A Q-learning-based routing scheme for smart air quality monitoring system using flying ad hoc networks

Air pollution has changed ecosystem and atmosphere. It is dangerous for environment, human health, and other living creatures. This contamination is due to various industrial and chemical pollutants, which reduce air, water, and soil quality. Therefore, air quality monitoring is essential. Flying ad hoc networks (FANETs) are an effective solution for intelligent air quality monitoring and evaluation. A FANET-based air quality monitoring system uses unmanned aerial vehicles (UAVs) to measure air pollutants. Therefore, these systems have particular features, such as the movement of UAVs in three-dimensional area, high dynamism, quick topological changes, constrained resources, and low density of UAVs in the network. Therefore, the routing issue is a fundamental challenge in these systems. In this paper, we introduce a Q-learning-based routing method called QFAN for intelligent air quality monitoring systems. The proposed method consists of two parts: route discovery and route maintenance. In the part one, a Q-learning-based route discovery mechanism is designed. Also, we propose a filtering parameter to filter some UAVs in the network and restrict the search space. In the route maintenance phase, QFAN seeks to detect and correct the paths near to breakdown. Moreover, QFAN can quickly identify and replace the failed paths. Finally, QFAN is simulated using NS2 to assess its performance. The simulation results show that QFAN surpasses other routing approaches with regard to end-to-end delay, packet delivery ratio, energy consumption, and network lifetime. However, communication overhead has been increased slightly in QFAN.