International Journal of Science, Engineering and TechnologyNode Position Based Multi – Hop Routing For Fanets
Authors- Dhiren Kumar
Abstract- In this work, we provide a stateless position-based packet routing method for a flying ad-hoc network (FANET). The goal of this project is to collect all available geographical data about the network. The node is directed to its target using its address and coordinates. Here, we introduce UAV technology to improve position-based multicarrier transmission in 3D spaces. The DRL algorithm is used by this ad hoc network while in flight. Together, DRL and LAR ensure timely arrival at the final destination. In this research, we present a decentralized intelligent routing approach using deep reinforcement learning (DRL) that takes into account the state of symmetrical nodes between two hops. The approach takes into account the location, velocity, load degree, and connection quality of nodes in the establishing process of state components, allowing for a more complete understanding of the local dynamic of the network. Using the Q values computed by the model during training of Deep Q-Networks, the nodes may do an adaptive neighbor selection. The simulation and analytical results demonstrate that the suggested technique has superior performance to various widely used methods and offers strong convergence characteristics.
