Energy Efficient Data Dissemination for Large-Scale Smart Farming Using Reinforcement Learning

Abstract

Smart farming is essential to increase the crop production for which IOT in agricultural parameters is crucial for the growth of crops. To achieve this, modern technology is required to enable accuracy in fertilizing, watering and adding pesticides to the crops as well as monitoring the conditions of environment. Now a days, more and more sophisticated sensors are developed but to achieve this on larger scale managing them efficiently is very significant. We want to achieve sustainability in large scale farms by improving communication between the Wireless Sensor Network nodes and the Base Station through monitoring the energy and communication of sensor nodes through machine learning algorithms. The idea is to make multihop communication efficient, by balancing the energy consumption of sensors. The nodes with more energy will be used as relay to transmit data over the distance. The path selection will done based on remaining energy of the sensor nodes. Reinforcement learning is proposed to select best paths among the fields towards the base station. Reinforcement Learning is the area of machine learning in which is concerned with how involved agents are supposed to take actions in a specified environment to maximize the reward and to achieve a common goal. Reinforcement learning. In our network, a large number of sensors are deployed on large scale fields, reinforcement learning is used to find the best path towards a base station. After a number of successful paths have been developed, they are then used to transmit the sensed data from the fields. The simulation results have shown better performance over shortest paths and broadcasting techniques that were tested against reinforcement learning.