A Windowed Stochastic Learning Automaton Algorithm for Directional Neighbor Discovery
Wenxiang Bai, Yongchao Liu, Wei Zhang, Lijuan Zhang · 2023
In recent years, unmanned aerial vehicles (UAVs) have been widely deployed to form temporary flying ad hoc networks (FANETs) for sensor data collection and monitoring assistance. Neighbor discovery as the bootstrapping step to form a self-configured network plays a critical role in FANETs. With fully directional antennas used in UAVs, neighbor discovery is more challenging because of the beam alignment requirement. In this work, an efficient windowed stochastic learning automaton (WSLA) algorithm is proposed to accelerate the neighbor discovery process for directional FANETs. With the stochastic estimator and time window-based optimization method, WSLA is capable of adapting to changing environments. Simulation results demonstrate that the proposed algorithm has shorter convergence time than the best-performing related works.