Anti-Jamming Path Planning for Unmanned Aerial Vehicles with Imperfect Jammer Information

In this paper, we consider an anti-jamming problem for one unmanned aerial vehicle (UAV) reaching a given destination with maintaining the communication quality, where the ground-to-air communication link is spatially affected by an adversarial jammer. Note that the jammer's position and transmit power are initially unknown. To solve this problem, we propose a jammer detection-localization-evasion framework. Firstly, we employ a jammer detection method combining the Received Signal Strength (RSS) with Packet Error Rate (PER). After the jammer is detected, our designed RSS based jammer localization algorithm is presented, which can estimate the jammer position and jamming signal power simultaneously. However, localization noises often exist in practical scenarios, and therefore the estimated jammer information is usually with uncertainties. To overcome these uncertainties, we propose a robust path planning method of successful jammer evasion. More specifically, the proposed method can guarantee the worst-case performance so that the communication quality can be always better than a certain level under all possible uncertainties. Finally, a UAV communication jammed model is built on Network Simulator (NS3) and corroborates the effectiveness of our proposed method.