Abstract
In this article, we present a supervised machine learning multipath and administrative-distance-based load balancing algorithm for an Agriculture Internet of Things (AG-IoT) network. The proposed algorithm is known as an artificial intelligence or simply AI-enabled multihop and administrative-distance-based opportunistic routing (MHADBOR) algorithm, which processes the collected information from source to the destination by means of multihop count and administrative-distance-based communication infrastructure in the network. Beside that, we used cluster heads (CH), microbase stations (\boldsymbol {\Re BS}ℜBS), and macrobase stations (\boldsymbol {\aleph BS}ℵBS) in the network with a frequent rate to effectively utilize the administrative distance while managing the deployed network traffic in a congestionless communication environment. In addition, the MHADBOR algorithm empowers the participating devices to practice the administrative distance rather than hop count communication when they are in the vicinity of network special components, e.g., CH and \boldsymbol {\Re BS}ℜBS outcome statistics of the MHADBOR algorithm in the simulation environment exhibit an extraordinary improvement in contention, congestion, communication, and computing costs, accompanied by throughput and end-to-end (E2E) delay and packet loss ratio in the deployed AG-IoT network.