Agricultural Irrigation Recommendation and Alert (AIRA) system using optimization and machine learning in Hadoop for sustainable agriculture

Internet of Things (IoT) in the field of agriculture promises to continuously provide global access to the farming information. The smart agriculture system either gives alert regarding the farm or it recommends for best agriculture field. This paper addresses both irrigation and alert, i.e., Agricultural Irrigation Recommendation and Alert (AIRA) system that operates individually without any correlation. At first, the IoT users of each farm field registers in HDFS, i.e., Hadoop Distributed File System. All the registered farm field holders will receive alerts for water level status and others. The collected data will be processed in a hybrid classifier that combines k-nearest neighbor with a neural network (k-N⁴). The classifier classifies into five classes of irrigation alerts: low water level, high water level, maintained water level, low pressure, and cyclonic storm. For faster classification, firstly, the neural network is used. Secondly, the recommendation for agronomists is optimal. The collected data is clustered by modified fuzzy clustering, and then optimal weather conditions are recommended from attractiveness-based particle swarm optimization (APSO) algorithm. The main measurements taken into account from the farms are soil moisture, temperature, humidity, wind speed, and intensity. Also, the access for IoT users is authenticated with identity, password, and biometric. Here, biometric iris is used, which is more secure than the fingerprint. Furthermore, data security is assured based on M-RSA cryptography.