Packet Delay in Wi-Fi Networks of Unmanned Aerial Vehicles

An important but challenging problem in multiple unmanned aerial vehicle (UAV) cooperative control systems is how the UAVs should be networked over the wireless medium so that information can be transmitted in near real-time. In this memorandum, we conduct a statistical analysis of the packet end-to-end delay in a Wi-Fi network of UAVs. In the network, each UAV runs the distributed coordination function (DCF) of research 802.11 at the medium access control (MAC) layer. All UAVs are one-hop neighbors, and a pair of UAVs can communicate over the wireless channel using a fixed data rate. A light traffic condition is used (i.e., in each UAV, data packets are generated one at a time and the interval between two successive generation times is described by independent random variables following an exponential distribution with some fixed mean). By modeling each UAV to be an M/M/1 queueing system, we first derive the mean packet end-to-end delay under light traffic conditions. Numerical and simulation results show that the mean packet end-to-end delay derived in this memorandum is accurate for Wi-Fi networks under the light traffic condition. It increases with either the number of UAVs in the network or the packet generation rate. In addition, existing results in the literature, based on the saturated traffic condition (i.e., packets are always supplied for transmission), tend to overestimate by a large amount the mean packet end-to-end delay for networks with light traffic. In the second part of the document, we apply simulation data to the analysis of the probability distribution function of the packet end-to-end delay. Using a distribution fitting tool, we observe that the packet end-to-end delay can be better approximated by the sum of a deterministic delay and a random delay. The deterministic delay corresponds to the time period during which the node senses the medium and is able to perform a successful transmission, while the random delay follows a Gamma distribution function.


COMMENT Uncategorized