricultural parameters. The simulation results show that the total energy consumption and the reduction of the agricultural parameters are improved in the proposed ASIA, compared with other fixed sampling interval schemes. Moreover, the number of dead nodes, the time of the first node death, the energy balance, the neighbors count, and the number of evolved nodes in sensing activities are enhanced in the proposed OSSA, compared with other schemes. Hence, the network lifetime and crop yields are enhanced. Moreover, the proposed OSSA selects the sensor nodes based on the root age so that the agricultural parameters of the whole soil of the plants are sensed. [ABSTRACT FROM AUTHOR]