A Geographic Routing Protocol for Flying Ad-Hoc Network Based on Gauss-Markov Mobility Prediction

Flying ad-hoc Network (FANET) is a self-organizing network composed of multiple Unmanned Aerial Vehicles (UAVs). Due to the high mobility of UAVs, the topology of FANET is highly dynamic. Therefore, it is important and challenging for FANET to design routing protocols suitable for the high mobility of UAVs. This article proposes a new reliable geographic routing protocol, named greedy perimeter stateless routing based on Gauss-Markov (GPSR-GM). Different from the GPSR protocol, GPSR-GM use Gauss-Markov Mobility Model (GMMM) to estimate the probability distribution of the position of neighbors. Based on the prediction results, this article proposes a more practical metric named weighted routing cost, which is used in GPSR-GM. Then, the MAC extract strategy is proposed, which can save the overhead associated with ARP by obtaining the MAC of neighbors through beacons. The simulation is carried out in the NS-3.35 to compare the impact of GPSR-GM and MAC extract strategy. The simulation results show that GPSR-GM increases the packet delivery ratio (PDR) and throughput by 5% to 10% over GPSR at the speed of 10 m/s. When the speed is increased, this improvement increases up to 20%. And the MAC extract strategy can improve the packet delivery ratio (PDR) and throughput by about 10%.