A Stochastic Packet Forwarding Algorithm in Flying Ad Hoc Networks: Design, Analysis, and Evaluation

Over the past decade, unmanned airborne vehicles (UAVs; widely known as drones) are quickly being deployed in various civilian as well as military applications. Drones can self-organize into a connected swarm (Flying Ad Hoc Networks– FANETs) to complete various challenging missions. As the primary building block of Internet-of-Drones (IoD), FANETs have an important role to play in governing the autonomous movement of drones and supporting drone-to-everything (D2X) communications. However, factors such as flying characteristics and the highly dynamic topology of FANETs compound the challenges of packet forwarding; thus, the focus of this article. Specifically, a stochastic packet forwarding algorithm (hereafter referred to as
SPA
) is proposed for FANETs, where the data packets can be efficiently transferred to the destination. In
SPA
, the packet sender evaluates each next hop forwarding candidate drone based on different network metrics, prior to selecting the next hop forwarding drone according to the forwarding probability. In addition, an analytical model is developed to evaluate the performance of
SPA
. Then, the link expiration time and link throughput are chosen as network metrics for network simulation experiments, where
SPA
is evaluated and compared with prior schemes (e.g,
$DTN_{geo}$
and
GeoUAVs
). The performance of
SPA
is evaluated in a real-world testbed to complement the network simulation experiments. Findings from the evaluations demonstrate that
SPA
obtains better performance in terms of packet delivery ratio, packet delivery latency, the number of delivered data packets, average link lifetime, and bit error rate.