Modeling the Performance of IEEE 802.11ac WLAN Network for Remote Visual SLAM

The paper presents the results of analyzing existing approaches and models to estimate bandwidth and delays in the wireless channel, as well as energy consumption and maximum operating time of a robot/unmanned aerial vehicle (UAV) when using visual simultaneous localization and mapping (SLAM) algorithms with remote data transmission and processing via an IEEE 802.11ac wireless network. Simulation modeling for estimating channel throughput, mean value of delay, packet delivery ratio, energy and power cost for transmission (based on the considered models and approaches), and the maximum possible flight time of a UAV in hovering mode when transmitting data between the robot and the conditional remote data processing server was carried out. In general, data pre-processing on the UAV side (feature identification, extraction, and coding, etc.) and sending the resulting data to the server for further processing and retrieval of keyframes/map updates should be considered for practical field tests of remote visual SLAM algorithms.