Collision risk assessment in Flying Ad Hoc aerial wireless networks

Abstract

A stochastic model is introduced that accurately models collisions in Flying Ad Hoc wireless Networks (FANETs) where Unmanned Aerial Vehicles (UAVs) are flying within the same shared 3D airspace. The model has two input parameters, the number of flying UAVs and the average time for an arbitrary UAV to come into contact with another UAV (the inter contact time). Using only these two parameters, we provide simple, yet accurate closed-form expressions for different collision related metrics such as safety periods, survival probabilities and number of collisions. Two scenarios are particularly considered. The first assumes that UAVs are equipped with perfect collision detection and avoidance capabilities. This is essentially to answer the question of whether detection and resolution tools are required for a given UAVs fleet to accomplish its mission. The second scenario assumes that small UAVs cannot satisfy the requirements of collision detection and avoidance equipage due to their size, weight or power constraints.While the number of deployed UAVs is easily known, the setup of the inter contact time is rather problematic for the end user (space controller) of the system. To this end, we developed a generic simple expression of the inter contact time to ease the administration task. This generic expression is then instantiated to two mobility models.Extensive simulations based on OMNeT++ are used to validate the obtained analytical results. The simulation results are shown to be in a remarkable agreement with those of the conducted mathematical analysis.