A Fast Neighbor Discovery Algorithm Based on Q-learning in Wireless Ad Hoc Networks with Directional Antennas

Abstract

Battlefield information interaction has high requirements for its effectiveness, but traditional algorithms are still inadequate in this respect. In this paper, the neighbor discovery process in wireless Ad hoc networks with directional antennas is discussed and an efficient neighbor discovery algorithm based on Q-learning theory is proposed. This paper takes traditional blind algorithm of all sectors scanning as the basis, then a fast neighbor discovery algorithm with the use of Q-learning is analyzed, which divides the neighbor discovery process into three stages, the initial stage without prior location information, the reinforcement learning stage, and the completion stage for mutual discovery in the shortest time. Finally, OPNET Modeler 14.5 is used to simulate this model, and the result show that the algorithm can improve the efficiency of neighbor discovery by nearly 86%.